Routledge Handbook of Early Chinese History 1138775916, 9781138775916

Table of contents :
Cover
Title
Copyright
CONTENTS
List of figures
List of tables
List of maps
Chronology
Acknowledgements
Contributors
Introduction: what is early Chinese history?
PART I Chronology
1 Main issues in the study of the Chinese Neolithic
2 Of millets and wheat: diet and health on the Central Plain of China during the Neolithic and Bronze Age
3 The Bronze Age before the Zhou dynasty
4 The Western Zhou state
5 The age of territorial lords
6 The Qin dynasty
7 The Former Han empire
8 The Latter Han empire and the end of antiquity
PART II Topical studies
9 The Old Chinese language
10 Early Chinese writing
11 The spirit world
12 Religious thought
13 Political thought
14 Food and agriculture
15 Warfare
16 Currency
17 Women in early China: views from the archaeological record
18 An overview of the Qin-Han legal system from the perspective of recently unearthed documents
19 Literature
20 Art
21 ‘Medicine’ in early China
22 Mathematics
23 Astronomy
Index

Citation preview

ROUTLEDGE HANDBOOK OF EARLY CHINESE HISTORY

The study of early China has been radically transformed over the past fifty years by archaeological discoveries, including both textual and non-textual artefacts. Excavations of settlements and tombs have demonstrated that most people did not lead their lives in accordance with ritual canons, while previously unknown documents have shown that most received histories were written retrospectively by victors and present a correspondingly anachronistic perspective. This handbook provides an authoritative survey of the major periods of Chinese history from the Neolithic era to the fall of the Latter Han empire and the end of antiquity (ad 220). It is the first volume to include not only a comprehensive review of political history but also detailed treatments of topics that transcend particular historical periods, such as: • • • •

Warfare and political thought Cities and agriculture Language and art Medicine and mathematics

Providing a detailed analysis of the most up-to-date research by leading scholars in the field of early Chinese history, this book will be useful to students and scholars of Chinese history, Asian archaeology, and Chinese studies in general. Paul R. Goldin is Professor of East Asian Languages & Civilizations at the University of Pennsylvania, USA. His recent publications include the Dao Companion to the Philosophy of Han Fei (2012) and A Concise Companion to Confucius (2017).

ROUTLEDGE HANDBOOK OF EARLY CHINESE HISTORY

Edited by Paul R. Goldin

First published 2018 by Routledge 2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN and by Routledge 711 Third Avenue, New York, NY 10017 Routledge is an imprint of the Taylor & Francis Group, an informa business © 2018 selection and editorial matter, Paul R. Goldin; individual chapters, the contributors The right of Paul R. Goldin to be identified as the author of the editorial material, and of the authors for their individual chapters, has been asserted in accordance with sections 77 and 78 of the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging-in-Publication Data A catalog record has been requested for this book ISBN: 978-1-138-77591-6 (hbk) ISBN: 978-1-315-77360-5 (ebk) Typeset in Bembo by Apex CoVantage, LLC

CONTENTS

CONTENTSCONTENTS

List of figuresviii List of tables xi xii List of maps Chronologyxiii Acknowledgementsxiv Contributorsxv Introduction: what is early Chinese history? Paul R. Goldin

1

PART I

Chronology13   1 Main issues in the study of the Chinese Neolithic Gideon Shelach-Lavi   2 Of millets and wheat: diet and health on the Central Plain of China during the Neolithic and Bronze Age Kate Pechenkina

15

39

  3 The Bronze Age before the Zhou dynasty Robert Bagley

61

  4 The Western Zhou state Li Feng

84

v

Contents

  5 The age of territorial lords Chen Shen

108

  6 The Qin dynasty Charles Sanft

146

  7 The Former Han empire Vincent S. Leung

160

  8 The Latter Han empire and the end of antiquity Wicky W.K.Tse

180

PART II

Topical studies197   9 The Old Chinese language Axel Schuessler

199

10 Early Chinese writing Luo Xinhui; tr. Zachary Hershey and Paul R. Goldin

217

11 The spirit world Jue Guo

229

12 Religious thought Ori Tavor

261

13 Political thought Yuri Pines

280

14 Food and agriculture Roel Sterckx

300

15 Warfare Wicky W.K.Tse

319

16 Currency François Thierry

336

17 Women in early China: views from the archaeological record Anne Behnke Kinney

367

vi

Contents

18 An overview of the Qin-Han legal system from the perspective of recently unearthed documents Kyung-ho Kim and Ming-chiu Lai

386

19 Literature Stephen Durrant

405

20 Art Wang Haicheng

425

21 ‘Medicine’ in early China Miranda Brown

459

22 Mathematics Karine Chemla

473

23 Astronomy David Pankenier

493

Index517

vii

FIGURES

FIGURESFIGURES

1.1 The locations of all sites mentioned in the chapter  1.2 Typical stone artifacts of the late Pleistocene and early Holocene period  1.3 A. Map of the Chahai site; B. Storage pit from the Kuahuqiao site with acorns in it; C. Bone spades from Hemudu  1.4 A. Longshan eggshell ceramic cups; B. A Liangzhu jade cong; C. A wealthy Liangzhu grave from the Sidun site  1.5 A. Grave M1 from the Xizhufeng site; B. Some of the ceramic artifacts found in grave M1  1.6 Grave M564 from the Liuwan cemetery and the grave goods found in it  2.1 Locations of archaeological sites discussed in the chapter  2.2 δ13C and δ15N values of human bone collagen samples from Yangshao sites  2.3 δ13C and δ15N values of human bone collagen samples from Yangshao sites compared to isotopic values of human collagen samples from other cultures and periods  2.4 δ13C and δ15N values of human bone collagen samples from Eastern Zhou sites  3.1 Stone relief in the Temple of Sety at Abydos, ca. 1300 bc, showing Sety I and his son Prince Ramesse revering a list of their predecessors  4.1 Bronze deer from Shaigushan 石鼓山 in Baoji  4.2 The Lai pan 逨盤, vessel and inscription  4.3 Bronzes newly discovered in the Yejiashan 葉家山 cemetery  4.4 Rules of Name Differentiation I and II  5.1 A. Hu ritual wine vessel; B. The inscription from the neck of the vessel  5.2 One of a pair of Biao Qiang bells (A) and the rubbing of its inscription (B)  5.3 Roof tile with character Shang  5.4 Dragon pendant  5.5 Fu food-container (A), detail of the lid with inscription (B), and rubbings of inscriptions (C)  5.6 City layout of the Yanxiadu ruin  5.7 Ridge tile (A) and detail (B)  5.8 Pictorial Hu wine vessels  viii

16 19 23 27 28 31 40 46

50 54 77 91 94 97 102 110 115 116 120 123 127 131 134

Figures

5.9 A full-scale drawing illustrating descriptive scenes from the pictorial bronze Hu wine vessel  5.10 Detailed imagery from the pictorial bronze Hu wine vessel  5.11 A newly discovered battlefield site in Longhu, near Xinzheng  16.1 Sea-cowrie  16.2 Bronze cowrie, dukedom of Jin (seventh–sixth c. bc)  16.3 Hollow-handled spade, Henan (sixth c. bc)  16.4 Hollow-handled spade, dukedom of Jin (sixth c. bc)  16.5 Pointed knife, northeast China (sixth c. bc)  16.6 Square-feet spade, Zhao, city of Anyang (fourth–third c. bc)  16.7 Yi knife of Yan, or Mingdao (fourth–third c. bc)  16.8 Qi knife, Qi fahua (fourth–third c. bc)  16.9 Yi si hua of Qi (fourth–third c. bc)  16.10 Early spade of Wei, city of Anyi (fourth c. bc)  16.11 Early Banliang of Qin (ca. 378–360 bc)  16.12 Banliang of Duke Xiao of Qin (360–338 bc)  16.13 Wufen qian of Empress Gaohou of the Western Han (182 bc)  16.14 Four zhu banliang of Han Wendi (175 bc)  16.15 Baijin sanpin of Wudi, dragon design coin (119 bc)  16.16 Junguo wuzhu (118–113 bc)  16.17 Chice wuzhu (115–113 bc)  16.18 Sanguan wuzhu (from 113 bc)  16.19 Gold inlaid knife of Wang Mang (ad 9–10)  16.20 Great cash of 50, daquan wushi of Wang Mang (ad 7–14)  16.21 Small cash worth 1, xiaoquan zhiyi of Wang Mang (ad 10–14)  16.22 Spade of 1000, dabu heng qian of Wang Mang (ad 10–14)  16.23 Huoquan of Wang Mang (ad 14–23)  16.24 Huobu of Wang Mang (ad 14–23)  16.25 Wuzhu of Eastern Han (Guangwudi, ca. ad 40–58)  20.1 Earthenware bottle painted in black, from Gansu Lanzhou Xinghetai  20.2 A. Black earthenware goblet; B. This electron microscope image shows the interior of a shard from another eggshell pot  20.3 Jade from Shaanxi Fengxiang Shangguodian  20.4 Jade cong from Zhejiang Yuhang Fanshan  20.5 Bronze you from Hubei Huangpi Panlongcheng  20.6 Bronze altar set from Shaanxi Baoji Shigushan  20.7 Bronze zun from Shanxi Yicheng Dahekou  20.8 Bronze hu from Shaanxi Meixian Yangjiacun  20.9 A. Bronze hu cast at the Houma foundry in Shanxi; B. Clay pattern block, also from Houma  20.10 A. Bronze zun and pan set from the tomb of Marquis Yi of Zeng at Hubei Suizhou; B. Close-up of openwork in the zun created by the lost-wax process  20.11 Drawing of a pattern repeat on a piece of polychrome weave from Hubei Jiangling Mashan  20.12 A. Lacquer box from Hubei Jingzhou Baoshan; B. and C. Detail of the register on the rim of the box’s lid  ix

135 136 141 337 340 341 343 344 345 346 348 349 350 351 351 354 354 355 356 357 357 359 359 360 360 361 361 362 426 429 430 431 435 437 439 442 444

446 448 449

Figures

20.13 A. Bronze chariot canopy shaft fitting with inlays of gold, silver, and semiprecious stones from Hebei Dingxian Sanpanshan; B. Detail of inlay  20.14 Rubbing of the west wall of the Xiaotangshan shrine in Shandong Changqing  20.15 An inventory from Gansu Juyan, ink on wooden slips tied with cords, each slip ca. 13.5 cm  22.1 The execution of a multiplication and a division, according to the Mathematical Canon by Master Sun. Digits written with rods are replaced by Hindu-Arabic numerals  23.1 Star chart showing the Ding constellation perpendicular to the horizon in late fall mid-seventh century bce, at the time when this accurate alignment technique is documented in the Book of Odes  23.2 The conjunction in late May 1059 bce at the “beak” of the Vermilion Bird  23.3 The Supernal Lord in imperial garb attended by spirit officials and driving the Dipper like a carriage 

x

451 454 456

486

500 501 508

TABLES

LIST OF TABLESLIST OF TABLES

1.1 The chronological and geographic framework of this chapter and some of the main associated Neolithic archaeological ‘cultures’  2.1 Stable isotope values in human bone collagen samples from the Yangshao archaeological sites  2.2 Stable isotope values in human bone collagen samples from the Neolithic and early dynastic archaeological sites  2.3 Frequency of carious teeth in Yangshao skeletal collections  2.4 Frequency of porotic hyperostosis and cribra orbitalia in Yangshao skeletal collections  2.5 Stable isotope values in human bone collagen samples from the Eastern Zhou archaeological sites  2.6 Animal stable isotope data from Neolithic and Bronze Age archaeological contexts of the Chinese Central Plain  4.1 Dates of Western Zhou Kings  9.1A 魚部 yúbù *-a  9.1B 鐸部 duóbù *-ak  9.2 歌部 or 戈部 gēbù *-ai  18.1 Overview of punishments under Qin law  20.1 Distribution of pictorial themes in the Xiaotangshan shrine  20.2 Pictorial themes in the Xiaotangshan shrine and their symbolisms  23.1 The four cardinal emblems and twenty-eight lunar lodges 

xi

17 45 47 50 51 53 55 85 202 202 203 396 454 455 503

MAPS

LIST OF MAPSLIST OF MAPS

4.1  The Zhou central region in the Wei River Valley, Shaanxi  4.2  Geography of the Western Zhou state 

xii

86 87

CHRONOLOGY

CHRONOLOGYCHRONOLOGY

Precise dates are not always ascertainable (as in the case of the Shang and Zhou kings), and there is not always a consensus regarding the beginning and ending years of conventionally named historical periods (such as Warring States). The dates of the Zhou kings in this chart are taken from the Xia Shang Zhou Chronology Project (Xia Shang Zhou duandai gongcheng 夏商周斷代工程) of the People’s Republic of China.The names of Kings Yì 懿 and Yí 夷 are distinguished in Modern Mandarin by tone.The import of the name Gong He 共和 is disputed, but it probably refers to Gong Hefu 共 龢父 (i.e., Hefu of Gong), who assumed control of the government in the mid-ninth century bce. Shang dynasty 商 Zhou dynasty 周 Western Zhou 西周 King Wu 武王 King Cheng 成王 King Kang 康王 King Zhao 昭王 King Mu 穆王 King Gong 共王 King Yì 懿王 King Xiao 孝王 King Yí 夷王 King Li 厲王 Gong He 共和 King Xuan 宣王 King You 幽王 Eastern Zhou 東周 Spring and Autumn 春秋 Warring States 戰國 Qin dynasty 秦 Han dynasty 漢 Western Han 西漢 Xin dynasty 新 Eastern Han 東漢

xiii

?–1046 bce 1046–256 bce 1046–771 bce r. 1046–1043 bce r. 1042–1021 bce r. 1020–996 bce r. 995–977 bce r. 976–923 bce r. 922–900 bce r. 899–893 bce r. 892–886 bce r. 885–878 bce r. 877–841 bce 841–828 bce r. 827–782 bce r. 781–771 bce 770–256 bce 722–481 bce 453–221 bce 221–207 bce 206 bce –220 ce 206 bce –9 ce 9–23 ce 25–220 ce

ACKNOWLEDGEMENTS

ACKNOWLEDGEMENTSACKNOWLEDGEMENTS

The editor would like to thank Deven M. Patel and Yuri Pines for helpful suggestions as he was composing the Introduction. All of us are indebted to the contributors, not only for their scholarship but also for their unstinting patience in the face of the many obstacles that delayed the publication of this book. We hope it has been worth the trouble.

xiv

CONTRIBUTORS

CONTRIBUTORSCONTRIBUTORS

Robert Bagley is Professor Emeritus of Art and Archaeology at Princeton University. He has written books and articles on Neolithic and Bronze Age art, archaeology, and metallurgy, on the origin of the Chinese writing system, and on pre-Han music theory. His most recent books are Max Loehr and the Study of Chinese Bronzes (2008) and Gombrich Among the Egyptians (2015). Miranda Brown is Professor of Chinese studies in the Department of Asian Languages and Cultures, University of Michigan. She is the author of The Art of Medicine in Early China: The Ancient and Medieval Origins of a Modern Archive (2015) and The Politics of Mourning in Early China (2007) and the co-editor of Fragments: Interdisciplinary Approaches to Ancient and Medieval Pasts. Karine Chemla is Senior Researcher, French National Center for Scientific Research (CNRS), group SPHERE (CNRS & University Paris Diderot), and from 2011 to 2016, Principal Investigator of the ERC Advanced Research Grant “Mathematical Sciences in the Ancient Worlds” (SAW, https://sawerc.hypotheses.org). She researches the history of mathematics in ancient China and modern mathematics in Europe from a historical anthropology viewpoint. Chemla published, with Guo Shuchun, Les neuf chapitres (2004). She edited The History of Mathematical Proof in Ancient Traditions (2012) and, with J. Virbel, Texts, Textual Acts and the History of Science (2015); with R. Chorlay and D. Rabouin, The Oxford Handbook of Generality in Mathematics and the Sciences (2016); and with Evelyn Fox Keller, Cultures Without Culturalism: The Making of Scientific Knowledge (2017). Stephen Durrant is Professor Emeritus at the University of Oregon. He specializes in early Chinese language and literature, with a particular interest in Chinese narrative and historiography. His publications include The Cloudy Mirror: Tension and Conflict in the Writing of Sima Qian (1995), The Siren and the Sage: Knowledge and Wisdom in Ancient Greece and China (with Steven Shankman, London, 2000), The Letter to Ren An and Sima Qian’s Legacy (with Wai-yee Li, Michael Nylan, and Hans van Ess, Seattle, 2016), and a three-volume translation of the Chinese classic Zuozhuan with Wai-yee Li and David Schaberg (Seattle, 2016). Jue Guo is Assistant Professor of pre-modern Chinese Humanities and Civilizations at Barnard College. She specializes in ritual practices, material culture, and social, religious, and cultural xv

Contributors

history of Early China. She has published “Divination” in The Blackwell Companion to Chinese Religions (2012) and “Concepts of Death and the Afterlife Reflected in Newly Discovered Tomb Objects and Texts From Han China” in Mortality in Traditional Chinese Thought (2011). Kyung-ho Kim received his Ph.D. from Sungkyunkwan University. He is currently a professor of the Academy of East Asian Studies at Sungkyunkwan University. A specialist in political and social history of Warring States and Qin-Han periods, he is currently studying Warring States and Qin-Han bamboo and wooden documents. His works include “A Study of Excavated Bamboo and Wooden-strip Analects: The Spread of Confucianism and Chinese Script” (2011), “The Changing Characteristics of the Shi in Ancient China and Their Significance” (2013), “The Contents and Nature of the Bamboo and Wooden Slips of Qin-Han Law: Focusing on Analysis of Qin Slips Collected by Yuelu Academy (III) • (IV)” (in Korean, 2016), and numerous articles. Anne Behnke Kinney is Professor of Chinese at the University of Virginia. She received a Ph.D. in Chinese language and literature from the University of Michigan in 1986. As a graduate student she was affiliated with the department of History and Archaeology at Peking University in Beijing. Her publications include Representations of Childhood and Youth in Early China, Chinese Views of Childhood, and Exemplary Women of Early China: The Lienü zhuan of Liu Xiang. She is the director of the digital research collection, Traditions of Exemplary Women. Her work focuses on literature and the social and intellectual history of early China. Ming-chiu Lai holds a doctorate from the University of Toronto. He is a professor of the department of history and director of the Centre for Chinese History at the Chinese University of Hong Kong and specializes in the political and social history of the Qin-Han and Wei-Jin periods. Currently, he is studying Qin-Han bamboo and wooden documents. He is the author of Fucou yu zhixu: Handiguo difang shehui yanjiu (Power Convergence and Social Order: The Study of Local Society of the Han Empire) (in Chinese, 2013) and the co-author (with Lam Shuk-kuen) of Han Yue heji: Han-Tang lingnan wenhua yu shenghuo (Cultural Interaction Between Han and Yue: Culture and Life in Han-Tang Lingnan Region) (in Chinese, 2013), as well as numerous articles. Vincent S. Leung is Associate Professor at the Department of History at Lingnan University, Hong Kong. He holds a doctorate in Chinese history from Harvard University. His research focuses on the political and intellectual history of ancient China, especially the transition from the Bronze Age to the rise of empires in the first millennium bce. He is the author of the forthcoming monograph Why History Mattered:The Politics of the Past in Early China. Li Feng is Professor of Early Chinese History and Archaeology at Columbia University. He received his M.A. from the Institute of Archaeology of CASS (1986) and Ph.D. from the University of Chicago (2000). Both a historian and a field archaeologist, Li Feng is broadly interested in the rise of complex society, early state organization, interregional cultural relations, the workings of bureaucracy, the nature of early writing, and the social and economic dynamics of early state and empire. He is particularly known for work on bronze inscriptions and led Columbia’s first international collaborative archaeological fieldwork in China in Guicheng, Shandong Province, in 2007–2010. He is the author of Landscape and Power in Early China (Cambridge 2006), Bureaucracy and the State in Early China (Cambridge 2008), Early China (Cambridge 2014), and co-editor of Writing and Literacy in Early China (Washington 2011), Daijiawan and Shigushan: A Catalog of Bronze Vessels Originating From Baoji, Shaanxi Province (Academia Sinica, 2015), and Guicheng: A Study of the Formation of States on the Jiaodong Peninsula in Late Bronze-Age China, xvi

Contributors

1000–500 bce (Science Press, forthcoming 2018). Li Feng received the Columbia University Lenfest Distinguished Faculty Award in 2015 and was appointed a Changjiang Scholar at Jilin University in China the same year. Luo Xinhui is Professor of Chinese Ancient History in the Department of History at Beijing Normal University, where she earned her doctorate in 1998. Her current research examines the origin and development of the concepts of tian 天 and di 帝, with a focus on oracle-bone and bronze inscriptions, as well as ancestor worship in the Shang and Zhou periods. Her recent publications include Annotations on Bronze Inscriptions from Shouyangzhai 首阳吉金注疏 (2015) and articles on early Chinese intellectual history. David Pankenier is Professor Emeritus of Modern Languages and Literatures at Lehigh University and earned his Ph.D. in Asian Languages from Stanford University in 1983. He is an International Fellow in the Swedish Collegium for Advanced Study, Uppsala, and member of the Institute for Advanced Study in Princeton. Dr. Pankenier served on the Executive Committee of the International Conferences on the Inspiration of Astronomical Phenomena (INSAP), the European Society for Astronomy in Culture (SEAC) and the advisory board of the Journal for the History of Astronomy.The author of many articles and four books on how astronomy, astrology, and cosmology shaped early Chinese culture, his most recent book, Astrology and Cosmology in Early China: Conforming Earth to Heaven, was published in 2013. Kate Pechenkina is Professor of anthropology and head of the anthropology department at Queens College, City University of New York. She is a bioarchaeologist and skeletal pathologist whose research is focused in reconstructing changes in human diet and health in early China. Yuri Pines is Michael W. Lipson Professor of Asian Studies at the Hebrew University of Jerusalem. His research focuses on early Chinese political thought, traditional Chinese political culture, early Chinese historiography, and the history of pre-imperial (pre-221 bce) China. His major monographs include The Book of Lord Shang: Apologetics of State Power in Early China (2017); The Everlasting Empire: Traditional Chinese Political Culture and Its Enduring Legacy (Princeton 2012); Envisioning Eternal Empire: Chinese Political Thought of the Warring States Era (Honolulu, 2009); Foundations of Confucian Thought: Intellectual Life in the Chunqiu Period, 722–453 b.c.e. (Honolulu, 2002). He co-authored (with Gideon Shelach and Yitzhak Shichor) the three-volume All-underHeaven: Imperial China (in Hebrew, Raanana, first volume 2011, second volume 2013, the third is forthcoming); co-edited together with Lothar von Falkenhausen, Gideon Shelach, and Robin D.S.Yates the Birth of an Empire:The State of Qin revisited (Berkeley, 2014), and with Paul R. Goldin and Martin Kern the Ideology of Power and Power of Ideology in Early China (Leiden, 2015). He also has published over 100 articles and book chapters. Charles Sanft is Associate Professor in the Department of History at the University of Tennessee, Knoxville. His monograph, Communication and Cooperation in Early Imperial China: Publicizing the Qin Dynasty, was published by the State University of New York Press in 2014, and his articles have appeared in Early China, Environmental History, and other journals. Axel Schuessler is Professor Emeritus at Wartburg College in Waverly, Iowa (USA). After high school Latin and Greek, he studied classical Chinese, Tibetan, and Mongolian as well as Sanskrit and Middle Indian languages at the Universität München. Publications include Etymological Dictionary of Old Chinese (2007) and Minimal Old Chinese and Later Han Chinese (2009). xvii

Contributors

Gideon Shelach-Lavi is the Louis Freiberg Professor of East Asian Studies and the director of the Institute of Asian and African studies at the Hebrew University. He is an archaeologist specializing in the Neolithic and Bronze Age of north China, with a Ph.D. from the Department of Anthropology, University of Pittsburgh (1996). Since 1994 he has been conducting archaeological field works in Northeast China and is currently co-heading the Fuxin Regional Archaeological Project in Liaoning province. He has published eight books and more than 60 papers in leading academic journals (including Science, Antiquity, Journal of Anthropological Archaeology, Journal of Archaeological Science, and more, including academic journals in China). Among his recent books are The Archaeology of China: From Prehistory to the Han Dynasty (2015) and Prehistoric Societies on the Northern Frontiers of China: Archaeological Perspectives on Identity Formation and Economic Change During the First Millennium bce (2009). Chen Shen currently serves as the Vice President World Cultures and is Senior Curator of Chinese Art and Archaeology at the Royal Ontario Museum, as well as being cross-appointed as a professor at the Department of East Asian Studies of the University of Toronto. He is the author of Anyang and Sanxingdui: Unveiling and Mysteries of Ancient Chinese Civilizations (2002) and Ancient Chinese Jades from the Royal Ontario Museum (2016), and the senior editor of Current Research in Chinese Pleistocene Archaeology (2003). Roel Sterckx is Joseph Needham Professor of Chinese History, Science and Civilisation at the University of Cambridge and Fellow of Clare College. His publications include The Animal and the Daemon in Early China (Albany: SUNY Press, 2002) and Food, Sacrifice and Sagehood in Early China (New York: Cambridge University Press, 2011). Ori Tavor is a lecturer in Chinese Studies at the department of East Asian Languages and Civilizations, University of Pennsylvania. His research focuses on early and medieval Chinese religion, philosophy, and self-cultivation practices. His current project examines the role of ritual theory in the development of organized religion. His work has been featured in Dao: A Journal of Comparative Philosophy, Studies in Chinese Religions, and Body and Religion. François Thierry studied in history, Chinese and Vietnamese language and civilization and graduated as agrégé in fine arts and art history. He first joined the Paris Mint Museum to write the Catalogue of the Far Eastern Coins Collection, then was recruited to the Coins and Medals Department of the National Library as Curator of Oriental Coins, where he was appointed Chief Curator of Oriental Coins (1999) and General Curator (2010). He is a specialist in Chinese numismatics and was awarded the Medal of the Royal Numismatic Society (2006). His recent publications include Les Monnaies de la Chine ancienne (Paris: Les Belles Lettres, 2017). Wicky W.K. Tse received his doctorate from the University of Pennsylvania in 2012 and is now Assistant Professor in the Department of Chinese Culture at the Hong Kong Polytechnic University. His main research interests are the history of violence and warfare in early and early medieval China and the formation of real and imagined frontiers of early Chinese empires as well. He is working on a monograph of the Latter Han empire and its northwestern frontier and chapters for the Cambridge History of War and the Cambridge World History of Violence (both forthcoming). Wang Haicheng is Mary and Cheney Cowles Endowed Associate Professor of Chinese Art in the School of Art + Art History + Design, University of Washington, Seattle. His research focuses xviii

Contributors

on the art and archaeology of early China, especially comparative studies between Bronze Age China and other early civilizations. Recent and forthcoming publications include Writing and the Ancient State (Cambridge, 2014), a chapter on the material culture of the Erligang civilization in Art and Archaeology of the Erligang Civilization (2014), a chapter on urbanization and writing in The Cambridge World History (2015), an article on administrative reach and documentary coverage in ancient states (Archéo-Nil 26, 2016), a chapter on Western Zhou despotism in Ancient States and Infrastructural Power, and articles on calligraphy and the archaeology of agency.

xix

INTRODUCTION

PAUL R. GOLDININTRODUCTION

What is early Chinese history? Paul R. Goldin

In this volume, researchers on three continents join forces to offer a concise but scholarly overview of the foundation of Chinese civilization. The pace of progress in the study of early China has been especially brisk since the reawakening of Chinese academic life at the end of the Cultural Revolution. Mostly because of new discoveries – though new approaches and methodologies have also played a role – what once might have seemed the dustiest and least controversial branch of Chinese studies has recently become one of the most vibrant.The concomitant reemergence of China as one of the world’s leading powers has also engendered considerable interest in the story of its genesis, with diverse viewpoints and values at stake. All of these trends are discussed in the pages that follow. As any good reference work ought to define its scope at the outset, the task of this introduction is to explain what is meant by “early Chinese history.” For each of the three words, there are pitfalls and ambiguities requiring exposition. To take each one in turn:

Early For the purposes of this volume, “early” is the easiest of the three words to define: conventionally, the period known as “early China” is said to end with the fall of the Eastern Han dynasty in ad 220 (see “The Latter Han empire and the End of Antiquity,” by Wicky W.K. Tse, Chapter 8). Early China, the journal that has lent the field its very name, specifies this as the endpoint of its coverage. The date of 220, however, is one of the most adventitious in Chinese history: it is the year when the great warlord Cao Cao 曹操 (155–220) happened to die, whereupon his son Cao Pi 曹丕 (187–226) decided that the time had come to end the moribund Han dynasty and declare himself the emperor of a new one, which he called Wei 魏. If Cao Cao had died in 219, we would say that the Han dynasty ended in 219; if he had died in 221, we would say that the Han dynasty ended in 221. Still, there are real socioeconomic, political, and intellectual changes underlying the arbitrary date, and thus it serves as a convenient cut-off point.The ensuing period, known as the Six Dynasties (220–589), was marked by social upheaval, political disunion, and the expansion of major religions, including both Daoism and Buddhism. Many scholars have observed that the seeds of these developments were sown in the Eastern Han (e.g., Ebrey 1990; Holcombe 1994), but accounting for their full germination would require the mastery of very different categories of sources. No one would seriously doubt that they belong to a different era. 1

Paul R. Goldin

But when does “early China” begin? This is a more difficult question to answer, in part because “China,” as we shall see presently, is not easy to define. There is a patently wrongheaded approach to this question, namely to commence the story of Chinese history with Homo erectus, the earliest known hominin in the region, whether in the form of Peking Man (e.g., Tung 1959: 7–8) or even earlier fossils such as Yuanmou Man 元謀猿人 (Yu Weichao et al. 1997: I, 12), whose remains were discovered in a county that was not even recognized as part of China until centuries later. But Homo erectus’s culture cannot plausibly be called Chinese in any respect, and it is doubtful that anyone in China today is directly descended from such species. Thus it is unproductive at best, and propagandistic at worst, to begin an account of Chinese history with Peking Man (Howells 1983: 298). For these reasons, this volume begins with the Neolithic cultures whose characteristics, from the perspective of social organization and material culture, demonstrably anticipated those of historical China. As we learn more about such societies with each archaeological discovery, these murky origins are continually reinterpreted (see the chapters by Shelach-Lavi and Pechenkina, Chapters 1 and 2). It must be borne in mind that the people who produced the artifacts that interest us did not necessarily speak an ancestral form of the Chinese language and could not have considered themselves ethnically or culturally Chinese (since neither the word nor the concept yet existed, as we shall see). Thus the phrase “Neolithic China” is a useful fiction: it refers to the complex of cultures in the region that we now identify as China, whose interactions ultimately led to the emergence of states that used written Chinese and displayed other traditional hallmarks of Chinese civilization (e.g., Liu 2005; Liu and Chen 2012; Shelach-Lavi 2015). It is not a concept that would have made sense in the Stone Age itself.

Chinese This brings us to a more difficult question, namely what is meant by “China” and “Chinese.” The name China is probably (but not assuredly) derived from Qin 秦 (Old Chinese *dzin, according to the system of reconstruction in Baxter and Sagart 2014), which was the dominant Chinese state in the western regions for most of the first millennium bc and went on to establish the first unified Chinese empire in 221 bc under the notorious First Emperor (e.g., Rao Zongyi 1993: 230–35; for a very different suggestion, see Wade 2009). One oft-heard objection to this hypothesis is that Chinese people do not normally identify themselves with Qin, which they have regarded as a brutal and failed regime (see the chapter by Charles Sanft, Chapter 6); rather, they have typically adopted the names of more successful dynasties, such as Han and Tang. Most readers will be familiar with the use of “Han” as an ethnonym (Hanzu 漢族 or Hanren 漢人 in Modern Mandarin; e.g., Mullaney 2012), and the standard Chinese term for “Chinatown” is Tangren jie 唐人街, literally “street [or neighborhood] of the Tang people.” So why do we say “China” and not “Hana” or “Tanga”? As China has always been an exonym rather than an endonym (apparently first attested in Sanskrit as cīna), its history has more to do with foreign than with Chinese usage (e.g., Kleine 2008). If we bear in mind that Qin was the westernmost Chinese state, it stands to reason that Central and South Asian travelers would have reached it before any other part of China, and thus the name could have come to stand for the whole subcontinent (Olivelle 2005: 22). Moreover, there is underappreciated evidence that the name Qin was used to refer to Chinese people even after the fall of the Qin dynasty (Krjukov et al. 1983: 353–54).1 Commenting on two such instances, the scholiast Yan Shigu 顔師古 (581–645) wrote: “Referring to Chinese people as ‘Qin people’ was an ancient way of speaking” 謂中國人為秦人，習故言也 and “In Qin times, there were people who fled to the Xiongnu; today, their descendants are still called ‘Qin people’ ” 2

Introduction

秦時有人亡入匈奴者，今其子孫尚號秦人 (Ban Gu et al. 1962: 96B.3913 and 94A.3782, respectively). Such locutions were evidently rare enough in Yan Shigu’s day that he felt obliged to explain them, but the best scholars were still aware of them.2 Yet more interesting is the term that Yan Shigu used to refer to Chinese people, namely Zhongguo ren 中國人, which remains the most common endonym today. Zhongguo is often translated as “the Middle Kingdom” (e.g., Pomfret 2016), but in antiquity it would probably have been construed as plural: “the Central States,” i.e. the Chinese domains along the lower Yellow River valley, closest to the royal seat at Luo 洛 and hence presumed to be closest in customs and mores to the ideal of the Sage King. Significantly, the connotations tended to be cultural rather than geographical. For example, in the Zuo Commentary to the Springs and Autumns (Chunqiu Zuozhuan 春秋左傳), the designation zhongguo is always contrasted with barbarians,3 who are identified by their uncivilized conduct rather than their ethnicity (Pines 2005; Goldin 2011) and are accordingly called yi 夷, “the destructive,” man 蠻, “the savage,” or rong 戎, “the warlike,” rather than by true ethnonyms, in this text. A typical example: “The zhongguo are pacified through virtue; the barbarians of the four directions are overawed [only] through punishment” 德以柔中國，刑以威四夷 (Yang Bojun 1990: I, 434 [Xi 僖 25]; compare the translation in Durrant et al. 2016: I, 391).4 Thus the effective meaning of zhongguo ren was “someone who behaves as a virtuous person from the Central States ought to behave,” not necessarily “someone from the Central States” (let alone “an ethnic Chinese person”). How old is this concept of zhongguo as “the place where people know how to behave”? Many Chinese scholars recognize the first use of the phrase in an inscription on a bronze vessel called He zun 何尊, cast during the reign of King Cheng 成王 (i.e. 1042–1021 bc). This text tells us that the king established a new capital at Luo and quoted his renowned father, King Wu 武王 (d. 1043 bc), as saying: “Let me dwell in this central territory and from here govern the people” (tr. David W. Pankenier in Cook and Goldin 2016: 18). At issue is the term rendered here as “central territory,” which in the inscription appears with the underdetermined graphs 中 或. Most Chinese palaeographers interpret this as zhongguo 中國, and some go so far as to call it the first record of “China” as a nation-state. (He Zhenpeng 2011 is a solid review.) A major unacknowledged problem with this interpretation is that what we call zhongguo – whether in the sense of “the Middle Kingdom” or “the Central States” – would probably have been written zhongbang 中邦 in the Bronze Age,5 because bang was systematically replaced by guo in received texts in order to avoid the taboo of writing the personal name of Emperor Gao of Han 漢高祖 (r. 202–195 bc), Liu Bang 劉邦 (Yoshimoto 2003: 582–84; for such taboos generally, see Chen Yuan 1928 and Adamek 2015). Palaeographical texts, which routinely write Bangfeng 邦風 rather than the familiar Guofeng 國風 for the section of the Odes known as “The Airs of the States,” or bangjia 邦家 rather than guojia 國家 for the set phrase “the state and its families,” permit the inference that imperial redactors dutifully changed bang to guo whenever they encountered it. Consequently, every instance of guo in texts that underwent such editing has to be viewed with suspicion. Since 中或 clearly cannot be zhongbang 中邦, what does it mean? There are two possibilities, which amount to nearly the same thing.The original meaning of guo 國 is “citadel” (as opposed to ye 野, the wilderness beyond the walls), a sense that is preserved in the phrase guoren 國人, which originally referred to “the denizens of the capital,” not “the people of the state.” Thus zhong guo in the He zun inscription could mean “the central citadel,” that is to say, the capital. “The People Toil” (“Minlao” 民勞), No. 253 in the Odes, contains the line “We appreciate this zhongguo” 惠此中國, for which the canonical Mao 毛 commentary supplies the gloss: “Zhongguo is the capital” 中國， 京師也 (Li Xueqin et al., 2000:VI, 1338a). Alternatively, 或 could be interpreted as yu 域,“region,” yielding zhongyu 中域,“the central region.” (Guo 國, Old Chinese 3

Paul R. Goldin

*C-qʷˤək, and yu 域, Old Chinese *ɢʷrək, were near homophones in the archaic language and probably cognate.)6 In either case, the king would have been referring to a narrowly delimited geographical area, not a kingdom or cultural “China.” With the He zun inscription eliminated as the source of Zhongguo in the sense of Zhongguo ren, “Chinese people,” the likeliest conclusion is that the concept is not attested until centuries later – perhaps not much earlier than the examples from the Zuo Commentary mentioned earlier. Thus we need to look elsewhere for early Chinese endonyms. These also tended to be cultural rather than geographic, and often self-congratulatory, such as hua 華, “luxuriant,” and xia 夏, “in full bloom” (Behr 2007) or perhaps “elegant, refined” (if xia is interpreted as a phonetic loan for ya 雅). One standard contrast is between “destructive” barbarians and “refined” Chinese (yixia 夷夏). What characteristics qualified someone as “luxuriant” or “refined”? One of the most important seems to have been the ability to read, write, and declaim texts in Old Chinese. Since the region was multilingual, and ethnicity still played a relatively minor role in determining cultural membership, this was true regardless of the speaker’s mother tongue, which could be completely unrelated (Pines 2005: 70; Wai-yee Li 2014: 243–46). One consequence is that Chinese was chosen to be the sole written language until at least the fifth century bc, and possibly even several centuries later (Goldin 2017: 125–26). Thus Chinese culture, history, and identity were intertwined, as early as the Bronze Age, with the Chinese language. The prestige accorded to the Chinese language derived in part from its status as the first to be encoded in writing (see the chapter by Luo Xinhui, Chapter 10), but there must have been other reasons, because this pattern – the first language to be written remains the only language to be written – is not common elsewhere in the world.The acceptance of Chinese as a marker of cultural attainment must also have had something to do with the acknowledged success of the Shang and Zhou dynasties, both of which used it in their written documents (see the chapters by Bagley and Li Feng, Chapters 3 and 4).7 Ancient Chinese thinkers identified “ritual” (li 禮) as another prime index of civilization (Creel 1970: 197), as in this famous statement by He Xiu 何休 (ad 129–182): “The Central States are states of ritual and morality” 中國者，禮義之國也 (Li Xueqin et al. 2000: XX, 68a [Yin 隱 7]).8 The impressive ritual vessels now on display in every major museum attest to the perceived significance of such ceremonies, but we know embarrassingly little about them.There is compelling evidence that they changed over time (e.g., Rawson 1999: 433–40; Falkenhausen 2006: 48–52) and even varied synchronically across regions and communities. In inscriptional literature, often we know that a certain graph must refer to a specific ritual, but can do little more than speculate as to its nature. Likewise, received texts may expatiate on high rituals of state, but we cannot say how closely such accounts reflect true practice, especially since many of them are refracted through poetry (e.g., Kern 2009). Jaded moderns might suspect that a concept as fluid as “ritual” lent itself to chauvinistic attitudes toward aliens (how could “barbarians” ever live up to such a standard if Chinese sources clearly reflect variation, if not outright disagreement?), and no society is completely devoid of xenophobic impulses in search of rationalization, but, at least in surviving texts, an accusation of violating “ritual” was most commonly leveled by one Chinese lord against another (e.g., for the Zuo Commentary, Schaberg 2001: 139–48; Pines 2002: 89–118; Wai-yee Li 2007: 295–320). Geographical and essentialist constructions of Chinese identity were voiced alongside humanistic ones. As civilization came to be associated with the glorious societies of the North Chinese heartland (often called zhongyuan 中原, “the central plain”), “China” was plotted spatially: not only the region where people behave as people should, but also the territory at the center of the world (Keightley 2000: 82–6; Ge Zhaoguang 2011, with further 4

Introduction

thoughts in 2014; also Zhang Longxi 2015). The association between people’s environment and their way of life was so strongly perceived that some sources evince a kind of geographical determinism, attributing specific character traits to inhabitants of different regions (Lewis 2006: 202–12; Goldin 2015: 38–40; Shao-yun Yang 2015). Such notions were hardened by encounters with mounted nomads from the steppe, who, unlike so many previous ethnic groups, proved unwilling to accept the supremacy of Chinese mores (largely because their arid territory was not compatible with Chinese agrarianist assumptions). This process led, in some Chinese writers, to a rigid conception of human nature: Heaven simply created some people differently, and it is foolish to pretend that everyone can be civilized (Goldin 2011: 228–35). Thus the story of early China is the story of an emerging, constructed, and repeatedly renegotiated Chinese identity. Asking what it means to be Chinese was always connected with asking what it means to be civilized. I shall close this section by listing seven basic features of Chinese civilization that endured despite regional diversity and historical change: (1) ancestor worship and respect for elders, especially parents; (2) the use of written Chinese as a lingua franca; (3) belief in the superiority of Chinese culture and the wisdom of Sinicizing foreigners; (4) a lack of any native tradition of democracy; (5) a preference for civil over military methods of control; (6) openness to religious diversity combined with intolerance of autonomous religious authority on the part of the imperial government; and (7) an imbalanced sex ratio exacerbated by polygyny among the elite, resulting in a large and restive population of unmarried males. Each of these has been the subject of many books and cannot be defended in extenso here.

History Lastly, “history” itself is problematic and contested, not least because it is not a Chinese word. Chinese civilization has always had a profound historical consciousness: the combination of China’s large population, which posed unique administrative challenges, and its relatively early invention of writing produced a voluminous historical record that is the envy of the world. “No other ancient nation possesses records of its whole past so voluminous, so continuous, or so accurate,” wrote Charles S. Gardner as early as 1938 (Gardner 1938: 105). Nevertheless, the classical language does not have a word that is precisely coterminous with “history” in a modern sense. Any other expectation would be anachronistic, after all. The closest word is shi 史 (as in modern words like shixue 史學, “historical studies”), but that originally denoted court officials with diverse administrative and clerical duties, including the recording of history, but certainly not limited to it (Harbsmeier 1995: 60–6; Vogelsang 2007: 17–91). For the Bronze Age, one scholar has proposed the sensible translation “secretary,” as in our “Secretary of State” (Kern 2007: 115–18), but by the Eastern Zhou, shi had come to refer to éminences grises at court who served as archivists, diviners, dispensers of wisdom who could be consulted on matters of ritual and strategy – and historiographers, since truthfully recording affairs of state was regarded as crucial to the project of judging rulers fairly and learning from the past.9 But what is meant by “truthfully”? Two famous anecdotes from the Zuo Commentary confirm that moral truth was prized, even to the extent that factual truth could be sacrificed in its behalf (Schaberg 2001: 262–64).10 In the first (Yang Bojun 1990: II, 662–63 [Xuan 宣 2]), the tyrannical Lord Ling of Jin 晉靈公 (r. 620–607 bc) plotted to kill his chief minister, Zhao Dun 趙 盾, because of the latter’s inconvenient remonstrances. Zhao escaped and was on his way out of the country, but had not yet reached the border, when he heard that his cousin had assassinated 5

Paul R. Goldin

Lord Ling, whereupon he returned to the capital and installed a new (and worthier) ruler. At this juncture, we read: 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？」. . . . . . 孔子曰：「董狐，古之良史也，書法不隱。趙宣子，古之良大夫也，為法 受惡。」 The Grand Historian wrote: “Zhao Dun assassinated his lord,” and displayed it in the court. Xuanzi [i.e. Zhao Dun] said: “It is not so.” He replied: “Sir, you are the chief minister. You fled but did not cross the border; when you returned, you did not punish the criminal. If it was not you, who was it?” . . . Confucius said: “Dong Hu was a fine historian of old; in writing history, his principle was not to conceal. Zhao Xuanzi was a fine grandee of old; he accepted this disgrace for the sake of principle.” (Compare the translation in Durrant et al. 2016: I, 597) Confucius’s final comment explains that by recording the event as he did, the historian Dong Hu achieved two things that were much more important than settling the pedestrian question of who in fact stabbed Lord Ling. First, he emphasized the principle that a chief minister cannot condone the assassination of the sovereign, even if the sovereign is wicked and deserves to be killed. Second, he afforded Zhao Dun the chance to forbear and let the comment stand, and thereby exhibit his own commitment to such high-minded principles. Although Zhao Dun goes down in history as a regicide, sensitive readers are expected to discern that he must have been a deeply ethical man. A vastly less ethical man, Cui Zhu 崔杼, occasioned a similar historiographical dilemma (Yang Bojun 1990: III, 1099 [Xiang 襄 25]) when he laid a deadly trap for Lord Zhuang of Qi 齊莊公 (r. 553–548 bc), who had been cuckolding him. At the fateful moment, Cui withdrew, permitting his guards to bring the matter to a close. The narrative then addresses the inevitable problem of historical judgment. When the Grand Historian recorded that “Cui Zhu assassinated his lord” 崔杼弒其君, Cui killed him – as well as his younger brother, who wrote the same thing as soon as he succeeded to the post (a hint that “Grand Historian” was a hereditary position). Only after a third brother recorded the same verdict did Cui finally relent. Presumably, Cui was hoping that because he had not in fact killed Lord Zhuang – the fatal arrow was loosed by an unidentified henchman – he could escape the verdict of the historians. In a world where moral correctness counted for more than factual correctness, he was soon to be disabused.11 Even Confucius is said more than once to have declined to correct historical records that he knew were factually incorrect if he could thereby teach readers a more important lesson (Yang Bojun 1990: I, 473 [Xi 28]; Li Xueqin et al. 2000: XXI, 567b-69a).12 There is much to admire in the scrupulousness of this didactic historiography, which one modern scholar has characterized as ad usum delphini, that is to say, intended for the instruction of statesmen (Vogelsang 2005: 151; see also Vogelsang 2007: 251–54). But when we read such documents today, how do we know which events were recorded as the historian saw them with his eyes, and which were recorded as he saw them with his heart? We cannot be sure of the answer. There are further complexities. The two anecdotes about historians’ dilemmas from the Zuo Commentary are what might be called meta-historiography: historians writing about 6

Introduction

historians writing about history. Consequently, we need to consider not only how Dong Hu and other straitened historians would have chosen to record a messy assassination, but also how an unrelated (and unnamed) set of historians chose to present such instructive anecdotes for posterity. We do not, and probably never will, have Dong Hu’s original text; indeed, we take it on faith that there was any such thing. Although there are reasonable differences of opinion (see, e.g., Pines 2002 and Van Auken 2016 for different perspectives), my view is that a large proportion of early Chinese “historical” literature is best interpreted as rhetorical impersonation rather than records of fact. Writers vaguely knew (or had heard – perhaps there was not much difference) that Zhao Dun had been implicated in Lord Ling’s assassination by virtue of his position, even though he did not personally participate, and imaginatively reconstructed the quandary facing the court historian as he was obliged to render judgment. The same theory of imaginative reconstruction can be applied to the lengthy and elegant speeches in a similar text, Discourses of the States (Guoyu 國語). Western authors and orators were trained in comparable rhetorical exercises, which they called ēthopoeia (Kennedy 2003). Compounding our interpretive problem is the fact that premodern audiences were undoubtedly more familiar with the circumstances of such events than we are today. But we can compensate with information that was unavailable to anyone living before the twentieth century: the transformational results of archaeological excavation. (My own reflections in Goldin 2005: 3–6 are already out of date.) These include not only previously unknown texts (or unknown versions of them) but also indispensable evidence relating to habitation, social organization, manufacture, and trade (e.g., Underhill 2013; Campbell 2014; Barnes 2015). A prime example is the Shang dynasty, whose very historicity was questioned before the excavations at Anyang 安陽 revealed not only one of the major bronze-producing civilizations of the ancient world but also thousands of oracle-bone inscriptions that had not seen the light of day in over three millennia (Li Chi 1957 is still informative). With such unprecedented sources at our disposal, it is safe to say that we know more about Shang society and culture than anyone who lived during the many centuries of imperial China. The Anyang archaeological project continues to this day. Yet even with this expanded inventory, there are many subjects for which we simply do not have adequate sources. Oracle-bone inscriptions provide fascinating details about the royal cult, procedures of divination, conception of the cosmos, and so on, but virtually nothing about the lives of ordinary men and women. Most of what we know about early Chinese history has to do with the activities of elite men. Legal and administrative texts, which yield glimpses of life beyond the royal and noble estates, are an important exception (see the chapter by Kyong-ho Kim and Ming Chiu Lai, Chapter 18). As we have seen, Chinese historians have been pondering the question of how to document important events for many centuries, and one of the greatest of them, Sima Qian 司馬遷 (145?–86? bc), left behind a work, Records of the Historian (Shiji 史記), whose format has inspired the present volume. Sima’s masterwork is divided into five parts: “basic annals” (benji 本紀), or reign-by-reign accounts of the emperors; “tables” (biao 表), which arrange genealogical and chronological data in convenient form; “treatises” (shu 書), which are essays covering important topics such as ritual and finance; and two final sections, “hereditary houses” (shijia 世家) and “arrayed traditions” (liezhuan 列傳), which include many biographies of exemplary figures as well as discussions of pre-imperial states and foreign peoples. The two parts of our book are akin to the “basic annals” and “treatises”: we begin with an overview of each major period and then turn to chapters on topics that transcend any particular 7

Paul R. Goldin

era. (It would not have been feasible to reproduce all five of Sima Qian’s divisions in a modern work.) This structure allows us to offer both synchronic and diachronic perspectives. With the meaning and significance of “early Chinese history” now clarified, it is time to let the contributors tell the story from here.

Notes 1 Revealingly, Rome was called “Great Qin” (Da Qin 大秦) because it was “like China in some respects” (you lei Zhongguo 有類中國; Fan Ye et al. 1965: 2919). 2 Complicating the matter is that “Qin” was sometimes used to refer to the short-lived Later Qin 後秦 dynasty (ad 384–417) or North China in that period more generally (Ji and Zhou 2009). This usage is fundamentally distinct. 3 Zhongguo and barbarians are also explicitly contrasted in the Odes (Shijing 詩經) and Gongyang Commentary (Gongyang zhuan 公羊傳). See Li Xueqin et al. 2000: V, 738b (the Minor Preface to “Liuyue” 六月); and XXI, 594b (Zhao 昭 23), respectively. 4 A similar distinction between virtue and punishment as modes of motivation appears in Analects 2.3. The three other appearances of zhongguo in Zuozhuan are Yang Bojun 1990: I, 249 (Zhuang 莊 31); II, 832 (Cheng 成 7); and IV, 1309 (Zhao 9). In addition, there are two citations from “Minlao,” a canonical ode that contains the phrase (to be discussed later):Yang Bojun 1990: I, 472; and IV, 1421. 5 Zhongbang is rare in received literature, but attested.The only well-known instance is from the “Yugong” 禹貢 chapter of Exalted Documents (Shangshu 尚書): “He established the revenues of the zhongbang” 成   ue (Yuejue shu 越絕書), 賦中邦 (Li Xueqin et al. 2000: II, 198a). But Documents on the Excellence of Y neglected until very recently, contains two others (Li Bujia 2013: 81 and 367); for the first, there is a revealing parallel in the Gongyang Commentary that reads zhongguo instead (Li Xueqin et al. 2000: XXI, 644b [Ding 定 4]). Perhaps Yuejue shu preserves zhongbang because it was not widely transmitted and thus escaped the systematic replacement of bang for guo. Incidentally, bang (Old Chinese *pˤroŋ) is obviously cognate with feng 封 (*proŋ), which refers to establishing the borders of land that has been awarded by the king and consequently became an important administrative term (e.g., Ren Wei 2004). Li Feng 2008: 48n.10 recognizes the semantic connection, but not the phonological one. 6 As Schuessler 2007: 268 observes, xu 淢/洫 (moat), yu 閾 (threshold), and even you 囿 (enclosure, ranch), must belong to the same etymon. 7 In philosophy and poetics, the supremacy of the Chinese language led to the widespread cultural assumption that objects and concepts with similar-sounding Chinese names or similar-looking Chinese graphs must supervene on some categorical connection in reality (Peterson 1982: 110–16; Pauline Yu 1987: 37–43; Bao 1990). This made rhyme, assonance, and paronomasia especially powerful literary devices (Behr 2005b; Goldin 2005: 14ff.). 8 Similarly, after Empress Dowager Lü 呂太后 (d. 180 bc) rejected a marriage proposal from the barbarian chieftain Modu 冒頓 (d. 174 bc), which she considered impertinent, he is reported to have stated: “I have not yet learned the ritual and morality of the Central States” 未嘗聞中國禮義 (Ban Gu et al. 1962: 3755). This appears in a Chinese text, naturally. 9 N.b.: Shǐ 史 is distinct from shì 士, “men of service,” a term discussed in the chapter by Yuri Pines, Chapter 13 (see also Pines 2009a: 115–84; and for a very different view, Yan Buke 1996: 29–72). The two words are cognate (Old Chinese *s-rəʔ and *m-s-rəʔ, respectively; see Behr 2005a: 16–18), but their usage and connotations grew apart. In Modern Mandarin, their pronunciations are distinguished only by tone. (Shì 事, “to serve,” Old Chinese *m-s-rəʔ-s, is manifestly cognate too, as was noticed even in antiquity: Jiang Renjie 1996: 78.) 10 The following discussion is condensed from Goldin 2008: 86–8. 11 Pines (2009b: 329–31) discusses another case of regicide and the historiographical distortions that it occasioned. Not all historians had high-minded motivations. 12 On the latter, see Gentz 2001: 96–9.The principle was well understood by premodern readers. For example, Su Xun 蘇洵 (1009–1066) remarked: “In the classics, sometimes false announcements are recorded, and sometimes things are not recorded for reasons of taboo-avoidance; there is a multitude of such cases, which are all merely for the sake of convenience in instruction” 經或從偽赴 [=訃] 而書,或隱諱而不 書，若此者眾，皆適於教而已 (Zeng Zaozhuang and Jin Chengli 1993: 230; cf. Klein 2010: 112).

8

Introduction

Works cited Adamek, Piotr. (2015) A Good Son Is Sad if He Hears the Name of His Father:The Tabooing of Names in China as a Way of Implementing Social Values, Leeds, UK: Maney. Ban, Gu 班固 (A.D. 32–92) et al. (1962) Hanshu 漢書, Beijing: Zhonghua. Bao, Zhiming. (1990) “Language and World View in Ancient China,” Philosophy East and West 40.2: 195–219. Barnes, Gina L. (2015) Archaeology of East Asia: The Rise of Civilization in China, Korea and Japan, Oxford and Philadelphia: Oxbow. Baxter, William H., and Laurent Sagart. (2014) Old Chinese: A New Reconstruction, Oxford: Oxford University Press. Behr,Wolfgang. (2005a) “Language Change in Premodern China – Notes on Its Perception and Its Impact on the Idea of a ‘Constant Way,’ ” in Helwig Schmidt-Glintzer et al. (eds.) Historical Truth, Historical Criticism, and Ideology: Chinese Historiography and Historical Culture from a New Comparative Perspective, Leiden and Boston: Brill. Behr, Wolfgang. (2005b) “Three Sound-Correlated Text Structuring Devices in Pre-Qín Philosophical Prose,” Bochumer Jahrbuch zur Ostasienforschung 29: 15–33. Behr, Wolfgang. (2007) “Xià: Etymologisches zur Herkunft des ältesten chinesischen Staatsnamens,” Asiatische Studien 61.3: 727–54. Campbell, Roderick B. (2014) Archaeology of the Chinese Bronze Age: From Erlitou to Anyang, Los Angeles: Cotsen Institute of Archaeology, UCLA. Chen Yuan 陳垣 (1880–1971). (1928) “Shi hui juli 史諱舉例,” Yanjing xuebao 燕京學報 4: 537–651. Cook, Constance A., and Paul R. Goldin. (eds.) 2016 A Source Book of Ancient Chinese Bronze Inscriptions, Berkeley: Society for the Study of Early China. Creel, Herrlee G. (1970) The Origins of Statecraft in China, Chicago and London: University of Chicago Press. Durrant, Stephen, et al. (trs). (2016) Zuo Tradition/Zuozhuan 左傳: Commentary on the “Spring and Autumn Annals”, Seattle: University of Washington Press. Ebrey, Patricia Buckley. (1990) “Toward a Better Understanding of the Later Han Upper Class,” in Albert E. Dien (ed.) State and Society in Early Medieval China, Stanford: Stanford University Press. Falkenhausen, Lothar von. (2006) Chinese Society in the Age of Confucius (1000–250 bc): The Archaeological Evidence, Los Angeles: Cotsen Institute of Archaeology, UCLA. Fan,Ye 范曄 (398–445) et al. (1965) Hou-Han shu 後漢書, Beijing: Zhonghua. Gardner, Charles S. (1938) Chinese Traditional Historiography, Cambridge, MA: Harvard University Press. Ge, Zhaoguang 葛兆光. (2011) Zhai zi Zhongguo: Chongjian youguan “Zhongguo” de lishi lunshu 宅兹中 國：重建有關「中國」的歷史論述, Beijing: Zhonghua. Ge, Zhaoguang. (2014) He wei “Zhongguo”? Jiangyu, minzu, wenhua yu lishi 何爲「中國」？疆域、民 族、文化與歷史, Hong Kong: Oxford University Press. Gentz, Joachim. (2001) Das Gongyang zhuan: Auslegung und Kanonisierung der Frühlings- und Herbstannalen (Chunqiu), Wiesbaden: Harrassowitz. Goldin, Paul R. (2005) After Confucius: Studies in Early Chinese Philosophy, Honolulu: University of Hawaii Press. Goldin, Paul R. (2008) “Appeals to History in Early Chinese Philosophy and Rhetoric,” Journal of Chinese Philosophy 35.1: 79–96. Goldin, Paul R. (2011) “Steppe Nomads as a Philosophical Problem in Classical China,” in Paula L.W. Sabloff (ed.) Mapping Mongolia: Situating Mongolia in the World from Geologic Time to the Present, Philadelphia: University of Pennsylvania Museum of Archaeology and Anthropology. Goldin, Paul R. (2015) “Representations of Regional Diversity During the Eastern Zhou Dynasty,” in Yuri Pines et al. (eds.) Ideology of Power and Power of Ideology in Early China, Leiden and Boston: Brill. Goldin, Paul R. (2017) “Some Shang Antecedents of Later Chinese Ideology and Culture,” Journal of the American Oriental Society 137.1: 123–29. Harbsmeier, Christoph. (1995) “Some Notions of Time and of History in China and in the West: With a Digression on the Anthropology of Writing,” in Chun-chieh Huang and Erik Zürcher (eds.) Time and Space in Chinese Culture, Leiden: E.J. Brill. He Zhenpeng 何振鵬. (2011) “He zun mingwenzhong de ‘Zhongguo’ ” 何尊銘文中的「中國」, Wenbo 文博 6: 32–4. Holcombe, Charles. (1994) In the Shadow of the Han: Literati Thought and Society at the Beginning of the Southern Dynasties, Honolulu: University of Hawaii Press.

9

Paul R. Goldin Howells, W.W. (1983) “Origins of the Chinese People: Interpretations of the Recent Evidence,” in David Keightley (ed.) Origins of Chinese Civilization, Berkeley: University of California Press. Ji, Xiangxiang 計翔翔 and Zhou Yan 周燕. (2009) “Faxian zhuan buneng zuo yuwai cheng Zhongguoren wei ‘Qinren’ de lizheng – Cihai ‘Qinren’ tiao jiucuo” 《法顯傳》不能作域外稱中國人為’秦人’的例 證 – 《辭海》’秦人’條糾錯, Zhejiang Daxue xuebao (Renwen shehui kexue ban) 浙江大學學報（人文 社會科學版） 1: 111–17. Jiang, Renjie 蔣人傑. (1996) Shuowen jiezi jizhu 説文解字集注, ed. Liu Rui 劉銳, Shanghai: Guji. Keightley, David N. (2000) The Ancestral Landscape: Time, Space, and Community in Late Shang China (ca. 1200–1045 B.C.), Berkeley: Institute of East Asian Studies. Kennedy, George A. (tr). (2003) Progymnasmata: Greek Textbooks of Prose Composition and Rhetoric, Atlanta: Society of Biblical Literature. Kern, Martin. (2007) “The Performance of Writing in Western Zhou China,” in S. La Porta and D. Shulman (eds.) The Poetics of Grammar and the Metaphysics of Sound and Sign, Leiden and Boston: Brill. Kern, Martin. (2009) “Bronze Inscriptions, the Shijing and the Shangshu: The Evolution of the Ancestral Sacrifice During the Western Zhou,” in John Lagerwey and Marc Kalinowski (eds.) Early Chinese Religion, Part One: Shang through Han (1250 bc–220 ad), Leiden: Brill. Klein, Esther Sunkyung. (2010) “The History of a Historian: Perspectives on the Authorial Roles of Sima Qian,” Ph.D. diss., Princeton University. Kleine, Christoph. (2008) “Anmerkungen zu Herkunft, Gebrauch und Bedeutung des Toponyms ‘Shina’ 支那 und verwandter Bezeichnungen für China,” Bochumer Jahrbuch zur Ostasienforschung 32: 115–36. Krjukov, M.V., et al. (1983) Drevnie Kitajcy v èpoxu centralizovannyx imperij, Moscow: Nauka. Lewis, Mark Edward. (2006) The Construction of Space in Early China, Albany: State University of New York Press. Li, Bujia 李步嘉. (2013) Yuejue shu jiaoshi 越絕書校釋, Beijing: Zhonghua. Li, Chi. (1957) The Beginnings of Chinese Civilization, Seattle: University of Washington Press. Li, Feng. (2008) Bureaucracy and the State in Early China: Governing the Western Zhou, Cambridge: Cambridge University Press. Li, Wai-yee. (2007) The Readability of the Past in Early Chinese Historiography, Cambridge, MA: Harvard University Asia Center. Li, Wai-yee. (2014) “Poetry and Diplomacy in the Zuozhuan,” Journal of Chinese Literature and Culture 1: 241–61. Li, Xueqin 李學勤 et al. (2000) Shisan jing zhushu: Zhengli ben 十三經注疏：整理本, Beijing: Beijing Daxue. Liu, Li. (2005) The Chinese Neolithic:Trajectories to Early States, Cambridge: Cambridge University Press. Liu, Li, and Xingcan Chen. (2012) The Archaeology of China: From the Late Paleolithic to the Early Bronze Age, Cambridge: Cambridge University Press. Mullaney, Thomas S. (ed.) (2012) Critical Han Studies: The History, Representation, and Identity of China’s Majority, Berkeley: University of California Press. Olivelle, Patrick. (2005) Manu’s Code of Law: A Critical Edition and Translation of the Mānava-Dharmaśāstra, Oxford: Oxford University Press. Peterson,Willard J. (1982) “Making Connections: ‘Commentary on the Attached Verbalizations’ of the Book of Change,” Harvard Journal of Asiatic Studies 42.1: 67–116. Pines, Yuri. (2002) Foundations of Confucian Thought: Intellectual Life in the Chunqiu Period, 722–453 b.c.e., Honolulu: University of Hawaii Press. Pines, Yuri. (2005) “Beasts or Humans: Pre-Imperial Origins of Sino-Barbarian Dichotomy,” in Reuven Amitai and Michal Biran (eds.) Mongols,Turks, and Others: Eurasian Nomads and the Sedentary World, Leiden and Boston: Brill. Pines,Yuri. (2009a) “Chinese History Writing Between the Sacred and the Secular,” in John Lagerwey and Marc Kalinowski (eds.) Early Chinese Religion, Part One: Shang Through Han (1250 bc-220 ad), Leiden: Brill. Pines,Yuri. (2009b) Envisioning Eternal Empire: Chinese Political Thought of the Warring States Era, Honolulu: University of Hawaii Press. Pomfret, John. (2016) The Beautiful Country and the Middle Kingdom: America and China, 1776 to the Present, New York: Henry Holt and Company. Rao, Zongyi 饒宗頤. (1993) Fanxue ji 梵學集, Shanghai: Guji.

10

Introduction Rawson, Jessica. (1999) “Western Zhou Archaeology,” in Michael Loewe and Edward L. Shaughnessy (eds.) The Cambridge History of Ancient China: From the Origins of Civilization to 221 B.C., Cambridge: Cambridge University Press. Ren, Wei 任偉. (2004) Xi-Zhou fengguo kaoyi 西周封國考疑, Beijing: Shehui kexue wenxian. Schaberg, David. (2001) A Patterned Past: Form and Thought in Early Chinese Historiography, Cambridge, MA and London: Harvard University Asia Center. Schuessler, Axel. (2007) Abc Etymological Dictionary of Old Chinese, Honolulu: University of Hawaii Press. Shelach-Lavi, Gideon. (2015) The Archaeology of Early China: From Prehistory to the Han Dynasty, Cambridge: Cambridge University Press. Tung Chi-ming. (1959) An Outline History of China, Peking: Foreign Languages Press. Underhill, Anne P. (ed.) (2013) A Companion to Chinese Archaeology, Chichester, UK: Wiley-Blackwell. Van Auken, Newell Ann. (2016) The Commentarial Transformation of the Spring and Autumn, Albany: State University of New York Press. Vogelsang, Kai. (2005) “Some Notions of Historical Judgment in China and the West,” in Helwig SchmidtGlintzer et al. (eds.) Historical Truth, Historical Criticism, and Ideology: Chinese Historiography and Historical Culture from a New Comparative Perspective, Leiden and Boston: Brill. Vogelsang, Kai. (2007) Geschichte als Problem: Entstehung, Formen und Funktionen von Geschichtsschreibung im Alten China, Wiesbaden: Harrassowitz. Wade, Geoff. (2009) “The Polity of Yelang (夜郎) and the Origins of the Name ‘China,’ ” Sino-Platonic Papers 188. Yan, Buke 閻步克. (1996) Shi dafu zhengzhi yansheng shi gao 士大夫政治演生史稿, Beijing: Beijing Daxue. Yang, Bojun 楊伯峻. (1990) Chunqiu Zuozhuan zhu 春秋左傳注, 2nd edition, Beijing: Zhonghua. Yang, Shao-yun. (2015) “ ‘Their Lands Are Peripheral and Their qi Is Blocked Up’: The Uses of Environmental Determinism in Han (206 bce-220 ce) and Tang (618–907 ce) Chinese Interpretations of the ‘Barbarians,’ ” in Rebecca Futo Kennedy and Molly Jones-Lewis (eds.) The Routledge Handbook of Identity and the Environment in the Classical and Medieval Worlds, Abingdon, UK: Routledge. Yoshimoto Michimasa 吉本道雅. (2003) “Shunjū kokujin saikō” 春秋國人再考, Ritsumeikan bungaku 立 命館文學 578: 581–92. Yu, Pauline. (1987) The Reading of Imagery in the Chinese Poetic Tradition, Princeton: Princeton University Press. Yu, Weichao et al. (1997) A Journey into China’s Antiquity, Beijing: Morning Glory. Zeng, Zaozhuang 曾棗莊 and Jin Chengli 金成禮. (eds.) (1993) Jiayou ji jianzhu 嘉祐集箋注, Shanghai: Guji. Zhang, Longxi. (2015) “Reconceptualizing China in Our Time: From a Chinese Perspective,” European Review 23.2: 193–209.

11

PART I

Chronology

1 MAIN ISSUES IN THE STUDY OF THE CHINESE NEOLITHIC GIDEON SHELACH-LAVIMAIN ISSUES IN THE CHINESE NEOLITHIC

Gideon Shelach-Lavi

This chapter covers an exceptionally long and complex period. It starts with the incipient social and economic processes that began during the pre-Neolithic periods some 20,000 years ago and eventually led to the transition to agriculture and sedentary lifeways some 8,500 years ago and ends some 4,000 years ago, when state-level societies began to emerge in different regions of China. The Chinese Neolithic, so defined, includes a vast number of societies, each evolving in its unique local environment and developing unique cultural, economic and social forms.Thus, even an entire book could not describe them all or do justice to their variations, let alone a brief summery chapter such as this.Writing this chapter, more than any other chapter in this book, requires that I select not only what data to include in it but, even more fundamentally, which issues are crucial to include in such a chapter and which to omit. This sort of selection is subjective, of course: topics that I deem important may be less important to other scholars and vice versa. Interested readers are, therefore, advised to seek more comprehensive treatments in recently published books on the archaeology of ancient China (e.g., Liu and Chen 2012; Shelach-Lavi 2015; Zhongguo 2010). Despite popular but contested claims that the origins of Chinese written language trace back to the late Neolithic period (e.g., Keightley 2006) and even earlier (Li et al. 2003), the earliest documents in which longer combination of characters convey complex messages of any historic value discovered so far in China are the oracle bones (Chinese: jia-gu-wen), dated to the late second millennium bce (see Bagley this volume). Thus, studying the prehistoric societies of Neolithic China is based on the painstaking collection, sorting and analysis of archaeological data. While this data can be studied using different approaches, with different theoretical and methodological tools, my own approach is anthropologically based. In this chapter, I therefore emphasize issues that are relevant to our understanding of the human society and the processes of change it underwent. Alongside the presentation of relevant archaeological data, I also discuss some of the outstanding questions that are still unresolved or await further research. Rather than focusing on the Yellow River basin, which is sometime considered the “cradle of Chinese civilization”, I choose in this chapter to emphasize the cultural diversity of societies from different regions of China and to examine the extent to which these societies were in contact with each other. Because of the compressed nature of this chapter, I use simplified, and admittedly quite simplistic, geographical and chronological terminology. ‘China,’ itself an anachronistic term for 15

Gideon Shelach-Lavi

11 12 9 2

8 7

45

3

16 17 18

6

1 13

20 21

15

25

19

14 28 29 31

22 23

24

10

30 32

33 27

26

Figure 1.1 The locations of all sites mentioned in the chapter. 1. Liuwan (柳湾); 2. Shimao (石峁); 3. Shizitan (柿子滩); 4. Longwangchan (龙王辿); 5. Taosi (陶寺); 6. Xiachuan (下川); 7. Jiangou (澗溝); 8. Nanzhuangtou (南庄头); 9. Zhuannian (转年); 10. Donghulin (东胡林); 11. Xiaohexi (小河西); 12. Chahai (查海); 13. Wangchenggang (王城岗); 14. Xinzhai (新砦); 15.Xishuipo (西水坡);16.Xizhufeng (西朱封);17.Bianbiandong (扁扁洞);18.Liangchengzhen (两城镇); 19.Yaowangcheng (尧王城); 20. Chengtoushan (城头山); 21. Pengtoushan (彭頭山); 22. Bashidang (八十垱); 23.Yuchanyan (玉蟾岩); 24. Zengpiyan (甑皮岩); 25. Miaoyan (庙岩); 26. Dingsishan (顶蛳山); 27. Shixia(石峡); 28. Sidun (寺墩); 29. Fanshan (反山); 30. Chuodun (绰墩); 31. Shangshan (上山); 32. Tianluoshan (田螺山); 33. Kuahuqiao (跨湖桥)

the Neolithic period but used here as a convenient shorthand, is divided, for the purpose of my discussion, into the north (the Wei and Yellow River basins region and areas to their north), center (the middle and lower Yangtze basin), south (areas south of the Yangtze River basin) and west (areas currently in Sichuan and Yunnan) (see Figure 1.1 for the location of all sites mentioned in this chapter). Chronologically, the Neolithic is commonly divided into numerous spatio-temporal entities (or archaeological ‘cultures’), but is here divided into the pre-Neolithic (ca. 18,000–6,500 bce); Early Neolithic (ca. 6,500–5,000 bce), Middle Neolithic (ca. 5,000–3,500 bce) and Late Neolithic (ca. 3,500–2,000 bce) (Table 1.1).

The deep background: processes during the pre-Neolithic period The transition to agriculture is a global phenomenon and one of the most significant processes in the history of humankind. It not only transformed the relationship between humans and their natural environment and the nature of human adaptation but also profoundly and fundamentally altered human social relations and culture. Everything that typifies our society today, or the historical era for that matter, including dense concentrations of populations, cities, states, food surpluses that support non-productive activities, advanced technologies and 16

Main issues in the Chinese Neolithic Table 1.1 The chronological and geographic framework of this chapter and some of the main associated Neolithic archaeological ‘cultures’ North China

Central China

South China

Pre-Neolithic (ca. 18,000–6,500 bce) Early Neolithic (ca. Dadiwan (大地湾); Pengtoushan (彭头 Dingsishan (顶蛳 山); Keqiutou 山); Shangshan 6,500–5,000 bce) Peiligang (裴李 (壳坵头) (上山); 岗); Cishan (磁 Kuahuqiao (跨湖 山); Houli (后李); 桥); Chengbeixi Xinglongwa (城背溪) (兴隆洼) Keqiutou (壳坵头) Lower Zaoshi Middle Neolithic (ca. Early and Middle (皂市下层); 5,000–3,500 bce) Yangshao (仰韶); Hemudu (河姆 Beixin (北辛); 渡); Majiabang Early Dawenkou (马家浜); (大汶口); Tangjiagang Zhaobaogou (赵 (汤家岗); Daixi 宝沟); Hongshan (大溪); Songze (红山) (松泽) Liangzhu (良渚); Dingsishan Phase Late Neolithic (ca. Late Yangshao; Qujialing (屈家 IV; Xiantouling 3,500–2,000 bce) Late Dawenkow; 岭); Shijiahe (咸头岭): Shixia Longshan (龙山); (石家河) (石峡);Yangliang Xiaoheyan (小河 (涌浪); 沿); Majiayao Tanshishan(马家窑); Niubishan (昙石 Banshan (半山); 山-牛鼻山) Machang (马厂)

West China

Baodun (宝墩)

professional specializations, could only have developed within the context of an intensive agriculture economy. Contrary to popular perceptions, ‘agriculture’ is not a one-time invention or historical event. In fact, archaeological and ethnographic studies challenge the notion of a clear-cut division between hunter-gatherer and agricultural societies. Contrary to their image as passive recipients of natural resources, some hunter-gatherer societies have also engaged in ‘resource management’ (Bellwood 2005: 12); the intentional burning of natural vegetation to stimulate the growth of desired plants, weeding and landscape modification, and selective intervention in the populations of plants and animals. On the other side of the same coin, societies that are often viewed as fully evolved agriculturalists continued to rely to a great extent on the procurement and consumption of wild resources. Thus, this chapter subscribes to the view that the transition to agriculture was a long process and that societies along this trajectory combined varying degrees of economic strategies, interactions with and modifications of the environment, social mechanisms and technologies (Smith 2001; Zeder 2015). Current research suggests that even the domestication of plants, considered one of the hallmarks of agriculture, was a much longer process than previously envisioned and sometimes took thousands of years (Fuller et al. 2014; Gross and Zhao 2014). Research on the long-term processes that eventually led to the flourishing of fully evolved sedentary agricultural societies is, thus, fundamental for our understanding of the history of 17

Gideon Shelach-Lavi

human society in general and of its varied manifestations in different regions of the world. China is among the handful of centers where agriculture developed independently and from which it spread to other regions. However, it is also the only center for which we are unable to fully reconstruct the entire trajectory from hunter-gatherer societies to agricultural communities (Bettinger et al. 2007; Shelach and Teng 2013). In spite of many recent advances in the identification and dating of early domesticated plants and animals (Flad et al. 2009; Liu 2015; Liu et al. 2015; Zhao 2011), few sites dating to the crucial transitional period are currently known. The social and archaeological context and consequences of those processes also remain poorly understood, as does the respective economic importance of domesticated food sources in different regions and during different phases of the trajectory. It is not even generally agreed upon whether there was a single process of transition to agriculture that encompasses north and central China (Cohen 2011), two or three independent centers of domestication in North China, the Yangtze River basin and the tropical South (Liu et al. 2015; Zhao 2011), or even multiple centers (Shelach 2000). What do we know, then, about the long-term background of the development of sedentary agricultural societies in different parts of China? Sites of the late Pleistocene to early Holocene are not much different from earlier sites. They are typically small, with little evidence of long-term residency or investment in permanent structures. Nanzhuangtou in Hebei Province, perhaps the best-known early Holocene site in north China, is a relatively small open-air site. Excavations here located the remains of fireplaces, but the shape and make-up of habitations is unclear, and investment in them was probably minimal (Hebei 2010). In central and south China, most known occupations are cave sites.Yuchanyan is a good example of a late Pleistocene site in this region. The area of this limestone cave is quite small – it is 12–15 m long and 6–8 m wide – suggesting that it was occupied by a small group of people. Fireplaces discovered inside the cave are the only clear indication of human modifications (Boaretto et al. 2009). A gradual change during this period is indicated, however, by the appearance of new technologies and cultural habits that started already during the peak of the last glacial age, some 20,000 years ago, and gained momentum during the Early Holocene.These technologies include the production of tiny stone artifacts (microblades), grinding stones, and ceramic containers, as well as the more common appearance of body ornaments (Qu et al. 2013; Sun and Wagner 2014; Wang 2005; Zhang et al. 2011). Collectively, the new technologies and the artifacts produced using these technologies suggest changes in human economic behavior, such as the development of new methods for procurement and processing of resources, and changing consumption habits, including a focus on new food resources and changes in social relations. Microblades or microliths (Chinese: xishiqi) are tiny flake artifacts, usually no more than 2 cm in length. The evolution of this technology began earlier, but it become ubiquitous during the late Pleistocene and early Holocene, and evidence of microblades industries during this time have been found in most parts of north China. For example, one study of sites associated with the Xiachuan culture in southern Shanxi Province, dated to ca. 18,000 bce, found that 95% of the stone tools were flaked blades and that the proportion of microliths among those tools is very high. This trend is also well represented in sites dated to the early Holocene and in other parts of north China (Bettinger et al. 2007; Chen 2007: 8–20). These tiny stone artifacts were probably embedded in wood or bone handles to form the cutting edge of composite tools such as knives, sickles or arrows (Figure 1.2) which could be used for harvesting and hunting, as well as the processing of food. Grinding stones, or querns (Chinese: mopan), are large flat stone slabs, one side of which is sometimes slightly concave and smoothly polished. Although they appear in much smaller quantities, their history and distribution are quite similar to that of microliths. A few such 18

Figure 1.2 Typical stone artifacts of the late Pleistocene and early Holocene period: A. microlith and flint microlith cores of the Xiachuan culture; B. a bone handle into which a microblade is inserted from Donghulin; C. grinding slab (quern) from Locality 1 at Longwangchan. A. and B., photographs by Prof. Zhao Chaohong; C. (after Zhang et al. 2011)

Gideon Shelach-Lavi

Figure 1.2  (Continued)

objects were recently found in late Pleistocene and early Holocene contexts in north China. For instance, querns were found alongside microliths at the Longwangchan site, in Shaanxi, in a stratum dated to ca. 23,000 bce (Figure 1.2C), as well as at different locations of the Shizitan site in Shanxi Province, dated more or less to the same period; and at sites such as Nanzhuangtou, and Donghulin in the Beijing area, dated to the beginning of the Holocene (Guo and Li 2002: 195–197; Hebei 2010; Liu et al. 2010; Liu et al. 2013; Shizitan 2010; Zhang et al. 2011). Those technologic changes are not reflected in the lithic industry of central and south China, which continued the older Paleolithic traditions dominated by large coarse tools, the so called “core-and-flake assemblages” (Qu et al. 2013). The production of ceramic vessels, however, began earlier in these areas, and, in fact, this seems to be the first region of the world in which pottery was produced and used. Potsherds found at cave sites such as Xianrendong (ca. 20,000 bp or ca. 18,000 bce) and Yuchanyan (18,300 to 15,400 bp, or ca. 16,000 to 13,500 bce) (Boaretto et al. 2009;Wu et. al. 2012) suggest the sporadic production of coarse, low-fired ceramics during the peak of the last glacial age. Early ceramic production was also identified at cave sites further to the south, such as Miaoyan in Guangxi Province, but its exact dating is still disputed (Qu et al. 2013: 53). In the north, modest quantities of potsherds found at terminal Pleistocene and early Holocene sites such as Nanzhuangtou and Donghulin (Hao et al. 2008; Hebei 2010) suggest a production mode that is not unlike that known from the Yangtze River area. What is the significance of these findings? Grinding stones and ceramics are traditionally associated with the processing of cereal grains and other domesticated food by agricultural societies. Scientific research, however, does not necessarily support such a connection. For, example, starch residue extracted from grinding stones excavated at Shizitan locality 14 and dated to the peak of the last glacial age has been identified as belonging to a range of plants including beans, tubers and different types of grasses, including Paniceae, the wild ancestor of millet (Liu et al. 2013). The same plant types were also identified on grinding stones from locality 9 of the same site, dated to the early Holocene; however, results from this locality, as well as from Donghulin, suggest that acorn was an additional major food source processed by the grinding stones (Liu 2015; Liu et al. 2010). On the other hand, analysis of starch residue extracted from potsherds found at Donghulin and Zhuannian identified both acorn and millet residue (Yang et al. 2014); thus, it is possible that the process of millet domestication was already underway at this time. Remains of animal bones and of plants found at those same sites suggest reliance on a broad spectrum of natural resources. Of the large quantities of animal bones and shellfish recovered 20

Main issues in the Chinese Neolithic

at Nanzhungtou, 67% belong to different types of deer, suggesting the continued importance of large-game hunting alongside the exploitation of a variety of other animals, including smaller ones. Recent research has identified some of the bones excavated from this site as belonging to domesticated dogs, thereby representing the oldest evidence for animal domestication in China (Yuan 2007). However, contrary to certain claims, bones of pigs from the same site belonged to wild boar and not to domesticated pigs (Yuan and Fled 2002). Because of the good preservation conditions inside cave sites, archaeo-zoologists and archaeobotanists have been able to recover large samples of animal bones and plant remains at sites across central and south China. For example, over 40 species of plants and 45 species of animals and shells were identified by the excavators of Yuchanyan, while 108 species of animals and shells (including various fish types, 20 bird types and 37 types of mammals) were found at Zengpiyan (Prendergast et al. 2009: 1034; Zhongguo 2003: 344–346). While it is possible that some of the animal bones were brought to the caves by carnivores, examinations of animal tooth marks and of cut marks left on the bones suggest that most were probably hunted and consumed by the human occupants of the caves (Prendergast et al. 2009). Botanical evidence from these cave sites suggest the exploitation of large number of plants, including acorns, water chestnuts and different grasses, including wild rice (Liu et al. 2015). In conclusion, it seems that, during the period between the peak of the last glacial age to the early Holocene, most areas in China were populated by small human groups. Those groups were quite mobile, although they may have spent more time in one place and invested slightly more energy in constructing their dwelling sites in comparison to earlier societies. They procured and consumed a broad spectrum of plants and animals, including some which were to become domesticated, though these may not necessarily have been those that were at this stage the most important caloric sources. Although current evidence suggests that these groups were unequivocally hunter-gatherer societies, this type of interaction with their natural environment, together with the development of new technologies, ‘prepared’ them, so to speak, for the transition to agriculture. It is possible that the cultivation of the wild progenitors of plants that were later domesticated, and even the incipient process of their domestication, had already started (Liu 2015;Yang et al. 2014). Evidence for the increased production and use of body ornaments, such as beads and hairpins, may furthermore suggest changes in social interactions that anticipated the more pronounced transitions that accompanied the beginning of agriculture and sedentary lifeways.

Early villages and village life in China With the transition to sedentary village life came the need to feed larger groups of people who resided at the same place all year round. Those groups needed to develop ways of managing their resources more efficiently, including not only procuring more food, be it domesticated or wild, from the immediate environment of their village, but also preserving and storing it for use during less productive seasons. This change was not merely economic, however. Living in large groups – larger than an extended family group – resulted, for example, in social tensions that needed to be resolved but, at the same time, also in opportunities for cooperation and exchange of goods and skills among the village members. These and other related social processes are fundamental issues for our understanding of human history and of the development of local cultures and identities. When the first sedentary villages which relied on domesticated food – if not fully, then at least to a meaningful extant – were first established in China is still an outstanding question. In north China, the tradition of low-level ceramic production and the exploitation of a variety 21

Gideon Shelach-Lavi

of food sources, including animal and plants that were later domesticated, which is associated during the tenth and ninth millennia bce with sites such as the aforementioned Nanzhuangtou and Donghulin, continued, perhaps with a slight intensification of production, during the eighth and seventh millennia at sites such as Bianbiandong in Shandong province and the Xiaohexi tradition of northeast China (Shelach and Teng 2013; Sun and Wagner 2014). By the mid- to late seventh millennium bce, relatively large sedentary villages appear throughout all of north China and are identified with various local ‘cultures’ (Table 1.1). It is in the context of these traditions that we also have the first conclusive evidence of domesticated plants (millet) and meat-producing animals (pigs) (Yuan and Flad 2002; Zhao 2011).The exact nature of this seemingly rapid transition is not yet fully understood, but as mentioned in the previous section, even after the establishment of relatively large sedentary communities throughout north China, a substantial part of their diet was still obtained from wild resources, and the process of transition to full reliance on domesticated resources took at least another millennium. No less important, the process of domestication and modification of staple food sources such as millet continued to evolve during the Neolithic, and certain other basic foods, including, for example, soybeans, were probably only domesticate during the later phases of the Neolithic, if not during the Bronze Age (Liu et al. 2015;Yuan 2007). In central China, the transition to agriculture and sedentary lifeways seems more gradual than in the north. The consumption of wild rice had already started during the terminal Pleistocene and Early Holocene, and large quantities of it were found at sites dated to the eighth and seventh millennia bce, such as Pengtoushan, Bashidang and Shangshan. However, it remains unclear when rice was domesticated: some researchers identify the botanical remains from those sites as belonging to domesticated rice, or at least rice that was at “an early stage of domestication” (Jiang and Liu 2006: 358), while others argue that that full domestication of rice only took place at around 4,000 bce, or some 4,000 years after it was extensively collected and perhaps even cultivated (Fuller et al. 2007). The large quantity of rice husks found at the Hemudu site (some estimates are as high as 20 tons; Fuller et al. 2007) are seen by many as evidence of intensive rice cultivation during the fifth millennium bce. Bone spades found in large numbers at this site (Figure 1.3B) are, likewise, evidence of intensified rice cultivation. Direct archaeological evidence of the development of paddy fields, one of the prominent features of the agriculture system in this region, including raised field boundaries, irrigation ditches and wells, were found at both the middle and lower Yangtze River areas, at sites such as Chengtoushan, Tianluoshan and Chuodun (Fuller et al. 2007; Hunan 2007; Zheng et al. 2009).Thus, it seems that at least by the fourth millennium bce, the two distinct agricultural systems – the dry agriculture of the north and the wet agriculture of central and south China – were already well established. However, the intensive exploitation and consumption of wild resources continued in the Yangtze area throughout the long period of rice domestication and human intervention in the environment. For example, at the waterlogged Kuahuqiao site, dated to ca. 6,000–5,000 bce, alongside large quantities of rice grains, most of the plant remains found belong to wild species, including water caltrop, acorn, water chestnut, Job’s tears and knotweed (Zhejiang 2004). Wooden-frame storage pits excavated at this site were filled with acorns (Figure 1.3C), illustrating the importance of wild food resources for the local community’s economy. Early villages in north China are surprisingly large and highly organized. A good example is the Chahai site located in western Liaoning province and dated to the beginning of the sixth millennium bce. It is 1.25 hectares in size, surrounded by a narrow ditch, inside of which fiftyfive houses and thirty-four storage pits were neatly clustered around an area where a grave and a pile of stones, sometimes described as having the shape of a ‘dragon’, are located (Figure 1.3A). 22

A.

I

II H23

H15

G2

H22

F26

H20

H21

H19

H14 H18

F34 H17 H27 H29

F35

H30

F38

F29

H24 H16 F24

F40 H33

H26

F39

H25

F32 H13

H32

G1 F50

F51 F46 F45

F36

F30 F31

F28

F47

F43

F52 F49

H36 M2

F27

F37

H31

F44

F54

H28

F25

F42

F41

F33

M3 M1 M7 M4 M10 M9 M5 M6

F53

F55

F48

M8

F23 T0508 T0408

F20

F22

H34

F8

H5

F21

No

D2

F9

F11

tY et

H7 D3

H8

T0407

F12

H35

D4 D1

H6

F2

F1

F5

H10

H4

F13

Exc av ate d

F15

F3 H9

F18

F10

F6

H3

F4

F16 F17

H2 H1

F7

F14 H12

H19

III

B.

IV

H11

0

5m

C.

Figure 1.3 A. Map of the Chahai site (after Liaoning 2012: 6); B. storage pit from the Kuahuqiao site with acorns in it (after Zhejiang 2004: 27); C. Bone spades from Hemudu (after Liu 2006: 65).

Gideon Shelach-Lavi

The rectangular semi-subterranean houses, which were built with wooden poles and equipped with hearths, varied in size between 15 and 150 sq m (Liaoning 2012). These remains suggest a community of at least 200 people which, judging by the substantial investment of labor in house construction and other immovable structures and objects (large grinding stones found on the floor of every house), resided in the same place for a relatively long period. During the subsequent Middle Neolithic period, village life continued to evolve. Sites are larger – some in the Wei and Yellow River basins are 40 hectares or larger – and more complex, and investment in houses, public structures (e.g., large ditches) and production facilities (e.g., kilns) are greater (Ma 2005; Peterson and Shelach 2010; Peterson and Shelach 2012; Shelach 2006).The clear division between domestic and cemetery areas that existed in some of these villages, the increased investment in and elaborateness of burial practices, and the production of decorated ceramics, body ornaments, figurines and musical instruments all suggest the development of belief systems and cultural mechanisms that regulated community life and increased shared communal identity (Shelach-Lavi 2015: 70–96). In central China, a shift from cave sites, which were common during the Pleistocene and early Holocene, to open-air sites seems to have taken place. However, villages in this area were not as extensive and well structured as Chahai and similar villages in north China during the seventh and sixth millennia bce. Many of the sites, such as Pengtoushan, are quite small and reflect societies that are comparable in size to those of the cave sites. Other open-air sites, nonetheless, are much larger, representing a substantial expansion in community size. Although remains of a surrounding ditch were found at Bashidang, the internal organization of sites in this region seems to be more fluid than in the north. In the middle Yangtze region, most houses are rectangular in shape and constructed on the surface level, although circular and semi-subterranean houses are also known. Postholes indicate a wooden structure that probably supported wattle-and-daub walls. The houses were quite small, ranging in size from about 8 to 40 sq m (Hunan 2006; Zhang and Hung 2008). During the same period a new tradition of house construction emerged in the lower Yangtze River region. At the waterlogged Kuahuqiao site, the excellent preservation of organic materials has made possible the recovery of wooden structures that seem to have supported houses in which the residential floors were raised above ground (Zhejiang 2004). These so-called ganlan, or pole houses, are well adapted to a swampy and often flooded environment, and subsequently became one of the most common types of dwelling in central and south China (as well as in regions of Southeast Asia). Remains of ganlan constructions are even better preserved at the Hemudu site. The wooden posts and planks, and especially the technique for joining them together (called mortise and tenon joints), indicate a high level of carpentry skills. Some of these houses were very large, reaching 23 m in length and 7 m in width (Chang 1986: 208–211). Such large habitation structures suggest, perhaps, the development of new social groupings. The situation in south China is quite different from that of north and central China. Similar to central China, a transition to open-air occupation occurred during the early Holocene, but the most typical archaeological sites are shell middens (Chinese: beiqiu yizhi) rather than sedentary villages with house remains. No habitation structures were identified at typical shell middens such as Dingsishan in Guangxi Province, though graves and a few ash pits were found together with ceramic vessels and shell, bone and stone artifacts (Zhongguo 1998). The economies of these communities were based on extensive exploitation of natural resources, with rice cultivation entering the area, probably from the Yangtze River basin, only during the fourth millennium bce (Zhao 2011). The ‘Neolithization’ of west China was also a relatively late process. In the fertile Sichuan basin, no evidence of human occupation is known prior to the third millennium bce. The 24

Main issues in the Chinese Neolithic

seemingly sudden appearance of dense human settlement in this region, including the construction of large and elaborate sites such as Baodun, is associated with intensive cultivation of not only domesticated rice but also of millet (d’Alpoim-Guedes et al. 2013; Flad and Chen 2013). How agriculture spread to this region and further west to Yunnan (Yao 2010), as well as whether this was the result of large-scale human migration from the east, are outstanding questions that await further studies.

The development of social complexity In two papers I published with Christian Peterson (Peterson and Shelach 2010; Peterson and Shelach 2012), we identified evidence of the early development of social, economic and political complexity; for example, the incipient division of labor, craft specialization, inter-family exchange and intra-communal hierarchy, which already appear during the Middle Neolithic period in north China. However, more substantial evidence for the development of social complexity and political hierarchy in north, central and west China appeared during the late fourth and, more dramatically, the third millennia bce. What does the term ‘social complexity’ imply? It is possible to differentiate between horizontal complexity – increased specialisation, people’s dependence on the expertise of others, and the need for different scales of cooperation – and vertical complexity – the formation of social and political hierarchies and the ability of a small segment of the society to control the majority and exploit their work and resources. Of course, those two types of complexity are inter-connected: increased craft specialisation often emerges in response to elite demands for prestige goods, and political subjugation is sometime made possible by technological improvements in the production of surpluses, weapons, mean of transportation, etc. However, I maintain the distinction between the two types of complexity, as a heuristic device, in my discussion here. Artifacts whose production require exceptional skills and large investment of labor are the clearest archaeological evidence of horizontal complexity, because they suggest that there were people who developed these skills and invested their time in the production of specific artifacts, and, therefore, other people must have supplied them with food and other life necessities, thereby creating a complex network of interdependency. Other indications of interdependency include evidence of economic specialization, the location of workshops, evidence for exchange of resources and artifacts, and more. Already during the middle Neolithic, we find substantial evidence of incipient craft specialization. this includes, for example, the production of high-quality painted ceramics at various areas of north China and the development of hard stone (jade) carving industries, especially in northeast China (Shelach and Teng 2013). In central China, at sites such as Hemudu, there is evidence not only of advanced ceramic production but also of skilled carpentry and even the earliest archaeological evidence of lacquer artifacts (Liu 2006: 102–104), a highly skilled and time-consuming industry which became one of the hallmarks of Chinese craft during the historical era. Evidence of a rapid increase in technological sophistication and craft specialization is found for late Neolithic societies throughout north, central and west China and, to a lesser extent, in south China as well. For example, pottery production, which by that time had a history of some 15,000 years, underwent dramatic changes with the introduction of the fast potter’s wheel and the development of kilns that could sustain higher firing temperatures for long periods (Underhill 2002).The Shandong Longshan culture is famous for its delicate eggshell cups (Figure 1.4A) and fine black burnished pottery, which have been found mostly in graves but also in domestic contexts. While each artifact was individually produced and had unique decorations, the 25

Gideon Shelach-Lavi

repetition of shapes and the consistency in size suggest the development of ‘production standards.’ Only highly trained potters could have produced such artifacts; firing them would have required not only the ability to reach high kiln temperatures but also absolute control over the firing process, so as to achieve the desired effects and prevent the fragile artifacts from breaking while being fired and cooled. All this suggests a highly skilled labor force of specialists.The division of the production process into sub-specializations (potters, kiln masters, etc.), furthermore, indicates a complex industry and the need for work coordination. The jade industry of the lower Yangtze region is another striking example of technological sophistication and craft specialization. The Liangzhu jades are much more numerous, labor intensive and technically sophisticated than earlier examples of jade. The carving of each jade object would have occupied highly specialised artisans for many hours and required the use of advanced technology. For example, drilling the shafts inside the famous Liangzhu jade tubes (cong) (Figure 1.4B), some of which are up to 25 cm long, is a challenging task, even with modern technology. At the Sidun site, a single grave contained no fewer than 109 jade artifacts, including thirty-three cong and twenty-four large discs (bi) (Figure 1.4C). At another site, Fanshan, eleven pit graves placed inside an artificial pounded earth mound contained some 1,200 grave goods, 90% of which were jades, suggesting the immense scale of jade production in this region (Huang 1992; Zhejiang 2005). Craft elaboration is not the only evidence of increased specialization and interdependency during the late Neolithic; religious specialists are another example. The importance of such specialists, be they shamans, priests, fortune tellers or others, is indicated by burial practices. In each region of China, we see not only increased investment in the building and furnishing of graves, but also a process of standardization (for each region or archaeological ‘culture’) of the way graves are constructed and the type and composition of artifacts placed in them. Such standardization suggests the accumulation of religious knowledge and norms, and the existence of persons whose expertise was to guide the construction of graves and to conduct burial ceremonies. Another, perhaps more explicit, sign of such persons is oracle-bone divination. Sites in northeast and northwest China, dated to the fourth and third millennia bce, have yielded animal bones (mostly bovine) with intentional burning marks that are associated with the practice of pyromancy (Flad 2008). Pyromancy became well established at the centers of the earliest states of China during the Bronze Age (Bagley, this volume, Chapter 3), but even at this earlier stage, the existence of such complex divination practices suggests the work of specialists. Evidence of increased vertical complexity (stratification) during the late Neolithic is likewise very extensive.The elaboration of burials, and especially the growing disparity between rich and poor graves, is one such indication. The scale of large graves in north China can be seen, for example, at the Xizhufeng site. Grave number M202 at this site is almost 7 m long and 3 m wide. It has a built ledge (Chinese: ercengtai) upon which some of the grave goods were placed, as well as inner and outer painted wooden coffins. Another grave at the same site, labeled M1, contains an even more elaborate structure of three nested wooden coffins and a special wooden chamber in which no less than fifty-four burial goods, including fine ceramic wares, were placed (Zhang 2006; Zhongguo 2010: 609–610) (Figure 1.5A and 1.5B). In some societies of central China, elite burial mounds and the furnishing of graves, such as described earlier for the sites of Sidun and Fanshan, suggest an even larger investment. The scale of the disparity in investment of labor and resources can be better seen by comparing a large number of graves belonging to the same community. The cemetery at Taosi, in which more than 1,000 graves were excavated, is a good example. Preliminary analysis of these graves suggests that they can be divided into three ranks: small pit graves, only big enough to contain a corpse in the extended position, made up 87% of the sample. Each of these small 26

A.

C.

0

5cm

B.

Figure 1.4  A. Longshan eggshell ceramic cups (after Zhang 2006). B. A Liangzhu jade cong (photo by Gideon Shelach). C. A wealthy Liangzhu grave from the Sidun site (after Nanjing Bowuguan 1984: 14).

Gideon Shelach-Lavi

A.

B.

Figure 1.5 A. Grave M1 from the Xizhufeng site; B. Some of the ceramic artifacts found in grave M1 (after Zhang 2006: 93–94)

graves contained few or no grave goods at all. Medium-sized graves comprise some 12.7% of the sample.They are larger and contain a wooden structure (coffin) with several dozen artifacts – primarily pottery, but also a number of prestigious offerings, such as jade ornaments and pig mandibles. Upper rank graves comprise only 1.3% of the sample; they are large pits, more than 3 m long, 2.5 m wide and up to 7 m deep, each containing a wooden coffin and hundreds of burial goods, including painted red pottery, jade artifacts, musical instruments such as alligator skin and ceramic drums and large chime stones, painted wooden artifacts, complete pig skeletons and more (Chang 1986: 276–277; Liu 2004: 135–137; Zhongguo 2010: 572–574). These disparities suggest that in Taosi and other similar communities in north China, elites could expend resources on death rituals (and presumably other aspects of life as well) that were at least an order of magnitude greater than those of mid-ranking people and commoners. In Taosi, both men and women were found buried in the small and medium-sized graves, but the upper rank graves contained only male skeletons, suggesting that gender played an important role in the construction of socio-political stratification during the late Neolithic. The growing gap between the sexes may have been the result of a growing division of labor, or it may have had to do with an intensification of inter-societal conflicts and violence. Comparisons of early and late Neolithic human skeletons do suggest an increase in physical stress (related perhaps to the intensification of agriculture) and dietary deficits. They also indicate greater differences between men and women, and between the elite and commoners, in terms of the foodstuffs they consumed (Pechenkina et al. 2002; Smith 2005). While mortuary data indicate a dramatic increase in socio-political stratification within the individual community, the development of larger polities and of regional-scale socio-political 28

Main issues in the Chinese Neolithic

hierarchies is best seen through settlement-patterns studies. It has been proposed, based on such studies, that late Neolithic polities in China had a three- or four-tier hierarchy (Liu 2004: 172– 178, 240; Underhill et al. 2008). I concur with the critique about our ability to accurately determine, based on currently available data, the exact number of hierarchical levels (Peterson and Drennan 2011).The development of regional-scale hierarchical polities is, nonetheless, apparent. The overall trend evident throughout most regions of north, central and west China is one of a substantial increase in population density, which is coupled with much greater variation than before in site sizes (with some sites now being very large) and in the amount of labor invested in public structures, such as walls and moats. For example, results of a systematic regional survey carried out at the Rizhao area of southern Shandong demonstrate phenomenal growth, from twenty-seven sites covering a total area of 47.3 ha during the fourth millennium bce to 463 sites covering an accumulated area of 2,005.4 ha during the third millennium bce. During the earlier phase, these sites were more or less of similar size and there was no apparent clustering of sites, but during the latter period, settlement in this region clearly became hierarchical and clustered around two very large sites – Liangchengzhen (272.5 ha in size) in the northern part of the region and Yaowangcheng (367.5 ha) in the southern part – indicating the formation of two regional-scale polities (Underhill et al. 2008: 6–8). Many, though not all, regions in north China display similar trends, although their scale differs from region to region (Shelach and Jaffe 2014). Many of the sites at the center of the late Neolithic polities were fortified by large walls. Remains of walls have been found at Liangchengzhen, and more than twenty other walled sites dated to the late Neolithic period have been found in the lower and middle Yellow River regions. Many of these fortified sites are distributed regularly at a distance of some 30 to 50 km from one another, suggesting perhaps the geographical size of each polity (Liu 2004). Taosi is one of the larger walled sites in north China. The walls at this site are made of pounded earth and are up to 10 m wide. At the peak of the site’s expansion they enclosed an area of some 280 ha, with internal pounded earth walls separating the residential and ceremonial quarters of the elite from the areas inhabited by commoners, thereby signifying the development of a stratified society (He 2013; Zhongguo 2005). A recent survey counted fifty-four sites that were contemporaneous with Taosi in this region; at least three of them, not including the Taosi site itself, are more than a 100 ha in size; twenty-three are between 10 and 99 ha, and the rest are smaller (He 2013). The range in site size, the more or less even distribution of the largest sites, and the association between labor investment and the largest site (Taosi) all suggest the development of a regional settlement hierarchy and the ability of the center(s) to recruit labor and accumulate resources. An even larger fortified site, named Shimao, has been reported preliminarily. It is said to include surrounding stone walls that enclosed an area of 400 ha, internal stone walls, fortified gates and watchtowers (Shaanxi 2013). If the preliminary reports are corroborated and the dating of the site to the late third millennium bce is established, these features would make Shimao the largest and most labor intensive of all late Neolithic sites found in China. Similar trands are found in central China as well. On the Liyang plain in northwest Hunan Province, the number and area of sites grew from forty-five sites with an accumulated size of 91 ha during the early and mid-third millennium bce to 163 sites covering an accumulated area of 169 ha during the late third millennium (Pei 2004). Large-scale moats and walls appeared in this region already by the fourth millennium bce. For example, at the Chengtoushan site, the moat and wall that enclosed this circular area were rebuilt and expanded four times between ca. 4,000 and 2,800 bce. During the final stage, the moat was 35 to 50 m wide and 4 m deep, while the walls were 5 m high, 27 m wide at the bottom and 16 m wide at the top (Hunan 2007: 84–87). 29

Gideon Shelach-Lavi

During the second half of the third millennium bce, Shijiahe was the largest fortified site in the central Yangtze River area. Well-preserved sections of the Shijiahe walls are 50 m wide at the base, 4 to 5 m wide at the top, and over 6 m tall. It is estimated that about 700,000 cubic m of earth were used in their construction. The wall itself is surrounded by a huge moat, which at some points reaches a width of 100 m. Inside the walls, the Shijiahe site is more than 100 ha in size, far larger than most contemporary sites. One other site in the same region, Taojiahu, is estimated to be around 60 ha in size, but the rest are 20 ha or smaller, suggesting again the existence of a settlement hierarchy (Hubei 2003). In west China, similar trends are observed in the Chengdu plain of Sichuan. A relatively large number of late Neolithic sites are known from this area, which was sparsely populated during the early and middle Neolithic. Some of the late Neolithic sites are walled and quite large, suggesting a rapid increase in population densities and the formation of a regional hierarchy (Flad and Chen 2013; Wang 2003b). So far, at least six walled sites are known in this region, of which Baodun is the largest and best known. The Baodun enclosure is rectangular in shape and covers an area of more than 60 ha. However, recent and as yet unverified reports note the discovery of a much larger outer wall, which was part of the late Neolithic occupation of Baodun (Flad and Chen 2013: 74). The better preserved sections of the piled earth wall at Baodun are 3 to 5 m high, 30 m wide at the base, and about 8 m wide at the top (Wang 2003b). Large walls that stood 10 m tall or more must have been impressive symbols, visible from far away, of the paramount position of these fortified sites in the political hierarchy of the emerging polities. In central China, large walls and moats may have functioned as protection against floods (Wang 2003a). However, alongside these symbolic and functional purposes, such walls should also be seen as fortifications that enhanced the military powers of the elite. Fortified gates and other clearly militaristic installations found not only at Shimao but at other sites as well (Chang 1986: 265–266; Shaanxi 2013) are clear evidence of this function. The large quantities of stone arrowheads found at late Neolithic sites also provide testimony of endemic warfare during this period (Liu 2004: 64). Intra-societal violence during the late Neolithic period is also indicated by an increase in the number of mutilated skeletons found at late Neolithic sites, as well as by evidence of human sacrifices.While human sacrifices are known from middle Neolithic sites, their number increased dramatically during the late Neolithic, and there is evidence of these practices from almost all regions of China. At many sites, mutilated human skeletons are found inside postholes and house foundations (Liu 2004: 46–47), and mutilated human skeletons are found next to the main occupants in elite graves. Regardless of the specific ritual functions of these sacrifices, the fact that the elite could regularly enforce the killing of other people, presumably from the lower strata of society or ‘war captives’ from other societies, is evidence of the very real power over life and death that the elite held. Discoveries of larger concentrations of mutilated human bodies may indicate the execution or sacrifice of prisoners. Such evidence has been found at Jiangou, for instance, where five layers of human skeletons were found buried inside a well. Some of the bodies display signs of violence, while others may have been buried alive (Chang 1986: 270–271). An increase in inter-site and inter-polity violence may have been the outcome of growing social stratification, but it probably also facilitated the rise of ambitious leaders and the concentration of more and more power in the hands of the elite. Unique artifacts found exclusively in elite graves, such as elaborate jades or eggshell ceramic cups, demonstrate the elite’s monopoly over symbols of power and prestige. It is notable that musical instruments are among the elite symbols of power found in wealthy graves in many parts of China. This suggests that performances and rituals played an important part in establishing, maintaining and reproducing sociopolitical prestige and power across most regions. 30

Main issues in the Chinese Neolithic

It should also be remembered that while the third millennium bce was a period of increased population densities in most areas, some regions, such as northeast China, experience a period of population decline (Chifeng 2011). Thus, much more research is needed on the causes and consequences of the development of social complexity and political hierarchy in China.

Regional variation and cross-cultural interactions Because of the abbreviated nature of this chapter, the discussion thus far has focused mainly on similar characteristics and shared developments among societies from different regions of China. However, it is not difficult to imagine that the vast and varied geography of China has led to diverse modes of adaptation (see Pechenkina in this volume) and a variety of cultural norms that evolved in different regions and localities. This diversity is best exemplified by attributes such as the shape and decorations of ceramic vessels and prestige items, the shape and content of graves, and even the style of domestic structures.This kind of diversity, which is already clearly indicated during the early and middle Neolithic (e.g., Shelach- Lavi 2015: 79–91), became much more pronounced during the late Neolithic, with the production of sophisticated ceramic vessels and elaborate burial modes. A comparison between elite graves from the lower Yangtze River area (see Figure 1.4C) the lower and the upper Yellow River (Figures 1.5 and 1.6) clearly illustrates such regional differences. While all three graves are large and very richly furnished, their shapes and the placement of artifacts within them is very different, and so are the artifacts themselves, representing very different consumption rules, aesthetics and perhaps religious values.While the emphasis in lower Yellow River graves is on the elaborate and difficult-to-produce shapes of the ceramic vessels, the shapes of vessels in the upper Yellow River grave are consistent, but distinguished from one another by their elaborate coloured decorations. Moreover, while the grave goods in the lower Yellow River area may have been valued for their unique aesthetics and their function as part of the burial rituals, it is possible that what was valued in the upper Yellow River area was the actual resources (grains) placed inside the large vessels (Allard 2001; Underhill 2002).

N

0

10m

Figure 1.6 Grave M564 from the Liuwan cemetery and the grave goods found in it (after Zhongguo Shehui 2010: 631)

31

Gideon Shelach-Lavi

The elite graves in the lower Yangtze are not part of a larger cemetery, as is the norm in the Yellow River region; rather, they are placed inside man-made earthen platforms, suggesting a very different ritual preparation, as well as, perhaps, clearer segregation between elites and commoners. The furnishing of graves with jade artifacts rather than with ceramic vessels along with the fact the many of those costly artifacts were intentionally broken as part of the burial ceremony (Huang 1992) is another distinctive feature of the local ‘cultures’ and their aesthetic and ritual norms. These are but a few of the many examples of clear cultural differences that existed, not only between the broadly defined regions but also within each of the regions. These variations are not only more extensive than is possible to present here, but they may also reflect much deeper differences in attributes such as modes of socio-economic organization and the way political hierarchies are constructed. For example, using a heuristic distinction between two types of socio-political and economic strategies – ‘corporate’ and ‘network’ (Blanton et al. 1996) – I have elsewhere suggested that in many societies in the eastern parts of China, the position of paramount leader was gained and legitimated through the production, exchange and manipulation of non-utilitarian prestige artifacts, akin to the ‘network’ mode. In societies of northwest China (the Wei River and Gansu area) and west China (the Chengdu basin), it seems that leaders used a more ‘corporate’ strategy, which was linked to the management and exploitation of the subsistence economy (Shelach- Lavi 2015: 156–158). This is, of course, a very generalized description, which overlooks much of the variation that existed among different societies in the same region, as well as the non-linear nature of their transformation over time. It also deemphasizes the fact that, during the same time when local variation became much more pronounced, contacts and mutual influences between regions intensified. These two processes that seemingly contradict one another may, in fact, be two sides of the same coin: intensified regional and inter-regional interactions result in the creation of a shared cultural ‘language’, but they also, at the same time, catalyzed the need for demonstrations of local identities. Needless to say, such processes are quite familiar in the modern globalizing era. Evidence of increased inter-regional contacts during the late Neolithic of China were first coherently analyzed by the late K.C. Chang. He described these contacts through a “Longshanoid horizon, which began in the north and the Yangtze valley by the middle of the fourth millennium B.C. and continued along the eastern coast all the way to Taiwan and the Pearl River delta up to the middle of the third millennium B.C” (Chang 1986: 238). The clearest evidence of this shared ‘horizon’ is common ceramic vessels types (such as dou and ding) found throughout those regions. It is now clear that similar types of vessels are known from across an even larger region, including, for example, the middle Yangtze River and the Chengdu plain (Wang 2003b). Other evidence of the formation of inter-regional networks of contacts and the adaptation of certain shared cultural attributes includes the construction of large site walls using the pounded earth (hangtu) technology, as well as the use of jade objects and their unique shapes (cong tubes, for example), which appear quite far from their probable place of origin in the lower Yangtze region. Evidence of long-range exchanges of resources includes, for example, drums made of wood and covered with alligator skin that were found both in the largest graves at Taosi and in graves of a comparable size in Shandong Longshan sites. It has been suggested that during this period the coastal area of Shandong was a natural habitat for alligators and that alligator skins were transported from there some 500 km inland, to Taosi (Liu 2004: 122). Contrary to traditional centralistic models, this network of contacts, exchanges, and influences was probably multi-centered and multi-directional. Within this ‘interaction sphere’, to use Chang’s term, some regions probably interacted more intensively with certain others, 32

Main issues in the Chinese Neolithic

depending on distance, ease of transportation and perhaps also their history of past interactions. For example, contact between the lower Yellow River and the middle Yellow River regions seems to have been quite extensive, as suggested by the types of ceramic artifacts found in graves as well as the alligator skins discussed earlier. Moreover, the contacts were not limited to regions where stratified regional polities emerged. It appears that areas in south China, where societies were not yet highly stratified, were also part of this network. For example, some cong and bi jades at the Shixia site are identical to examples from the Liangzhu culture, suggesting that they were brought to the site from the upper Yangtze region. Cong of lesser quality, also found at this site, illustrate a process whereby foreign objects, which may have reached the region via indirect down-the-line exchange, were imitated and incorporated into the local culture (Allard 1997). We do not have direct evidence of artifacts or materials that made the trip in the opposite direction, from south to north, but during the historical period, the ‘Lingnan’ region is famous for the pearls and colorful birds’ feathers it traded with the north. Cowrie shells known from Neolithic sites in north and northwest China (Peng and Zhu 1995) are further direct evidence of such contacts. Because they are shells of sea creatures that only live in hot areas, their origin is either from the coastal area of south China or from the Indian Ocean.

The ‘collapse’ of the late Neolithic societies In recent years it has become popular to argue that late Neolithic societies collapsed in different parts of north and central China sometime around 2,000 bce, thereby paving the way for the appearance of state-level societies (e.g., Liu and Chen 2012: 250–252). The dramatic decline in social complexity and population density is often attributed to dramatic climatic fluctuations, which are sometimes equated to the Chinese legend of the ‘great flood’ (Stanley et al. 1999;Yu et al. 2000). Critical examination, however, suggests that while several regions did indeed experience a period of decline around this time, in other regions there was a clear continuity from the third to the second millennia bce and even an increase in population densities and complexity (Shelach and Jaffe 2014). Clear examples of a rapid process of depopulation are found along the eastern coast of north and central China (Underhill et al. 2008; Zhongguo 2010: 692), and this, indeed, may have been the result of climatic change and the rise of sea and underground water levels. In areas removed from the coast the process seems to be much more mixed. Taosi, the largest walled Neolithic site in all of north China, seems to have undergone a process of decline. Sometime around 2,000 bce, the large pounded earth enclosure was destroyed, and stone and bone debris found in the public buildings area suggest that it was converted into a workshop for craft production (Liu 2004: 110–111). Nonetheless, a range of evidence, including radiocarbon dates, suggests that the site remained occupied until around 1,700 bce (Zhongguo 2010: 566, 838), and its area may even have expanded during this period (He 2013). Another site in the central Yellow River area, Xinzhai, reached the peak of its development during the early decades of the second millennium bce. During this time Xinzhai expanded to about 100 ha and was surrounded by two concentric ditches and pounded earth walls (Zhao 2009). Thus, rather than perceiving a homogenous process of collapse of one tradition and the rise of a new one, we should examine the process, for each region, of long-term transitions. This perspective allows us to see that the multi-regional process was varied and included periods of expansion as well as decline. Such perspective is instrumental in understanding the dynamic of pre-state societies in China as well as the rise of the earliest states in this region (Shelach and Jaffe 2014) 33

Gideon Shelach-Lavi

The legacy of the Neolithic period and the ‘Chineseness’ of the Neolithic cultures So far in this chapter I have used the term ‘China’ as a shorthand for all the areas that were historically part of the Chinese states. But how ‘Chinese’ were those prehistoric cultures? In other words, what is the relevance of discussing the Neolithic period in a volume devoted to early Chinese history? One way of justifying this discussion is to argue that the social, economic and political processes I have described laid the foundations for the Chinese states and societies as we know them during the historical periods. However, the intellectual paradigm current in China today is one which argues for a much more concrete connection between the prehistory and history of China. According to this model, the genesis of ‘Chinese civilization’ occurred during the Neolithic period, when specific ‘Chinese’ attributes evolved. Indeed, the origins of ‘Chineseness’ during the Neolithic period is, in fact, a very popular research theme. One survey of the academic literature, for example, counted more than 800 publications on the origins of Chinese civilization by members of the Chinese Academy of Social Sciences alone (Chen 2009). The late Neolithic, in particular, is seen as a crucial period during which a unique Chinese culture emerged. Some researchers associate this period with the legendary heroes of Chinese mythology and identify sites, mainly in the middle Yellow River region, as the ‘capital’ cities where those legendary figures lived. Thus, for instance, despite a lack of any real corroborative evidence, the Taosi and Wangchenggang sites have been identified with such figures as Yao, Shun and Yu (e.g. Cheng 2005; Fang and Liu 2006; Ma 2008). More cautious scholars avoid using unverified histories that were written thousands of years after the fact but nevertheless identify as ‘Chinese’ attributes that are thought to have emerged during this period. These include socio-political hierarchies, walled cities and their association with political power, belief systems, including ancestor worship and the use of music in rituals, extended and internally stratified families, an incipient Chinese writing system, and traditional forms of artifacts, structures and symbols (e.g. Dematte 1999; Keightley 2006; Liu 2000; Zhongguo 2010). Evidence of the emergence of some of these attributes, such as the development of a Chinese writing system, is problematic, while others, such as the evolution of socio-political hierarchies, are not uniquely ‘Chinese’. Nonetheless, the transmission of Neolithic forms and symbols is a phenomenon that merits serious consideration. Clear examples of unique forms which first appeared during the Neolithic period and are known as important symbols during the early historical period and the Imperial era include, for example, the aforementioned cong jades. During the historical era, such artifacts are described in texts and are imbued with symbolic meanings (Chang 1989). A highly celebrated example is the Chinese dragon (long), which according to many studies evolved as an important symbol during the Neolithic. For example, hundreds of clam shells arranged on either side of the principal occupant’s body inside a large middle Neolithic grave (M45) from the Xishuipo site were identified as depicting a tiger and a dragon (Puyang 1989). A more or less contemporaneous large ‘C’-shaped jade artifact from Inner Mongolia was named the ‘jade dragon’ (yu long) and identified by some as the predecessor of the dragon symbol in China (Shelach-Lavi 2015: 337). Because transmission from earlier to later societies is seen by many as a natural, almost biological, phenomenon, scholars tend to project our knowledge of historical periods backwards onto prehistoric times. This kind of backward projection assumes immense ideological stability during times of tremendous socio-political and economic change as well as the direct and complete transmission of a symbolic representation over more than two millennia and across 34

Main issues in the Chinese Neolithic

a very large region. However, as I have argued elsewhere (Shelach-Lavi 2015: 159–160), the transmission process was far from being simple or direct.The form of cong, for example, seems to have been forgotten several times during the 2,000 years between the end of the Neolithic and the beginning of the Imperial era only to be rediscovered, probably when ancient jade artifacts were found, and reintroduced into the contemporary culture. It stands to reason that, after each such period of ‘amnesia’, when cong-shaped artifacts were produced again, they were imbued with new meanings that fit the current ideology.Thus, although we find Neolithic and historical artifacts and symbols that resemble each other, we cannot assume that their functions, religious significance and social meanings were the same. That said, we cannot ignore a long-term continuity between the Neolithic and the Bronze Age periods in different parts of China, nor can we overlook the importance of such continuity in the long-term evolution of social, cultural and religious aspects of China. One example of this continuity in shape and function is the remarkable stability of vessels forms used in Neolithic and Bronze Age rituals, such as three-legged ding and high-stem dou vessels. This continuity in forms, in spite of tremendous technologic change (from ceramic to bronze), does not by itself imply that the rituals in which those vessels were used or the belief system that underpinned those rituals remained the same – in fact, they most probably did not – but the transmission over such a long period of forms and symbols is nevertheless meaningful.This type of ‘Neolithic legacy’ is, to my mind, a topic worthy of further research.

Works cited Allard, F. (1997) ‘Growth and stability among complex societies in prehistoric Lingnan, Southeast China’, Papers from the Institute of Archaeology, 8: 37–58. Allard, F. (2001) ‘Mortuary ceramics and social organization in the Dawenkou and Majiayao cultures’, Journal of East Asian Archaeology, 3: 1–22. Bellwood, P. (2005) First Farmers:The Origins of Agricultural Societies, Malden: Wiley-Blackwell. Bettinger, R.L., Barton, L., Richardson, P.J., Boyd, R., Hui, W. and Won, C. (2007) ‘The transition to agriculture in northwestern China’, in D.B. Madsen, F.H. Chen and X. Gao (eds) Late Quaternary Climate Change and Human Adaptation in Arid China, Amsterdam: Elsevier. Blanton, R.E., Feinman, G.M., Kowalewski, S.A. and Peregrine, P.N. (1996) ‘A dual-processual theory for the evolution of Mesoamerican civilization’, Current Anthropology, 37: 1–14. Boaretto, E., Wu, X.,Yuan, J., Bar-Yosef, O., Chu,V., Pan,Y., Liu, K., Cohen, D., Jiao, T., Li, S., Gu, H., Goldberg, P. and Weiner, S. (2009) ‘Radiocarbon dating of charcoal and bone collagen associated with early pottery at Yuchanyan cave, Hunan Province, China‘, Proceedings of the National Academy of Sciences, 106: 9595–9600. Chang, K.C. (1986) The Archaeology of Ancient China, New Haven:Yale University Press. Chang, K.C. (1989) ‘An essay on cong’, Orientations, 20: 37–43. Chen, C. (2007) ‘Techno-typological comparison of microblade cores from East Asia and North America’, in V.Y. Kuzmin, S.G. Keates and C. Shen (eds) Origin and Spread of Microblade Technology in Northern Asia and North America, Burnaby: Archaeology Press. Chen,Y. (2009) ‘Yige cong “li” tansuo Zhongguo wenming qiyuan de moshi (One model of exploring the origins of Chinese civilization from the concept of “li”)’, Wenwu, 2: 90–93. Cheng, P. (2005) ‘Lun Taosi gucheng de fazhan jieduan xingzhi (Discussion of the development stages and characteristics of Taosi city)’, Jianghan Kaogu, 3: 48–53. Chifeng International Collaborative Archaeological Project. (ed) (2011) Settlement Patterns in the Chifeng Region, Pittsburgh: Center for Comparative Archaeology. Cohen, D.J. (2011) ‘The beginnings of agriculture in China: a multiregional view’, Current Anthropology, 52: S273–S293. D’alpoim-Guedes, J., Jiang, M., He, K., Wu, X. and Jiang, Z. (2013) ‘Site of Baodun yields earliest evidence for the spread of rice and foxtail millet agriculture to south-west China’, Antiquity, 87: 758–771. Dematte, P. (1999) ‘The role of writing in the process of state formation in late Neolithic China’, East and West 49 (1–4): 241–272.

35

Gideon Shelach-Lavi Fang,Y. and Liu, X. (2006) ‘Henan Dengfengshi Wangchenggang yizhi 2002, 2004 nian fajue jianbao (Preliminary report on the 2002, 2004 excavations at the Wangchenggang site, Dengfeng City)’, Kaogu, 9: 3–9. Flad, R.K. (2008) ‘Divination and power: a multiregional view of the development of oracle bone divination in early China’, Current Anthropology, 49: 403–437. Flad, R.K. and Chen, P. (2013) Ancient Central Asia: Centers and peripheries along the Yangzi River, Cambridge: Cambridge University Press. Flad, R., Yuan, J. and Li, S. (2009) ‘Zhongguo xibei diqu zuizao dongwu xunhua de dongwu kaoguxue ziliao (Zooarchaeological evidence for the earliest animal domestication in northwest China)’, Kaogu, 5: 80–86. Fuller, D.Q., Denham, T., Arroyo Kalin, M., Lucas, L., Stevens, C.J., Qin, L., Allaby, R.G. and Purugganan, M.D. (2014) ‘Convergent evolution and parallelism in plant domestication revealed by an expanding archaeological record’, Proceedings of the National Academy of Sciences, 111: 6147–6152. Fuller, D.Q., Harvey, E. and Qin, L. (2007) ‘Presumed domestication? Evidence for wild rice cultivation and domestication in the fifth millennium bc of the lower Yangtze region’, Antiquity, 81: 316–331. Gross, B.L. and Zhao, Z. (2014) ‘Archaeological and genetic insights into the origins of domesticated rice’, Proceedings of the National Academy of Sciences, 111: 6190–6197. Guo, R. and Li, J. (2002) ‘The Nanzhuangtou and Hutouliang sites: exploring the beginnings of agriculture and pottery in north China’, in Y. Yasuda (ed) The Origins of Pottery and Agriculture, New Delhi: Roli Books. Hao, S., Xue, J. and Cui, H. (2008) ‘Donghulin sihao ren muzang zhong de guohe (Fruit seeds found at grave no. 4, Donghulin)’, Renxue Xuebao, 3: 249–255. He, N. (2013) ‘The Longshan period site of Taosi in southern Shanxi province’, in A.P. Underhill, (ed) A Companion to Chinese Archaeology, Malden: Wiley-Blackwell. Hebei Sheng Wenwu Yanjiusuo. (2010) ‘1997 nian Hebei Xushui Nanzhuangtou yizhi fajue baogao (Report on the 1997 excavations at the Nanzhuangtou site, Xushui County, Hebei Province)’, Kaogu Xuabao, 3: 361–392. Huang,T.M. (1992) ‘Liangzhu: a late Neolithic jade-yielding culture in southwestern coastal China’, Antiquity, 66: 75–83. Hubei Sheng Wenwu Kaogu Yanjiusuo. (2003) Dengjiawan:Tianmen Shijiahe kaogu baogao zhi er (Dengjiawan: Archaeological work at Shijiahe,Tianmen II). Beijing: Wenwu Chubanshe. Hunan Sheng Wenwu Kaogu Yanjiusuo. (2006) Pengtoushan yu Bashidang (The Pengtoushan and Bashidang Sites), Beijing: Kexue Chubanshe. Hunan Sheng Wenwu Kaogu Yanjiusuo. (2007) Lixian Chengtoushan: Xinshiqi shidai yizhi fajue baogao (Chengtoushan, Lixian: A Report on the Neolithic Site Excavations), Beijing: Wenwu Chubanshe. Jiang, L. and Liu, L. (2006) ‘New evidence for the origins of sedentism and rice domestication in the Lower Yangzi River, China’, Antiquity, 80: 355–361. Keightley, D. (2006) ‘Marks and labels: early writing in Neolithic and Shang China’, in M.T. Stark. (ed) Archaeology of Asia, Malden: Wiley-Blackwell, 177–201. Li, X., Harbottle, G., Zhang, H. and Wang, C. (2003) ‘The earliest writing? Sign use in the seventh millennium bc at Jiahu, Henan Province, China’, Antiquity, 77: 31–43. Liaoning Sheng Wenwu Kaogu Yanjiusuo. (2012) Chahai: Xinshiqi shidai juluo yizhi fajue baogao (Chahai: Report on the excavation of a Neolithic Site), Beijing: Wenwu Chubanshe. Liu, J. (2006) Hemudu wenhua (Hemudu culture), Beijing: Wenwu Chubanshe. Liu, L. (2000) ‘Ancestor worship: an archaeological investigation of ritual activities in Neolithic north China’, Journal of East Asian Archaeology, 2 (1–2): 129–164. Liu, L. (2004) The Chinese Neolithic:Trajectories to early states, Cambridge & New York: Cambridge University Press. Liu, L. (2015) ‘A long process towards agriculture in the middle Yellow River valley, China: evidence from macro- and micro-botanical remains’, Journal of Indo-Pacific Archaeology, 35: 3–14. Liu, L., Bestel, S., Shi, J., Song,Y. and Chen, X. (2013) ‘Paleolithic human exploitation of plant foods during the last glacial maximum in North China’, Proceedings of the National Academy of Sciences, 110: 5380–5385. Liu, L. and Chen, X. (2012) The Archaeology of China: From the Late Paleolithic to the Early Bronze Age, Cambridge: Cambridge University Press. Liu, L., Field, J., Fullagar, R., Bestel, S., Chen, X. and Ma, X. (2010) ‘What did grinding stones grind? new light on early Neolithic subsistence economy in the middle Yellow River valley, China’, Antiquity, 84: 816–833.

36

Main issues in the Chinese Neolithic Liu, X., Fuller, D.Q. and Jones, M. (2015) ‘Early agriculture in China’, in G. Barker and C. Goucher (eds) Cambridge World History; Vol. 2: A World with Agriculture, 12,000 bce – 500 ce, Cambridge: Cambridge University Press, 310–334. Ma, S. (2008) ‘Dengfeng Wangchenggang chengyi you Yu du yincheng (On the ancient city site of Wangchenggang and the Yang city founded by Yu)’, Zhongyuan Wenwu, 2: 22–26. Ma, X. (2005) Emergent Social Complexity in the Yangshao Culture: Analyses of Settlement Patterns and Faunal Remains from Lingbao,Western Henan, China (c. 4900–3000 bc), Oxford: Archaeopress. Nanjing, bowuguan. (1984). “1982 nian Jiangsu Changzhou Wujin Sidun yizhi de fajue (1982 Excavation at the Sidun site, of Changzhou, Wujin, Jiangsu province).” Kaogu 1984 (2): 109–129. Pechenkina, E.A., Benfer, R.A. and Wang, Z.J. (2002) ‘Diet and health changes at the end of the Chinese Neolithic: the Yangshao/Longshan transition in Shaanxi Province’, American Journal of Physical Anthropology, 117: 15–36. Pei, A. (2004) ‘Liyang pingyuan shiqian juluo xingtai de yanjiu yu sikao (Analysis of prehistoric settlement patterns on the Liyang Plain)’, in Jilin Daxue Bianjiang Kaogu Yanjiu Zhongxin (ed.) Qingzhu Zhang Zhongpei Xiansheng Qishisui Lunwenji, Beijing: Kexue Chubanshe. Peng, K. and Zhu,Y. (1995) ‘New research on the origins of cowries used in ancient China’, Sino-Platonic Papers, 68: 1–21. Peterson, C. and Drennan, R.D. (2011) ‘Methods for delineating community patterns’, in Chifeng International Collaborative Archaeological Project (ed) Settlement Patterns in the Chifeng Region, Pittsburgh: Center for Comparative Archaeology, University of Pittsburgh, 80–87. Peterson, C. and Shelach, G. (2010) ‘The evolution of Yangshao period village organization in the middle reaches of northern China’s Yellow River valley’, in M. Bandy and J.R. Fox (eds) Becoming Villagers, Tucson: University of Arizona Press, 246–275. Peterson, C. and Shelach, G. (2012) ‘Jiangzhai: Social and economic organization of a Middle Neolithic Chinese village’, Journal of Anthropological Archaeology, 31: 265–301. Prendergast, M.E.,Yuan, J. and Bar-Yosef, O. (2009) ‘Resource intensification in the Late Upper Paleolithic: a view from southern China’, Journal of Archaeological Science, 36: 1027–1037. Puyang Xishuipo Yizhi Kaogudui. (1989) ’1988 nian Henan Puyang Xishuipo yizhi fajue jianbao (Preliminary report on the 1988 excavations of the Xishuipo Site, Puyang, Henan)’, Kaogu, 12: 57–66. Qu, T., Bar-Yosef, O., Wang,Y. and Wu, X. (2013) ‘The Chinese Upper Paleolithic: geography, chronology, and techno-typology’, Journal of Archaeological Research, 21: 1–73. Shaanxi Sheng Kaogu Yanjiuyuan. (2013) ‘Shǎanxī Shénmù xiàn Shímǎo yízhǐ (The Shimao site, Shenmu County, Shaanxi)’, Kaogu, 7: 15–24. Shelach, G. (2000) ‘The earliest Neolithic cultures of Northeast China: recent discoveries and new perspectives on the beginning of agriculture’, Journal of World Prehistory 14: 363–413. Shelach, G. (2006) ‘Economic adaptation, community structure, and sharing strategies of households at early sedentary sites in Northeast China’, Journal of Anthropological Archaeology, 25: 318–345. Shelach-Lavi, G. (2015) The Archaeology of Early China: From Prehistory to the Han Dynasty, Cambridge: Cambridge University Press. Shelach-Lavi, G. and Jaffe,Y. (2014) ‘The earliest states in China: a long-term trajectory approach’, Journal of Archaeological Research, 22 (4): 327–364. Shelach-Lavi, G. and Teng, M. (2013) ‘Earlier Neolithic economic and social systems of the Liao River region, Northeast China’, in A. Underhill (ed) A Companion to Chinese Archaeology, Oxford: Blackwell Publishers. Shizitan, Kaogudui. (2010) ‘Shanxi Ji xian Shizitan yizhi di jiu dian faxian jianbao (Preliminary report of the excavations at locality 9 of the Shizitan site, Ji County, Shanxi)’, Kaogu, 10: 7–17. Smith, B.D. (2001) ‘Low-level food production’, Journal of Archaeological Research, 9: 1–43. Smith, B.L. (2005) Diet, Health, and Lifestyle in Neolithic North China, Cambridge: Harvard University Press. Stanley, D.J., Chen, Z.Y and Song, J. (1999) ‘Inundation, sea-level rise and transition from Neolithic to Bronze Age cultures,Yangtze Delta, China’, Geoarchaeology, 14: 15–26. Sun, B. and Wagner, M. (2014) ‘Archaeological discovery and research at Bianbiandong Early Neolithic cave site, Shandong, China’, Quaternary International, 348: 169–182. Underhill, A.P. (2002) Craft Production and Social Change in Northern China, New York: Kluwer Academic & Plenum Publishers. Underhill, A.P., Feinman, G.M., Nicholas, L.M., Fang, H., Luan, F.S.,Yu, H.G. and Cai, F.S. (2008) ‘Changes in regional settlement patterns and the development of complex societies in southeastern Shandong, China’, Journal of Anthropological Archaeology, 27: 1–29.

37

Gideon Shelach-Lavi Wang, H. (2003a) ‘Cong Menbanwan chenghao juluo kan Changjiang chongyou diqu chenghaoquluo de qiyuan yu gongyong (The development and function of moats in the middle Yangzi River area)’, Kaogu, 9: 61–75. Wang,Y. (2003b) ‘Prehistoric walled settlements in the Chengdu Plain’, Journal of East Asian Archaeology, 5: 109–148. Wang, Y. (2005) Zhongguo yuangu renlen wenhua de yuanliu (The roots of Pleistocene hominids and cultures in China), Beijing: Kexue Chubanshe. Wu, X., Zhang, C., Goldberg, P., Cohen, D., Pan, Y., Arpin, T. and Bar-Yosef, O. (2012) ‘Early pottery at 20,000 years ago in Xianrendong cave, China’, Science, 336: 1696–1700. Yang, X., Ma, Z., Wang, T., Perry, L., Li, Q., Huan, X. and Yu, J. (2014) ‘Starch grain evidence reveals early pottery function cooking plant foods in North China’, Chinese Science Bulletin, 59: 4352–4358. Yao, A. (2010) ‘Recent developments in the archaeology of southwestern China‘, Journal of Archaeological Research, 18: 203–239. Yu, S., Zhu, C., Song, J. and Qu, W. (2000) ‘Role of climate in the rise and fall of Neolithic cultures on the Yangtze Delta’, Boreas, 29: 157–165. Yuan, J. (2007) ‘Dongwu kaoguxue jiemi gudai renlei he dongwu de xianghu guanxi (Zooarchaeology reveals the mutual relations between ancient humans and animals)’, Xibu Kaogu, 2: 82–95. Yuan, J. and Flad, R.K. (2002) ‘Pig domestication in ancient China’, Antiquity, 76: 724–732. Zeder, M.A. (2015) ‘Core questions in domestication research’, Proceedings of the National Academy of Sciences, 112: 3191–3198. Zhang, C. and Hung, H.C. (2008) ‘The Neolithic of Southern China: origin, development, and dispersal’, Asian Perspectives: Journal of Archeology for Asia & the Pacific, 47: 299–329. Zhang, J.F., Wang, X.Q., Qiu, W.L., Shelach, G., Hu, G., Fu, X., Zhuang, M.G. and Zhou, L.P. (2011) ‘The Paleolithic site of Longwangchan in the middle Yellow River, China: chronology, paleoenvironment and implications’, Journal of Archaeological Science, 38: 1537–1550. Zhang, X. (2006) Longshan Wenhua (The Longshan culture), Beijing: Wenwu Chubanshe. Zhao, C. (2009) ‘Xinzhai juluo kaogu de shijian yu fangfa (Archaeological methods and practice in research on the settlement of Xinzhai)’, Kaogu, 2: 48–54. Zhao, Z. (2011) ‘New archaeobotanic data for the study of the origins of agriculture in China’, Current Anthropology, 52: S295–S306. Zheng, Y.F., Sun, G.P., Qin, L., Li, C.H., Wu, X.H. and Chen, X.G.(2009) ‘Rice fields and modes of rice cultivation between 5000 and 2500 bc in east China’, Journal of Archaeological Science, 36: 2609–2616. Zhejiang Sheng Wenwu Kaogu Yanjiusuo. (2004) Kuahuqiao, Beijing: Wenwu Chubanshe. Zhejiang Sheng Wenwu Kaogu Yanjiusuo. (2005) Fanshan, Beijing: Wenwu Chubanshe. Zhongguo Shehui Kexueyuan Kaogu Yanjiusuo. (2003) Guilin Zengpiyan (The Zengpiyan site, Guilin), Beijing: Wenwu Chubanshe. Zhongguo Shehui, Kexueyuan Kaogu Yanjiusuo. (2010) Zhongguo Kaoguxue: Xinshiqi shidai Juan (The Archaeology of China:The Neolithic volume), Beijing: Zhongguo shehui kexue chubanshe. Zhongguo Shehui Kexueyuan Kaogu Yanjiusuo Shanxidui. (2005) ‘Shanxi Xiangfen Taosi chengzhi 2002 nian fajue baogao (Report on the 2002 year excavations at the Taosi site, Xiangfen, Shanxi)’, Kaogu Xuabao, 3: 307–346. Zhongguo Shihui Kexueyuan Kaogusuo Yanjiu Guangxi Gongzuodui. (1998) ‘Guangxi Yongning xian Dingsishan yizhi de fajue (Excavations at the Dingsishan site, Yongning County, Guangxi)’, Kaogu, 11: 11–33.

38

2 OF MILLETS AND WHEAT

KATE PECHENKINAOF MILLETS AND WHEAT

Diet and health on the Central Plain of China during the Neolithic and Bronze Age Kate Pechenkina

This chapter takes a bioarchaeological approach in examining how human diet and health changed on the Central Plain during the time preceding the unification of China by the Qin dynasty (秦朝) in 221 bc. Using the analysis of human skeletons as our starting point, we aim to reconstruct how people’s lives changed from the time of the Neolithic Yangshao to that of the Eastern Zhou dynasty, which witnessed agricultural intensification, introduction of new domesticated plants, the development of animal husbandry, population growth, and the rise of social inequality. The Central Plain (中原) of China is formed by deposits of the Yellow River and spans the fertile agricultural lands of Henan, southern Hebei, southern Shanxi, and the western portion of Shandong. In a broader sense, the Central Plain is frequently amplified to include the adjacent territories of the Guanzhong Plain surrounding the Wei River valley, the northwestern part of Jiangsu, and parts of Anhui and northern Hubei, as human communities of this entire territory shared many aspects of their sociocultural development in the past.The northern center of Chinese agriculture developed in the region of the Central Plain during the Neolithic, independent of the southern agricultural center of Yangzhe, where rice was the main staple cereal (Zhao 2011; Liu et al. 2012). On the Central Plain, broomcorn millet (Panicum miliaceum) and foxtail millet (Setaria italica), cereals characterized by superior resistance to drought and cold, were the two principal crop plants. Between 5000 and ca. 3000 cal. bc this region was occupied by settlements of the Yangshao (仰韶) archaeological tradition, best known for its black on red painted pottery produced by coiling and painted with zoomorphic, anthropomorphic, and, later, geometric motifs (Yan 1992; Ren and Wu 1999; The Institute of Archaeology Chinese Academy of Sciences 2010). The Yangshao was succeeded by the Longshan (龙山) cultural phase and then the early Chinese dynasties. Following the conquest and consolidation of territories by the Qin dynasty, led by Qin Shi Huang (秦始皇), the Central Plain became the core of an expanding Chinese empire (Li 2013).

Development of agriculture and animal husbandry on the Central Plain The trajectory of early agricultural development on the Central Plain is unusual, as the repertoire of cultivated plants changed considerably during the Bronze Age. The earliest evidence of 39

Kate Pechenkina

domesticated millet in this area dates back to more than ~11,000 cal bp (Lu et al. 2009, Yang et al. 2012). Millet agriculture flourished during the Yangshao, when the two species of millet became predominant in the paleobotanical record of the region. A flotation study conducted at the Yangshao tradition Yuhuazhai (鱼化寨) site (Figure 2.1), located in Shaanxi Province, found that millets accounted for almost 90% of the total number of recovered seeds (Zhao 2011), underscoring their economic importance. The ubiquity of the two species of millet was similar: 67% for Panicum miliaceum and 63% for Setaria italica. Other economically important plants recognized at this site from paleobotanical remains included Brassica (cabbage), Papaver (poppy), Vitis (grapevines), and Oryza (rice), the latter represented by only four charred grains (Zhao 2011). Similarly, analysis of plant remains at the nearby Yangshao site of Didong (底董) revealed that the two species of millet accounted for 81% of all plant remains (Wei 2014). Remains of acorns and wild legumes, as well as some plants that couldn’t be identified, accounted for the rest. Stable isotope studies of human and animal bones from Yangshao sites confirm that millets served as the major calorie sources for humans as well as for their domesticated animals, i.e. pigs and dogs, during the Yangshao (Pechenkina et al. 2005; Guo et al. 2011; Zhang et al. 2011). Recovery of millet-based noodles from the Late Neolithic Lajia (喇家) archaeological site, dating to around 4,000 bp (Lu et al. 2005), indicates that millets served as a foundation of indigenous cuisine through the end of the Neolithic. Evidence of both millet- and rice-based fermented beverages dating back to the Neolithic was identified by McGovern and colleagues based on pottery and bronze vessel residue analyses (Mcgovern et al. 2004). Largely provisioned with millet, pigs served as essential domesticated animals during the Neolithic. In the earlier phases of the Yangshao, hunting still made a substantial contribution to the human diet, as suggested by zooarchaeological assemblages. At the Early Yangshao site of Jiangzhai (姜寨), with radiocarbon dates ranging from 6790 to 5360 cal bp (Institute of

Figure 2.1 Location of archaeological sites discussed in the chapter: 1 – Shijia, 2 – Banpo and Yuhuajai, 3 – Jiangzhai, 4 – Baijia, 5 – Xipo, 6 – Guanjia, 7 – Xishan, 8 – Gouwan, 9 – Ancient Zhenghan city associated sites: Xiyasi, Xinghong, Chanxinyuan, Thermal Power Plant, 10 – Jiahu, 11 – Fujia, 12 – Dongjiaying, 13 – Liangwangchen, 14 – Huating

40

Of millets and wheat

Archaeology 1991), bones of domestic pigs comprised only 22% (674 out of 3096 identifiable animal bones) of the total faunal assemblage by count (Xi’an Banpo Museum 1988). Different species of deer, including sika deer (Cervus nippon), water deer (Hydropotes inermis), musk deer (Moschus moschiferus), and unidentified deer comprised 60% of the faunal assemblage. Other faunal remains identified from that site included the bones of macaque (Macaca mulatta), mole rats (Myospalax fontianieri), dhole (Cuon alpinus), raccoon dog (Nyctereutes procyonoides), wild cattle (Bos sp.), Mongolian gazelle (Prodorcas gutturosa), and indeterminate Artiodactyla, as well as single bones of other wild animals. Pig bones dominate the faunal assemblages of Middle and Late Yangshao sites. For instance, at Xipo (西坡) pig bones comprised 84% of the faunal assemblage by number, while dog bones were 1.3%. Wild animal bone altogether comprised 14.7% of the Xipo assemblage and included the remains of sika deer (Cervus nippon), water deer (Hydropotes inermis), and musk deer (Moschus moschiferus), along with a few fragments of mollusks, fish, and birds (Ma 2003: 129–148). Similarly, more than 80% of animal bones recovered from Xishan (西 山), a Miaodigou (庙底沟) phase Yangshao site, belonged to pigs. The remains of twelve species of wild mammals were also identified at Xishan, including those of sika deer (Cervus nippon), tufted deer (Elaphurus cephalophus), Père David’s deer (Elaphurus davidianus), water deer (Hydropotes inermis), leopard (Felis pardus), panther (Panther tigris), fox (Vulpes sp.), raccoon (Nyctereutes proycyonoides), and a number of small mammals (Chen 2006). Domesticated chickens were present in the area as early as 10,000 bp (Xiang et al. 2014), although assessment of their ubiquity is complicated by the difficulty of species diagnostics (Yuan 2001). Development of millet agriculture in China overlapped in time with the Holocene Megathermal, which manifested in this territory as higher mean annual temperatures and greater precipitation between ca. 7900 and 4450 cal yr bp (Xiao et al. 2004; Peng et al. 2005).This favorable episode was followed by the rapid onset of a cool and dry climatic episode around ca. 4450 to 3950 cal yr bp and a generally cooler and drier climate between ca. 4450–2900 cal yr bp, which likely resulted in lower agricultural outputs.The period of climatic instability and colder climate coincided in time with the introduction of new cereals and the expansion of domesticated herbivores into the area. Wheat (Triticum aestivum) and barley (Hordeum vulgare) first appeared in paleobotanical assemblages from northern China ca. 4600–3900 years ago, during the Late Neolithic (Crawford et al. 2005; Zhao and He 2006; Li et al. 2007; Flad et al. 2010; Dodson et al. 2013). Domesticated soybeans (Glycine max) also first appeared in this area during the Late Neolithic and became progressively more important over time (Fuller et al. 2014). Initially and through much of the Bronze Age, the proportion of these new grains remained miniscule, as compared to those of millets, but they seem to have gained in importance after 2500 bp (Fuller and Zhang 2007; Lee and Bestel 2007; Lee et al. 2007; Lan and Chen 2014). Domesticated herbivores were introduced into the region during the Late Neolithic and early dynasties. Domesticated cattle (Bos taurus) appeared on the Central Plain between 4500 and 4000 bp (Lv 2010). Sheep and goats also first appeared in the region during the Late Neolithic and became widespread during the Erlitou (二里头 3900–3500 bc) (Li et al. 2014). Sheep were predominantly raised for wool, as inferred from the advanced age of sheep in several faunal assemblages from this region (Dai et al. 2014), although it is suggested that at Yinxu (殷墟), an archaeological site in Henan, the sheep slaughter pattern was consistent with their utilization for meat (Li et al. 2014). The presence of domesticated horses in the region dating to around 3300 bp has also been documented at Yinxu (Yuan and An 1997). Domesticated water buffalo (Bubalus bubalus), used as draft animals, were introduced to China from South Asia as late as 3000 bp (Liu et al. 2006). How, why, and when wheat supplanted millets, which had servedas the core of indigenous cuisine on the Central Plain of China through the Neolithic, remains enigmatic. Strictly 41

Kate Pechenkina

pragmatically, wheat is more productive than millets per cultivated area of land. Although millets are a better source of dietary iron than wheat, they are extremely deficient in protein (FAO 1972). Yet millets had long been established as the foundation of an elaborate cuisine, while wheat and barley don’t seem to be appreciated initially. Historic records dating to the Han dynasty consistently refer to wheat, as well as beans, as a poor person’s diet, fall back or starvation food, plentiful, yet not very palatable (Yu 1977). Millets were priced at 2.5 times more than wheat or beans during the great famine of ad 194 (Yu 1977). Beans were likely prepared as congee (gruel), but it is not clear how wheat and barley were initially cooked. No technology yet existed for converting these much tougher cereals into thin flour for noodle production. Bray (1984) proposed that they could have been roasted and then pounded down and mixed with oils in the manner of Tibetan tsamba (Bray 1984: 462). Only at the end of the Han dynasty, when the development of hand-mills, as well as large water and animal powered mills, allowed converting wheat into fine flour for noodle production, were these new cereals finally appreciated in the area (Bray 1984: 461).

Stable isotope perspective on Yangshao diets Stable isotope method The two stable isotopes of carbon, 12C and 13C, accumulate at different rates in plants using the C3 and C4 pathways of photosynthesis (Farquhar et al. 1984; O’Leary 1988).The overwhelming majority of plants, close to 95% of known species, follow the C3 pathway, which can be utilized only when plant stomata are open. In hot and arid climates, some plants developed a waterconserving C4 pathway of photosynthesis that is more efficient than C3 and can proceed even when stomata are closed allowing plants to conserve moisture during the hot periods of the day. This latter pathway discriminates against heavier isotopes of carbon, including proportionally fewer atoms of 13C into the sugars produced. The consequent δ13C isotopic signatures of C4 plants are less negative than those of the C3 plants. The two types of plants can be recognized based on leaf anatomy, as C4 plants display a Kranz leaf structure, with bundle sheath cells possessing chloroplasts. The two millet species were the only C4 domesticates grown in early China. Setaria viridis, the wild progenitor of Setaria italica, also follows the C4 pathway of photosynthesis. It is a weedy plant that grows on the Central Plain and beyond. Its consumption would certainly shift isotopic signatures in human and animal bones towards less negative δ13C. However, it seems unlikely that humans would consume large quantities of this particular species, as Setaria viridis is always found in patches interspersed with other plants in the wild. Besides, isotopic signatures obtained from the bones of wild animals suggest that the overwhelming majority of wild grasses growing in the vicinity of Neolithic settlements in the area were C3. The C4 pathway results in δ13C isotopic values averaging around −12.5‰, as opposed to the −26.5‰ typical for C3 plants (van Der Merwe 1982; Schoeninger and Moore 1992; Ambrose 1993). Analysis of modern Setaria italica millet grain produced δ13C values averaging −11.81‰ (Pechenkina et al. 2005). Thus, consumption of millets can be detected by the stable isotope analysis of human and animal bone collagen, which preserves the isotopic composition of the diet plus an approximately 5‰ trophic level enrichment. Nitrogen isotopic composition (δ15N) for terrestrial food-webs is a reliable indicator of animal product consumption (Ambrose 1991), as δ15N undergoes a trophic level enrichment of 2–6‰ with every step of the food chain (Minagawa and Wada 1984; Schoeninger and Deniro 1984; Bocherens and Drucker 2003). In human communities where terrestrial diets are inferred 42

Of millets and wheat

from the geographic location and zooarchaeological evidence, higher δ15N values generally reflect greater proportions of animal products in the diet. Aquatic food-webs, where nitrogen is repeatedly recycled, display very high δ15N signatures. Therefore, δ15N in human bone collagen provides an estimate of the amount of animal protein in the diet as well as the degree of dependence on aquatic animal resources (Ambrose 1991). Heavy reliance on legumes (e.g. beans) can considerably depress bone δ15N values, because legumes maintain colonies of nitrogen-fixing bacteria in their roots and hence have δ15N values close to 0‰ (Deniro and Epstein 1981).

Brief summary of the archaeological background for Yangshao sites in the analysis Shaanxi sites Jiangzhai (姜寨) is an Early Yangshao archaeological site located in Lintong County of Shaanxi Province, south of the Wei River, on the northern bank of the Lin River (Xi’an Banpo Museum 1988) (Figure 2.1). Five radiocarbon dates obtained from charcoal, as well as animal and human bones, range from 4790 to 3360 cal bc (Institute of Archaeology 1991: 262). Jiangzhai had a settled area of approximately 5 hectares, demarcated by a substantial ditch (Xi’an Banpo Museum 1988). During phase I, the living community consisted of about 120 houses. These surrounded a central plaza encompassing circular structures interpreted as animal pens. Five of the largest houses, with a floor area of up to 124 m2 each, were initially interpreted as gathering centers for large clans. Each of these houses was associated with a number of medium-size houses of 20–40 m2 interpreted by the archaeologists as representing the activity areas of smaller matrilineal clans, each in turn with an aggregate of small houses of its own, each with living space sufficient for a small family of 2–5 people, or possibly a pair of families (Xi’an Banpo Museum 1988: 352–357). Based on the layout of the Jiangzhai site, Shelach (Shelach 2006) argued that its people were communally sharing oriented because its houses were arranged facing inward in circles, while storage pits are all located outside of the houses. Similarly, Lee (Lee 2007) concluded that the Jiangzhai community relied on communal food sharing. Among 376 excavated burials, 174 were single burials, whereas the rest contained multiple individuals. Burials contained grave goods typical for the Neolithic, including a variety of pottery, stone implements, and arrowheads, as well as jewelry made of stone, bone, and shell (Xi’an Banpo Museum 1988). The Shijia (史家) site from Weinan County in eastern Shaanxi is the type site for the Shijia phase of Early Yangshao culture in Shaanxi Province, which in this region succeeded the Banpo phase and was followed by Miaodigou.The site is situated on the western bank of the Qiu River, 12 kilometers south of the Wei River (Museum 1978; Wang 1993). A single radiocarbon date obtained for a human bone from Shijia is calibrated to 3779–3526 bc (Institute of Archaeology 1991: 264). Radiocarbon dates associated with the Shijia phase fall between 6140 and 5935 calibrated radiocarbon years ago, or 4190–3985 bc (Wang 1993). The overwhelming majority, 40 out of 43 burials at Shijia, represent multiple secondary interments with individual bodies that apparently partially decomposed before being packed in tight bundles. Multiple bundles were arranged in rows next to one another, with each grave containing multiple rows of funerary bundles (Gao and Lee 1993). Based on a craniometric analysis of Shijia individuals, Gao and Lee (Gao and Lee 1993) proposed that Shijia society was organized in patrilineal clans. Yuhuazhai (鱼化寨) is a 40,000 m2 Yangshao village site located in the western suburbs of Xi’an on a mound located 300 m from Yuhuazhai Village (Zhang and Guo 2003). The site is in close proximity to Banpo (半坡) (Institute of Archaeology 1963), surrounded by moats, 43

Kate Pechenkina

and is strikingly similar to Banpo in its layout, features, and artefacts. The site dates to between 7000–6000 cal bp (Zhao 2011). Therefore, paleobotanical remains from Yuhuazhai can be used as proxies for the Banpo assemblage (Zhao 2011). Many of the human bone samples analyzed come from urn burials of juvenile individuals (Zhang et al. 2011).

Henan sites Xipo (西坡) is situated in the upper Sha River valley, about 3 kilometers north of the Qinling Mountains, in the western corner of Henan Province (Ma 2003; Ma et al. 2005, 2006). Radiocarbon dates obtained from the human remains place Xipo burials toward the end of the Yangshao (5300–4900 bp). This was a relatively large settlement (40 hectares), surrounded by a moat, and with an estimated population of 640–900 people (Ma 2003: 85). A three-level size hierarchy among the houses, as well as the presence of several larger labor-intensive structures, suggests incipient social inequality. The biggest structure may have served as a gathering place for ritual or public functions (Ma 2003: 100). A cemetery outside the moat yielded 22 burials, all single interments. Associated grave goods ranged widely in number, suggesting a degree of status heterogeneity. Guanjia (关家), the other Middle Yangshao site, was a smaller settlement (9 hectares). It is located along a narrow terrace on the south bank of the Yellow River, about 70 kilometers northeast of Mianchi (Fan 2000). The major occupation at Guanjia was contemporaneous with that at Xipo, but Peiligang and Early Dynastic remains were also present. Ditches demarcate the western and southern limits of the site, where it is not bounded by the Yellow River. Only seven of the 52 burials unearthed had associated artifacts, limited to small personal adornments, such as beads and hairpins. Pottery fragments were recovered from one burial (Fan 2000). Xishan (西山) is a large Neolithic site in the northwestern suburbs of Zhengzhou City (Liu 1986). The site was initially excavated in 1984, and later for four consecutive years between 1993 and 1996. The total area of the site is 34,000 m2. The site has three distinct layers of occupation.Yangshao culture layers were dated to from 6500 to 5000 bp. Its major phase of occupation and the majority of burials correspond to the Miaodigou phase of Yangshao culture (ca. 6500–5800 bp). Gouwan (沟湾) is a Neolithic site located in southwestern Henan, Xichuan (淅川) county, in the Zhangying (张营) administrative district, approximately 800 m west of the Fengzi Mountain (Fu et al. 2010). Three layers of Yangshao occupation are recognized at the site, including Early, Yangshao 1 (7000–6600 bp); Middle, Yangshao 2 (6600–6000 bp); and Late, Yangshao 3 (6000–5500 bp).Yangshao occupation was succeeded at the site by the Qujialing (屈家岭) Culture (5000–4600 bp). Its location in the southern part of Henan suggests that rice cultivation, along with millet agriculture, could have been practiced at the site.

Dietary variation during Yangshao A considerable quantity of stable isotope data from human and animal bone samples obtained from Neolithic archaeological sites of the Central Plain and adjacent territories has been accumulated in recent years (Cai and Qiu 1984; Pechenkina et al. 2005; Hu et al. 2006, 2008; Barton et al. 2009; Guo et al. 2011; Zhang et al. 2011). These data indicate that domesticated millets became the principal caloric source for humans of the Central Plain during the Yangshao (Table 2.1). Based on published data and our research, Figure 2.2 shows the distribution of the stable isotope values from the Yangshao collagen bone samples. Several aspects of isotopic variation in Yangshao bone samples are noteworthy: 44

Of millets and wheat Table 2.1  Stable isotope values in human bone collagen samples from the Yangshao archaeological sites site1

sex

δ 13C ‰ mean

δ 15N ‰ SD

median

Yangshao (7000 bp to 4800 bp): Banpo Total −14.84 1.93 −14.2 Jiangzhai F −9.8 0.9 −9.9 Jiangzhai M −10.1 1.5 −9.9 Jiangzhai U −9.7 NA −9.7 Jiangzhai Total −9.91 1.12 −9.8 t-test (femalest = 0.63, p = 0.54 males) Shijia F −10.2 1.0 −10.2 Shijia M −10.0 0.6 −10.2 Shijia U −10.0 NA −10.0 Shijia Total −10.04 0.69 −10.15 t-test (femalest = −0.35, p = 0.74 males) Guanjia F −8.3 0.5 −8.3 Guanjia M −7.8 0.6 −7.6 Guanjia U −8.0 0.1 −8.0 Guanjia Total −8 0.58 −7.96 t-test (femalest =−1.79, p = 0.09 males) Xipo F −9.6 1.2 −9.5 Xipo M −9.5 1.1 −9.7 Xipo Total −9.55 1.11 −9.62 t-test (femalest =−0.18, p = 0.85 males) Xishan F −8.2 1.2 −7.9 Xishan M −8.3 1.6 −7.6 Xishan U −8.2 1.6 −7.8 Xishan Total −8.21 1.47 −7.82 t-test (femalest = 0.12, p = 0.90 males) Yuhuazhai Total −8.43 1.31 −8.14 Gouwan Total −14.30 1.92 −14.2

N

MAD2

mean

1.19 1.3 1.5 0.0 1.33

9.05 NA 9.05 8.4 0.3 8.4 9.0 0.6 9.1 8.6 NA 8.6 8.63 0.55 8.5 t = −2.79, p = 0.02

0.0 0.3 0.5 0.0 0.58

5 12 8 1 21

1.4 0.4 0.0 0.53

8.2 0.3 8.2 8.0 0.6 8.1 7.7 NA 7.7 8.07 0.46 8.08 t = 0.55, p = 0.59

0.4 0.5 0.0 0.52

3 5 1 9

0.7 0.5 0.1 0.61

6.1 0.8 6.2 6.3 0.6 6.2 6.5 0.2 6.5 6.21 0.64 6.23 t = −0.56, p = 0.58

1.0 0.3 0.2 0.55

8 13 2 23

0.9 1.0 1.02

9.5 0.7 9.3 9.3 1.2 9.3 9.35 1.04 9.26 t = 0.67, p = 0.50

0.4 0.8 0.63

12 28 40

0.1 0.6 1.0 0.53

8.8 0.6 8.6 9.0 0.5 8.9 9.3 1.3 9.0 9.01 0.81 8.8 t = −1.06, p = 0.30

0.3 0.5 1.4 0.65

10 18 11 39

0.88 1.63

9.37 8.34

0.74 1.33

22 39

SD

0.71 1.1

median

9.24 8.3

MAD

Sources: Banpo (Cai and Qiu 1984; Pechenkina et al. 2005), Jiangzhai (Pechenkina et al. 2005; Guo et al. 2011), Shijia (Pechenkina et al. 2005), Guanjia (Dong et al. 2017), Xipo (Gong 2007; Zhang et al. 2010), Xishan and Yuhuazhai (Zhang et al. 2011), Gouwan (Fu et al. 2010). 2 median absolute deviation Statistically significant p values are in bold. 1

Isotopic values vary considerably among Yangshao sites, albeit that all Yangshao human bone samples suggest a very heavy reliance on millets. Banpo and Gouwan are the two Yangshao sites where the remains stand out as having more negative δ13C values and hence evidencing a mixed C3/C4 diet. Only five bone samples were analyzed from Banpo, one by Pechenkina et al. (2005) and four more by Cai and Qiu (1984). Based on these samples, carbon isotopic values at Banpo range from −18.8 to −13.3‰, largely overlapping with the isotopic values from the chronologically earlier Baijia site of the Dadiwan culture (Table 2.2). 45

Kate Pechenkina

Figure 2.2  δ13C and δ15N values of human bone collagen samples from Yangshao sites. Values from the Baijia site of the Dadiwan culture are included for comparative purposes: A. intersite comparison; B. sex differences within each site.

Notwithstanding the contextual similarity and geographic proximity of the Banpo and Yuhuazhai sites, isotopic values of human bone samples from these two sites are quite different. Yuhuazhai δ13C values average −8.4‰, and their range doesn’t overlap with that from Banpo. Several explanations are possible for the disparity between Banpo and Yuhuazhai isotopic values. Given the small number of specimens analyzed from Banpo, it is possible that these are not representative of Banpo general diet and perhaps came from the earlier part of the Banpo occupation. Because of the high proportion of specimens from infant urn burials in the Yuhuazhai sample, its δ13C and δ15N values might be more reflective of juvenile diet and may be enriched by the consumption of breast milk, as well as by a high proportion of cereals in the weaning diet. The Gouwan site is in the southern part of the Yangshao range. More negative δ13C values of human bone samples from this site likely reflect a considerable proportion of cultivated rice in the human diet, as well as the presence of other C3 plants. Fu and colleagues (2010) estimated that millets constituted only half of the human diet there. The presence of a few migrants from the southern rice agricultural area at this site, seen as outliers with highly negative δ13C values, has been also considered possible (Fu et al. 2010). Comparing isotopic values between the three Yangshao phases represented at the site, Fu and colleagues noticed a considerable overlap of δ13C values, while δ15N increased during the two later phases of the Yangshao, perhaps due to greater availability of domesticated animals. A similar trend of increased rice contribution to the human diet in the south has been reported for the Dawenkou (6100 to 4600 bp) culture of Eastern China (Dong 2013). Bone samples from northern Dawenkou sites, Dongjiaying and Fujia, display isotopic values similar to those observed for the majority of Yangshao sites, suggesting strong reliance on millet and some contribution from millet-foddered domesticated animals (Table 2.2). Samples from southern Dawenkou sites are characterized by more negative δ13C values, likely because of a greater proportion of rice in the diet. Aside from the more negative δ13C values at Banpo and Gouwan, there seems to be a slight increase in δ13C values from earlier Yangshao sites of the Wei River area, i.e. Jiangzhai and Shijia, 46

Of millets and wheat Table 2.2 Stable isotope values in human bone collagen samples from the Neolithic and early dynastic archaeological sites site1

sex

δ 13C ‰ mean

δ 15N ‰ SD

median

MAD

Peiligang culture (southern Henan, 9000–7000 bp): Jiahu F −20.65 0.37 −20.6 0.37 Jiahu M −20.47 0.06 −20.5 NA Jiahu U −22 2.79 −20.62 0.66 Jiahu Total −21.66 2.5 −20.51 0.43 t-test (females-males) t = −0.97, p = 0.39 Dadiwan culture (Shaanxi, 7900–7200 bp) Baijia M −12.73 NA −12.73 0 Baijia U −13.33 1.25 −13.3 1.78 Baijia Total −13.18 1.06 −13.02 0.89 Houli culture (Shandong, 8500–7500 bp) Xiaojingshan Total −17.77 0.32 −17.85 0.35 Dawenkou culture (northeastern China, 6100–4600 bp): Liangwangcheng F −11.95 2.69 −11.4 2.97 Liangwangcheng M −10.49 1.42 −10.4 1.33 Liangwangcheng Total −11.3 2.3 −10.9 2.08 t-test (females-males) t = −1.8, p = 0.09 Huating F −14.35 0.49 −14.35 NA Dongjiaying Total −10.41 5.36 −7.6 1.33 Fujia F −7.6 0.45 −7.6 0.44 Fujia M −7.59 0.5 −7.4 0.59 Fujia U −7.8 NA −7.8 0 Fujia Total −7.6 0.46 −7.6 0.59 t-test (females-males) t = −0.04, p = 0.96 Qujialing culture (5000–4600 bp) Total −14.6 0.85 −14.6 NA Gouwan Qinglongquan Total −15.73 0.88 −15.9 0.60 Shijiahe (4600–4200 bp) Qinglongquan Total −14.18 1.1 −14.20 1.1 Xia dynasty (4070–3600 bp) Niedian F −7.2 0.39 −7.2 0.44 Niedian M −7.07 0.33 −7 0.3 Niedian U −7.18 0.32 −7.1 0.3 Niedian Total −7.14 0.35 −7.1 0.3 t-test (females-males) t = −1.11, p = 0.28

mean

N SD

median MAD

8.78 0.49 8.75 8.1 0.52 7.8 5.49 6.95 8.67 6.3 6.06 8.68 t = 1.74, p = 0. 17

0.44 NA 1.62 1.24

4 3 28 35

8.88 10.83 10.35

NA 1.15 1.35

8.88 10.8 10.25

NA 1.63 1.42

1 3 4

8.99

0.56

9.15

0.25

10

9.03 0.86 8.9 8.82 0.6 8.75 8.94 0.75 8.8 t = 0.77, p = 0.44 8.66 0.41 8.66 3.83 6.91 6.7 9.1 0.45 9.1 9.15 0.41 9.2 9.8 NA 9.8 9.16 0.43 9.2 t = −0.30, p = 0.77

0.74 0.59 0.59

15 12 27

NA 1.93 0.44 0.44 0 0.44

2 21 11 11 1 23

7.0 8.69

0.85 1.12

7.0 9.10

NA 0.80

2 7

8.89

1.20

9.2

0.70

17

0.59 0.67 0.44 0.59

17 22 21 60

10.31 0.59 10.5 10.51 0.82 10.55 10.52 0.75 10.6 10.46 0.73 10.6 t = −0.92, p = 0.36

 Sources: Jiahu (Hu et al. 2006; Gong 2007), Xiaojingshan (Hu et al. 2008), Dawenkou sites (Dong 2013), Baijia (Atahan et al. 2011) and unreported data, Qinglongquan (Yunxian County, Hubei Province) (Guo et al. 2011), Niedian (Wang et al. 2014)

1

to the chronologically later sites located in Henan province, i.e. Xipo, Guanjia, and Xishan (Figure 2.2 A).Whether this slight increase in the δ13C values reflects ecological differences between the Wei River area and the middle reaches of the Yellow River or a chronological trend related to a steady increase in the contribution of millet to the human diet over the Neolithic requires further investigation. Given even less negative δ13C values in the Taosi human samples of the 47

Kate Pechenkina

Longshan culture (Zhang et al. 2007) and at the Niedian site of the Xia dynasty (Wang et al. 2014) than in Yangshao samples, the temporal trend explanation seems more likely. Another important aspect of Yangshao isotopic variation, which is clearly seen in Figure 2.2, is that isotopic values of bones from the same site tend to cluster together, often forming spherical clouds that overlap only partially with those of the other sites. For instance, Guanjia samples are characterized by relatively high δ13C and unusually low, for the Yangshao, δ15N values. Thus, all Guanjia data points occupy the lower right part of Figure 2.2A. Xishan δ13C values largely overlap with those of Guanjia, while its δ15N values are considerably higher. Such clustering can be explained by fairly consistent and uniform dietary preferences and/or choices at each site. In addition, different ecology and soil composition, as well as use of fertilizers, could be driving these differences. This clustering also suggests the fairly low long-distance mobility of Yangshao people. A general lack of outliers seems to suggest that most people whose bones were tested for stable isotopes were local to the settlement. In this respect, the Xipo series seems to be an exception, with considerable scattering of isotopic values.The observed variation in isotopic values of human bone samples from Xipo is consistent with a regional centre status for this settlement, as was proposed by Ma Xiaolin and colleagues (2005, 2006).

Male-female differences during the Neolithic Discussions of the development of agriculture-based food production and male-female inequality have been intertwined in Chinese archaeology for over half a century (Shelach 2004, 2006; Chen 2014). In the Marxist archaeological thought of socialist China, F. Engels’ “The Origin of the Family, Private Property, and the State,” (1884) served as the interpretive framework describing basic stages of human social development, starting from the Paleolithic. Engels, building on the observations and ideas of Morgan and Bachofen, suggested that familial ties in early human societies could be described as Bachofen’s “Mutterrecht” – a female dominated commune that traced relationship through maternal lineages, as paternity was generally unknown or uncertain. In this formulation, early agricultural communities were organized around matrilineal clans controlling communal property, with related women performing necessary agricultural tasks cooperatively. The communistic household, in which most of the women or even all the women belong to one and the same gens, while the men come from various other gentes, is the material foundation of that predominancy of women which generally obtained in primitive times. (Engels 1891 [2004]) The dispossession and hence disempowerment of women followed a shift in gender contributions to food production with cattle domestication and the husbandry of large herds of domesticated animals. In archaeological interpretations derived within this framework, early agricultural communities of the Neolithic were generally assumed to be matriarchal, without rigorous testing of Engels’ model (Jiao 2001; Chen 2014).Through the twentieth century, archaeological discussion of gender roles in early China didn’t draw clear distinctions between matrilinearity, matrilocality, and matriarchy, oftentimes using these terms interchangeably and employing any evidence for – and against – the “Mutterrecht” to support or, conversely, reject all three. For instance, reports published on excavations at Banpo (Institute of Archaeology 1963; Museum 1975) referred to burials of females with infants and collective burials arranged according to sex and not in agreement with family units as an illustration of both matrilineage and matriarchy. Matriarchy was inferred for the Yangshao sites of Jiangzhai (Xi’an Banpo Museum 1988) and Yuanjunmiao (Zhang 1985), as well as for the Early Neolithic Houli Culture (Zhang and Lu 2004) from 48

Of millets and wheat

the presence of large communal houses that are ethnographically associated with matrilocality. Yangshao matriarchy was contested by Gao and Lee’s (1993) craniometric study of Shijia burials that found greater morphological variation in a female skull series as compared to the male one, which the authors interpreted as evidence of patrilocality and, by extension, of patriarchy. Because inequality usually goes hand in hand with uneven access to food resources, we compared male and female isotopic values for Yangshao sites. We used two-tailed Welch two-sample t-tests to evaluate the difference between male and female isotopic values (Tables 1 and 2). For the Yangshao samples, sex differences in isotopic values were minimal and generally didn’t attain the level of statistical significance, suggesting that male and female diets during this period were generally similar.The only exception was at Jiangzhai, where males had a significantly higher δ15N values: 9.0‰ vs. 8.4‰ for males and females respectively, suggesting a lesser proportion of animal products in the female diet (Dong et al. 2017). Other Neolithic sites for which sex differences in isotopic values have been reported show a similar absence of male-female dietary differences (Hu et al. 2008; Wang et al. 2014, also see Table 2.2). Although female δ13C values were more negative than those of males in a sample from the Liangwangcheng site of the Late Dawenkou culture from Shandong (4800–4500 cal. bp) (Dong 2013), these differences didn’t attain the level of statistical significance (Table 2.2) and seem to be driven by a few female outliers with unusual dietary signatures.Thus, isotopic data seem to provide little evidence for unequal distribution of resources during the Neolithic.The observed similarity in isotopic values among individuals from the same Yangshao sites seems to suggest a fairly equal distribution of resources among all community members.

Indicators of human health in Yangshao skeletal collections Several papers have compared indicators of oral health and cranial/skeletal lesions among Yangshao skeletal collections (Pechenkina et al. 2002, 2007, 2013). I will briefly discuss whether dietary variation evidenced by stable isotopes translated into unequal distribution of skeletal health markers. To test whether the observed variation in the Yangshao diet had an impact on different aspects of human health we compared the frequencies of carious lesions and anemia indicators from site to site. Both of these skeletal health indicators are known to be diet dependent and tend to have increased in frequency throughout the world along with plant domestication and increased reliance on agriculture (Cohen and Armelagos 1984).

Dental caries Caries is an infectious disease caused by bacterial pathogens including Streptococcus mutans and Lactobacilli. Reliance on domesticated cereals prompts the progression of caries for two reasons. First, sucrose, a disaccharide present in high concentration in many cereals, is necessary for Streptococcus to establish initial infection. Digesting sucrose, Streptococcus secretes a sticky matrix, hence forming a firm attachment with the dental surface. Cereals that contain low levels of sucrose, such as rice, are known to be less cariogenic (Tayles et al. 2000). Second, cariogenic bacteria require readily available water-soluble carbohydrates for their proliferation. Metabolizing sugars, both Streptococcus mutans and Lactobacilli produce acid that dissolves dental enamel, eventually leading to formation of a carious lesion. The frequency of carious lesions considerably and statistically significantly increased from the Early Yangshao to the Middle/Late Yangshao. This pattern is in agreement with the trend indicated by stable isotopes, suggesting an increase in the dietary contribution of millet in later Yangshao sites and particularly at Guanjia and Xishan. Although the paleodietary signatures for Yangshao males and females overlap completely, the observed differences in caries frequencies 49

Kate Pechenkina

between the sexes are not unexpected. Female physiology – specifically low saliva flow, different saliva composition, and physiological changes triggered by pregnancy – favors the initiation and progression of caries (Lukacs and Largaespada 2006). Despite low δ15N and high δ13C values, which suggest a millet-rich diet with low amounts of animal products, caries frequencies from Guanjia were similar to those at other Middle and

Figure 2.3  δ13C and δ15N values of human bone collagen samples from Yangshao sites compared to isotopic values of human collagen samples from other cultures and periods. Individuals from Jiahu and Dawenkou sites with δ15N values below 4‰ were not included to improve resolution. Because rice consumption has likely affected Gouwan isotopic values, data from that site were not shown on this chart. Table 2.3  Frequency of carious teeth in Yangshao skeletal collections Site

Jiangzhai Shijia Guanjia Xipo Xishan

males

females

present

affected

%

present

affected

%

262 137 568 424 695

5 5 65 43 98

2.0 3.6 11. 4 10.1 14.1

174 45 446 190 989

6 2 102 18 173

3.4 4.4 22.9 9.5 17.5

 Based on (Pechenkina et al. 2013)

1

50

χ2

p, 1 df

1.09 0.04 29.83 0.06 5.04

.2975 .8376 .0000 .7988 .0248

Of millets and wheat Table 2.4  Frequency of porotic hyperostosis and cribra orbitalia in Yangshao skeletal collections1 Cribra orbitalia

Jiangzhai Shijia Guanjia Xipo Xishan 1

porotic hyperostosis

present

affected

%

present

affected

%

28 38 31 21 43

5 1 4 3 6

17.9 2.6 12.9 14 14

28 39 37 23 45

0 9 4 2 7

0 2.3 11 8 15.5

based on (Pechenkina et al. 2013)

Late Yangshao sites (Figure 2.3). One peculiar aspect of Guanjia oral health is the notable frequency of caries on front teeth, incisors and canines. Samples from other Yangshao sites display carious lesions almost exclusively on posterior teeth. Crowns of anterior teeth are smooth and have no fissures. Regular saliva flow and mechanical pressures generated by drinking and chewing typically remove bacterial growth from the anterior teeth; hence caries on incisors and canines is very rare globally. Sugar-intense diets and habitual chewing of sweets have been known to lead to anterior carious infection. Nevertheless, there is no evidence of sweets consumption in early China. We can hypothesize that mushy and non-abrasive well-cooked millet could have caused anterior caries in Guanjia dentitions in the absence of more abrasive food products.

Porotic hyperostosis and cribra orbitalia Hyperporosity of cranial bones associated with diploe expansion has been linked to childhood anemia by paleopathological research (Walker et al. 2009). Compensatory hyperplasia of the red bone marrow contained in the medullary cavities of the spongy bone leads to substantial expansion of the spongy region sandwiched between the cortical layers of the cranial bones, along with thinning of the cortical layer, giving cranial bones a porous, hair-on-end appearance. Such cranial lesions are referred to as porotic hyperostosis or cribra cranii when found on the bones of the cranial vault. Porous lesions on the cranial roof are called cribra orbitalia. A transition to agriculture, agricultural intensification, and population growth have been known to lead to increasing frequencies of these anemia indicators in past human societies. Several contributing factors have been discussed in the literature. Overall, domesticated cereals furnish low amounts of dietary iron. Furthermore, absorption of iron from plant tissues is somewhat less efficient than of the heme iron found in animal tissues.Thus, cereal-rich diets are expected to cause iron-deficiency anemia. Overuse of cooked grains as the base for a weaning diet leads to increased anemia among infants. An increase in intestinal parasites in overcrowded settlements leads to considerable losses of iron through bloody diarrhea, further exacerbating iron deficiency. Similar to the trend observed for oral health markers, there is a significant increase in anemia indicators from the Early to Middle/Late Yangshao and a slight eastward trend toward an increase in these frequencies, although the latter is not statistically significant (Table 2.4). The easternmost skeletal series, from Xishan, displays the highest frequency of anemia indicators among the collections analyzed.

51

Kate Pechenkina

Isotopic evidence for a dietary shift toward wheat and barley agriculture Wheat and barley first appear in the paleobotanical assemblages left on the Central Plain during the Late Neolithic (ca. 4600–3900 years ago). However, based on the human collagen stable isotope values from early archaeological sites, the dietary contribution of these new cereals to human diet was minimal until the late Bronze Age. Human stable isotope values from the Wei River valley suggest that dependence on millet was increasing until approximately 4000 years ago, whereas a clear shift toward reliance on C3 plants occurred around 2500 cal bp (Atahan et al. 2014). Figure 2.3 shows that the contribution of millets may have further increased from Yangshao levels during Longshan and the Xia dynasty (ca. 4070–3600 bp) (Wang et al. 2014). The earliest clear evidence of the intrusion of plants utilizing the C3 pathway of photosynthesis into human diets in the Yellow River area comes from Eastern Zhou (770–221 bc or 2720–2171 bp) archaeological sites (Hou et al. 2012; Dong et al. 2017; Zhou et al. 2017). The range of variation in δ13C values from Eastern Zhou sites tends to be greater than the range at each Yangshao site (Table 2.5, Figure 2.4). Thus, approximately a third of human bone samples from Eastern Zhou contexts display isotopic values completely within the range of Yangshao sites from the Wei and Yellow River valleys. Male-female differences in isotopic values are more marked in Eastern Zhou samples, so that female skeletons on average display more negative δ13C and lower δ15N values. These differences attained the level of significance in the Changxinuan assemblage for both δ13C and δ15N, in δ13C only for Xiyasi, and in δ15N for Xinghong (Table 2.5). These sex-related differences in dietary signatures likely show that Eastern Zhou females had lesser access to animal products and an increased reliance on the less prized grains, such as wheat, barley, and beans. Observed unequal distribution of food sources between males and females suggest the rise of gender inequality (Dong et al. 2017). Alternatively, consumption of millet-based fermented beverages by males could have been responsible for the observed sex differences. Ligang Zhou (Zhou 2016) noticed that bone samples from more elaborate burials that included two nested coffins had less negative δ13C values, suggestive of a greater proportion of millet in the diet of wealthier people, supporting the hypothesis that C3 cereals were less valued than millet during the Eastern Zhou. Three skeletal series that show significant differences in isotopic values between males and females are associated with the urban population of the Ancient Zhenghan city and generally represent fairly wealthy urban dwellers. No such differences were observed for other Eastern Zhou sites, where people represented likely came from rural farming communities (Table 2.5). Thus, a gender divide in access to resources can be documented during the Eastern Zhou only in some settings and is associated with an overall wealthier population. Carbon isotope values for human bone samples excavated from burials of the Han dynasty (206 bc – 220 ad) (Hou et al. 2012; Zhou 2016) suggest a further increase of C3 plants in the human diet. Development of mills allowed converting wheat into fine flour for noodle production and likely prompted a greater appreciation of wheat, apparently elevating its status to a more desirable cereal toward the end of the Han dynasty (Bray 1984: 461).

The animal story The number of animal bone samples analyzed from the Central Plain remains low, yet there is better chronological continuity in the animal isotopic record (Table 2.6). As attested to by fairly high δ13C values in pig and dog bone samples from early Neolithic contexts, millet was 52

Of millets and wheat Table 2.5  Stable isotope values in human bone collagen samples from the Eastern Zhou archaeological sites site1

sex

δ 13C ‰ mean

δ 15N ‰ SD

median

Xinzheng sites: Changxinyuan F −11.24 1.31 −11.29 Changxinyuan M −9.21 0.51 −9.02 Changxinyuan Total −10.29 1.44 −9.84 t-test (females-males) t = −4.05, p = 0.002 −14.2 Xiyasi F −13.82 1.8 Xiyasi M −10.78 1.63 −10.23 Xiyasi Total −11.89 2.23 −11.45 t-test (females-males) t = −4.62, p = 0.0002 Thermal Plant F −11.21 1.35 −11.05 Thermal Plant M −10.97 0.67 −11.3 Thermal Plant Total −11.15 1.18 −11.3 t-test (females-males) t = −0.40, p = 0.69 Xinghong F −11.17 1.43 −10.95 Xinghong M −10.46 1.81 −10 Xinghong U −11.75 2.3 −11.65 Xinghong Total −11 1.67 −10.8 t-test (females-males) t = −1.52, p = 0.12 Nanyang City, Xichuan County, Henan Province: Shenmingpu Total −12.69 0.79 −12.75 Wenxian county, Henan Province: Chenjiagou F −10.32 1.2 −10.25 Chenjiagou M −9.22 0.78 −8.9 Chenjiagou U −9.97 1.89 −9.5 Chenjiagou Total −9.68 1.47 −9.4 t-test (females-males) t = −1.75, p = 0.16

MAD

mean

0.76 0.62 1.3

7 0.61 6.84 8.57 0.63 8.71 7.73 1.01 7.68 t = −4.88, p = 0.0003 7.77 0.99 7.73 8.33 0.85 8.4 8.12 0.93 8.18 t = −1.56, p = 0.13 8.19 0.76 8.25 8.83 0.81 9.2 8.36 0.79 8.4 t = −1.19, p = 0.31 8.62 0.66 8.55 9.22 0.86 9.4 8.95 0.52 8.85 8.85 0.76 8.8 t = −2.71, p = 0.010

0.71 0.64 1.51

8 7 15

0.7 1.07 0.93

11 19 30

0.82 0.3 1.04

8 3 11

0.52 0.89 0.15 0.82

36 20 6 62

0.89

8.74

9.2

0.74

14

1.41 0.59 0.74 0.74

8.9 1.02 9 9.55 0.66 9.7 8.9 0.68 8.9 9.18 0.76 9.1 t = −1.20, p = 0.30

0.96 0.74 0.74 0.89

4 17 18 39

1.1 1.75 2.92 1.48 0.15 1.48 1.33 1.04 0.44 1.41

SD

N

1.25

median

MAD

 Sources: Xiyasi and Changxinyuan (Dong et al. 2017), Xinhong, Thermal Power Plant, and Chenjiagou (Zhou et al. 2017), Shenmingpu (Hou et al. 2012).

1

Statistically significant p values are in bold.

an important component of their fodder from the initial stages of its cultivation on the Central Plain (Barton et al. 2009; Atahan et al. 2011). When domesticated cattle became available in the region, millet seems to have assumed an important role in their provisioning as well, likely in the form of millet straw (Hou et al. 2013; Chen et al. 2016; Dong et al. 2017). Pig bone samples from the Yangshao Xipo site, as well as from a number of Longshan sites, are characterized by very low δ15N values, generally lower than those of humans from the same time periods (Pechenkina et al. 2005; Wu et al. 2007; Chen et al. 2012; Dai et al. 2016), suggesting that pigs were receiving food refuse with a high millet content and a low proportion of animal products. Pig bones from the Kangjia site of the Longshan period and from later dynastic sites show a marked increase in δ15N values (Zhang et al. 2007; Hou et al. 2013; Ma et al. 2016). Combining pig sties with latrine areas would give pigs access to human feces and could increase their δ15N values via tropic level enrichment due to consuming that fecal matter. Adding fertilizer to millet fields during the early dynasties could potentially have increased the δ15N values 53

Kate Pechenkina

Figure 2.4  δ13C and δ15N values of human bone collagen samples from Eastern Zhou sites. Top: Eastern Zhou vs. Yangshao. Because rice consumption has likely affected Gouwan isotopic values, data from that site were not shown on this chart. Bottom: sex differences within each Eastern Zhou site.

of the cereals themselves. However, the latter explanation doesn’t seem to be supported by the high variation of δ15N values in human bone samples from early dynastic sites (Figure 2.3). Field fertilization should lead to greater uniformity of δ15N values, especially for sites that display a narrow range of δ13C variation.

54

Henan   Henan

Shaanxi

Shanxi

Henan

Henan

Henan   Henan

  Hebei

  Henan

Xipo   Xipo

Kangjia

Taosi

Xinzhai

Wadian

Erlitou   Zhangdeng

  Baicun

  Tianli and Changxinyuan

  Eastern Zhou

  Proto−Shang ca. 2000–1600 bc

Erlitou 1900–1500 bc Proto−Shang ca. 2000–1600 bc

Late Longshan 2600–1900 bc

Late Longshan

Late Longshan 2600–1900 bc

Middle Yangshao 4000–3500 bc Middle Yangshao 4000–3500 bc Longshan 3000–1900 bc

period

pig dog bovine sheep deer pig dog bovine sheep pig pig dog bovine sheep deer pig dog bovine sheep deer pig   pig dog bovine sheep pig dog bovine sheep deer bovine dog pig sheep

pig dog pig

species

2.4 3.9 0.6 NA NA 1.6 0.4 2.2 0.4 3.6 1 2.1 1.7 1.6 3.3 2.4 1 2 0.9 0.9 3 1.3 0.9 3.4 2.7 0.4 1.6 1.8 2.6 0.5 1.9 2.9 2.7 2.1

−7.7 −7.6 −9 −15.4 −7 −8 −8.3 −12.1 −20.8 −10.1 – 12.9 – 11.7 – 15.7

0.2 NA 0.4

−10.3 −11.8 −14.7 −18.8 −17.2 −7.1 −6.8 −11.3 −17.2 −9.2 −8.5 −10.4 −9.8 −14.4 −16.2 −11.4 −10.1 −12.7 −16.6 −20.8 −10.5

−7.5 −8.1 −7.0

−7.3 −7.5 −9.6 −15.1 −7.2 −8.0 −8.3 −11.9 −21 −10.4 – 13.2 – 13.1 – 15.8

−11.5 −11.8 −14.7 NA NA −6.6 −6.7 −11.8 −17.2 −9.2 −8.1 −10.4 −9.9 −14.8 −16.4 −11.6 −10.5 −13.2 −16.6 −21.1 −12

−7.5 NA −7.1

−11.2 −8.8 −18.9 −21.4 −7.2 −9.1 −10.9 −16.5 −21.3 −11.7 – 15.7 – 14.4 – 18.6

−11.8 −14.3 −15.1 NA NA −11.8 −7.5 −13.5 −17.7 −20.2 −10.7 −12.8 −12.5 −16.2 −12.3 −16.1 −11 −16 −17.3 −21.6 −12.5

−7.7 NA −7.5

min

−6.4 −6.7 −6.2 −10.7 −6.5 −6.9 −5.8 −6.4 −20.2 −6.8 – 9.4 – 7.9 – 13.1

−7.5 −9.0 −14.2 NA NA −6.2 −6.4 −7.3 −16.6 −8.1 −7.1 −7.9 −7 −11.5 −19.8 −8.1 −8.5 −9.4 −16 −18.8 −7.1

−7.4 NA −6.5

max

7.7 7.3 6.8 7.7 7.8 7.2 7.3 6.2 3.4 6.4 6.2 6.1 9.0

8.7 9.7 7.0 6.6 8.0 7.2 8.8 6.6 6.8 6.2 6.2 6.8 6.3 5.6 5.3 6.9 7.2 7.6 7.6 5 8.4

7.7 6.9 NA

mean

med

mean

SD

δ15N, ‰

δ13C, ‰

0.5 0.4 1.4 1.2 0.5 0.5 1.0 1.1 0.5 2.2 1.1 1.1 1.6

0.9 0.3 0.6 NA NA 0.4 1.1 1.2 1 1.8 0.9 1.2 0.9 0.5 0.8 1 1.1 0.8 0.1 1.2 0.6

0.3 NA NA

SD

7.7 7.4 6.2 8 7.7 7.2 7.3 5.8 3.3 7.2 6.3 6.1 9.3

8.7 9.7 7.0 NA NA 7.3 8.5 6.3 7.4 6.1 6.1 6.6 6.1 5.8 5.2 7 7.2 7.8 7.6 5.2 8.7

7.7 NA NA

med

6.6 6.7 4 5.2 7.4 6.9 5.9 5.1 2.8 2.4 4.9 5.1 6.6

7.8 9.5 6.6 NA NA 6.6 7.5 5.3 5.5 4.5 4.4 5.8 4.9 4.8 4.4 5.5 5.9 6 7.5 2.8 7.7

7.5 NA NA

min NA NA

8.4 7.7 8.7 10.8 8.3 7.6 8.3 8.3 3.9 7.7 7.3 7.7 10.7

9.6 9.8 7.4 NA NA 7.9 10 8.7 7.8 10 7.6 8.4 7.6 6.2 6.4 8.7 8.5 9.1 7.7 6.9 8.8

8

max

3 2 2 1 1 11 5 6 5 10 11 4 11 8 4 10 7 9 2 10 3   18 4 12 13 3 2 5 13 5 5 4 7 6

2 1 4

N

Xipo (Pechenkina et al. 2005; Zhang et al. 2011), Kangjia (Pechenkina et al. 2005),Taosi (Chen et al. 2012), Xinzhai (Wu et al. 2007; Dai et al. 2016),Wadian (Chen et al. 2016), Erlitou (Zhang et al. 2007), Zhangdeng (Hou et al. 2013), Baicun (Ma et al. 2016), Tianli and Changxinyuan (Dong et al. 2017; Zhou et al. 2017), and present chapter.

province

site

Table 2.6 Animal stable isotope data from Neolithic and Bronze Age archaeological contexts of the Chinese Central Plain

Kate Pechenkina

Eastern Zhou pig bone samples display a much larger range of δ13C variation than pig bones from earlier sites. Because domesticated pigs tend to have a short lifespan, this range in δ13C values from Eastern Zhou may reflect the variation in wheat harvests and wheat availability from year to year. Provisioned by typical household refuse, pig isotopic values may also mirror status differences among Eastern Zhou households. Finally, sheep and goats show highly negative δ13C values, similar to those of wild animals in the area and indicative of an overwhelmingly C3 pattern of wild vegetation in the pastures. A marked separation between human and sheep/goat isotopic values seems to suggest that these animals were raised for wool and/or ritual purposes, contributing to human diet only infrequently (Dong et al. 2017).

Conclusions A comparative analysis of stable isotope values among multiple Yangshao sites indicates that human diets varied considerably from site to site in terms of the proportion of millets and animal products in the diet. These dietary differences are reflected in the variation in oral health, as Yangshao skeletal assemblages with higher δ13C values also display a higher rate of caries. There is also a trend toward a greater frequency of cranial lesions suggestive of childhood anemia in human remains from Middle/Late Yangshao sites with higher δ13C, although these differences do not attain statistical significance. Despite considerable diet variation during Yangshao, the composition of male and female diets appears to have been very similar, suggesting that access to food resources was not gender biased at the time. An increased proportion of C3 cereals in the human diet, i.e. wheat and barley, as well as beans, is evident isotopically only after 2700 bp. At the time of the Eastern Zhou, human isotopic values and hence human diets became more variable within each site yet overlapped considerably among different sites. The composition of male diets as reflected by stable isotopes is often significantly different from that of females. A greater proportion of C3 plants and lesser proportion of animal products became characteristic of female diets during the Eastern Zhou. These differences suggest gender-biased access to food resources and likely gender segregated meal consumption. The presence of C3 cereals in pig feed is also seen during the Eastern Zhou, indicating that C3 cereals had become a common part of the household refuse, which was used to provision pigs.

Works cited Ambrose, S.H. (1991) ‘Effects of diet, climate and physiology on nitrogen isotope abundances in terrestrial foodwebs’, Journal of Archaeological Science, 18: 293–317. Ambrose, S.H. (1993) ‘Isotopic analysis of palaeodiets: methodological and interpretive considerations’, in: M. K. Sandford (ed.) Investigations of Ancient Human Tissue, Langhorne: Gordon and Breach Science Publishers. Atahan, P., Dodson, J., Li, X.Q., Zhou, X.Y., Chen, L., Barry, L., and Bertuch, F. (2014) ‘Temporal trends in millet consumption in northern China’, Journal of Archaeological Science, 50: 171–7. Atahan, P., Dodson, J., Li, X., Zhou, X.Y., Hu, S., Chen, L., Bertuch, F., and Grice, K. (2011) ‘Early Neolithic diets at Baijia, Wei river valley, China: stable carbon and nitrogen isotope analysis of human and faunal remains’, Journal of Archaeological Science, 38: 2811–7. Barton, L., Newsome, S.D., Chen, F.H., Wang, H., Guilderson, T.P., Bettinger, R.L., and Hole, F. (2009) ‘Agricultural origins and the isotopic identity of domestication in northern China’, Proceedings of the National Academy of Sciences of the United States of America, 106: 5523–8. Bocherens, H., and Drucker, D. (2003) ‘Trophic level isotopic enrichment of carbon and nitrogen in bone collagen: Case studies from recent and ancient terrestrial ecosystems’, International Journal of Osteoarchaeology, 13: 46–53.

56

Of millets and wheat Bray, F. (1984) Biology and Biological Technology: Agriculture, Cambridge: Cambridge University Press. Cai, L., and Qiu, S. (1984) ‘Carbon-13 evidence for ancient diets in China (碳十三测定和古代食谱研 究)’, Kaogu (考古), 10: 949–54. Chen, Q. (2006) ‘A study of the faunal remains from the Xishan site in Zhengzhou (郑州西山遗址出土 动物遗存研究)’, Acta Archaeologica Sinica (考古学报), 3: 385–418. Chen, S.X. (2014) ‘The creation of female origin myth: a critical analysis of gender in the archaeology of Neolithic China’, Totem:The University of Western Ontario Journal of Anthropology, 22: 24–30. Chen, X., Fang,Y., Hu,Y., Hou,Y., Lü, P.,Yuan, J., Song, G., Fuller, B.T., and Richards, M.P. (2016) ‘Isotopic reconstruction of the late Longshan period (ca. 4200–3900 bp) dietary complexity before the onset of state-level societies at the Wadian site in the Ying river valley, Central Plains, China’, International Journal of Osteoarchaeology, 26: 808–17. Chen, X.,Yuan, J., Hu,Y., He, N., and Wang, C. (2012) ‘A preliminary exploration to the domestic animal raising strategy: the evidences from carbon and nitrogen isotope analyses (陶寺遗址家畜饲养策略初 探:来自碳、氮稳定同位素的证据)’, Archaeology and Technology(考古与科技), 9: 843–50. Cohen, M. N. and Armelagos, G. J. Editors. (1984) Paleopathology at the Origins of Agriculture. Orlando, FL: Academic Press. Crawford, G., Underhill, A., Zhao, Z., Lee, G-A., Feinman, G., Nicholas, L., Luan, F.,Yu, H., Fang, H., and Cai, F. (2005) ‘Late Neolithic plant remains from northern China: preliminary results from Liangchengzhen, Shandong’, Current Anthropology, 46: 309–17. Dai, L., Li, Z., Hu, Y., Zhao, Q., and Wang, C. (2014) ‘The age profile of sheep and the strategy of animal resources utilization at Xinzhai site [新砦遗址出土羊的死亡年龄及畜产品开发策略]’, Kaogu (考 古), 1: 94–103. Dai, L.L., Li, Z.P., Zhao, C., Yuan, J., Hou, L., Wang, C., Fuller, B.T., and Hu, Y. (2016) ‘An isotopic perspective on animal husbandry at the Xinzhai site during the initial stage of the legendary Xia dynasty (2070–1600 bc)’, International Journal of Osteoarchaeology, 26: 885–96. Deniro, M.J., and Epstein, S. (1981) ‘Influence of diet on the distribution of nitrogen isotopes in animals’, Geochimica et Cosmochimica Acta, 45: 341–51. Dodson, J.R., Li, X., Zhou, X., Zhao, K., Sun, N., and Atahan, P. (2013) ‘Origin and spread of wheat in China’, Quaternary Science Reviews, 72: 108–11. Dong,Y. (2013) ‘Eating identity: food, gender, and social organization in Late Neolithic Northern China‘, unpublished Ph.D. thesis, University of Illinois at Urbana-Champaign. Dong,Y., Morgan, C., Chinenov,Y., Zhou, L., Fan,W., Ma, X., and Pechenkina, K. (2017) ‘Shifting diets and the rise of male-biased inequality on the Central Plans of China during Eastern Zhou’, Proceedings of the National Academy of Sciences of the United States of America, 114: 932–7. Engels, F. (1891 [2004]) The Origin of the Family, Private Property and the State, Chippendale, NSW: Resistance Books. Fan, W. (2000) ‘Results of excavations at guanjia (关家遗址发掘获重要成果)’,Zhongguo Wenwu Bao (中 国文物报): 1. FAO, Food Policy and Nutrition Division. 1972. Food composition table for use in East Asia. Rome: Food Policy and Nutrition Division, Food and Agriculture Organization of the United Nations. Rome: FAO. Farquhar, G.D., O’Leary, M.H., and Berr, J.A. (1984) ‘On the relationship between carbon isotope discrimination and the intercellular carbon dioxide concentration in leaves’, Australian Journal of Plant Physiology, 9: 121–37. Flad, R., Li, S.,Wu, X., and Zhao, Z. (2010) ‘Early wheat in China: results from new studies at Donghuishan in the Hexi corridor’, The Holocene, 20: 955–65. Fu, Q., Jin, S., Hu, Y., Ma, Z., Pan, J., and Wang, C. (2010) ‘Agricultural development and palaeodietary study of Gouwan site, Xichuan, Henan [河南淅川沟湾遗址农业发展方式和先民食物结构变化]’, Zhongguo Kexue Zazhishe (中国科学杂志社), 55: 589–95. Fuller, D.Q., Denham, T., Arroyo-Kalin, M., Lucas, L., Stevens, C.J., Qin, L., Allaby, R.G., and Purugganan, M.D. (2014) ‘Convergent evolution and parallelism in plant domestication revealed by an expanding archaeological record’, Proceedings of the National Academy of Sciences of the U S A, 111: 6147–52. Fuller, D.Q., and Zhang, H. (2007) ‘A preliminary report of the survey archaeobotany of the upper Ying valley (Henan province) (颍河中上游谷地植物考古调查的初步报告)’, in: School of Archaeology and Museology, Peking University and Henan Provincial Institute of Cultural Relics and Archaeology (ed.) Archaeological Discovery and Research at the Wangchenggang Site in Dengfeng (2002–2005) (登封王城 岗考古发现与研究 [2002～2005]), Zhengzhou: Great Elephant Publisher. Gao, Q., and Lee, K.Y. (1993) ‘A biological perspective on Yangshao kinship’, Journal of the Anthropological Archaeology, 12: 266–98.

57

Kate Pechenkina Gong JZ (2007) ‘Inequalities in diet and health during the intensification of agriculture in Neolithic China’. Unpublished BA Thesis (Harvard University). Guo,Y., Hu,Y., Gao, Q., Wang, C., and Richards, M.P. (2011) ‘Stable carbon and nitrogen isotope evidence in human diets based on evidence from the Jiangzhai site, China’, Acta Anthropologica Sinica (人类学学 报), 30: 149–57. Hou, L., Li, S., Hu,Y., Hou,Y., Lu, P., Cao, L., Hu, B., Song, G., and Wang, C. (2013) ‘A preliminary study of the mode of domestic animal husbandry in the pre-Shang culture period. (先商文化时期家畜饲养 方式初探)’, Huaxia Kaogu (华夏考古), 2: 130–9. Hou, L., Wang, N., Lu, P., Hu,Y., Song, G., and Wang, C. (2012) ‘Transition of human diets and agricultural economy in Shenmingpu site, Henan, from the Warring States to Han dynasties’, Zhongguo Kexue: Diqiu Kexue (中国科学: 地球科学), 42: 1018–25. Hu,Y., Ambrose, S.H., and Wang, C. (2006) ‘Stable isotopic analysis of human bones from Jiahu site, Henan, China: implications for the transition to agriculture’, Journal of Archaeological Science, 33: 1319–30. Hu, Y., Wang, S., Luan, F., Wang, C., and Richards, M.P. (2008) ‘Stable isotope analysis of humans from Xiaojingshan site: implications for understanding the origin of millet agriculture in China’, Journal of Archaeological Science, 35: 2960–5. Institute of Archaeology, Academia Sinica. (1991) Radiocarbon Dates 1965–1991, Beijing: Cultural Relics Publishing House. Institute of Archaeology, Academia Sinica, Xi’an Banpo Museum. (1963) The Neolithic Village at Banpo, Xian (西安半坡), Beijing: Wenwu. Jiao, T. (2001) ‘Gender studies in Chinese Neolithic archaeology’, in: B. Arnold and N. L. Wicker (eds.) Gender and the Archaeology of Death, Walnut Creek, CA: AltaMira Press. Lan, W., and Chen, C. (2014) ‘Plant macroremains from Guanzhuang site in Xingyang, Henan (荥阳官庄 遗址浮选样品植物大遗存分析)’, Dongfang Kaogu (东方考古): 78–89. Lee, G-A., and Bestel, S. (2007) ‘Contextual analysis of plant remains at the Erlitou period Huizui site, Henan, China’, Bulletin of the Indo-Pacific Prehistory Association, 27: 49–60. Lee, G.A., Crawford, G.W., Liu, L., and Chen, X. (2007) ‘Plants and people from the Early Neolithic to Shang periods in north China’, Proceedings of the National Academy of Sciences of the U S A, 104: 1087–92. Lee,Y.K. (2007) ‘Centripetal settlement and segmentary social formation of the Banpo tradition’, Journal of Anthropological Archaeology, 26: 630–75. Li, F. (2013) Early China: A Social and Cultural History (New Approaches to Asian History), Cambridge: Cambridge University Press. Li, X., Dodson, J., Zhou, X., Zhang, H., and Masutomoto, R. (2007) ‘Early cultivated wheat and broadening of agriculture in Neolithic China’, The Holocene, 17: 555–60. Li, Z., Brunson, K., and Dai, L. (2014) ‘The zooarchaeological study of wool exploration from the Neolithic age to the early Bronze age in the Central Plains (中原地区新石器时代到青铜时代早期羊毛 开发的动物考古学研究)’, Quaternary Sciences (第四纪研究), 34: 149–57. Liu, D. (1986) ‘Survey of the Neolithic village of Xishan in Zhengzhou city (郑州市西山村新石器时代 遗址调查简报)’, Cultural Relics of Central China (中原文物）, 2: 23–6. Liu, L.,Yang, D., and Chen, X. (2006) ‘On the origin of the Bubalus bubalis in China (中国家养水牛起源 初探)’, Acta Archaeologica Sinica (考古学报), 2: 141–78. Liu, X., Fuller, D.Q., and Jones, M. (2012) ‘Early agriculture in China’, in: G. Barker and C. Goucher (eds.) The Cambridge World History, Cambridge: Cambridge University Press. Lu, H.,Yang, X.,Ye, M., Liu, K.B., Xia, Z., Ren, X., Cai, L.,Wu, N., and Liu,T.S. (2005) ‘Culinary archaeology: Millet noodles in late Neolithic China’, Nature, 437: 967–8. Lu, H., Zhang, J., Liu, K.B.,Wu, N., Li,Y., Zhou, K.,Ye, M., Zhang,T., Zhang, H.,Yang, X., Shen, L., Xu, D., and Li, Q. (2009) ‘Earliest domestication of common millet (Panicum miliaceum) in East Asia extended to 10,000 years ago’, Proceedings of the National Academy of Sciences of the U S A, 106: 7367–72. Lukacs, J., and Largaespada, L. (2006) ‘Explaining sex differences in dental caries prevalence: saliva, hormones, and “life-history” etiologies’, American Journal of Human Biology, 18: 540–55. Lü, P. (2010) ‘On the origin of Chinese domestic cattle’, in: Zooarchaeology (动物考古). Beijing: Cultural Relic Publishing House. Ma, X. (2003) ‘Emergent social complexity in the Yangshao culture: Analyses of settlement patterns and faunal remains from Lingbao, western Henan, China’, unpublished Ph.D. thesis, La Trobe University, Australia. Ma, X., Li, X., and Yang, H. (2005) A Great Breakthrough in the Fifth Excavation at Xipo in Lingbao, Henan. 26. Zhengzhou: Henan Provincial Institute of Cultural Relics and Archaeology., Zhengzhou: Henan Provincial Institute of Cultural Relics and Archaeology.

58

Of millets and wheat Ma, X., Li, X., and Yang, H. (2006) ‘A study of jade articles from the Yangshao cultural cemetery at Xipo in Lingbao. (灵宝西坡仰韶文化墓地出土玉器初步研究)’, Cultural Relics of Central China (中原文 物), 3: 72–6. Ma, Y., Fuller, B.T., Wei, D., Shi, L., Zhang, X., Hu, Y., and Richards, M.P. (2016) ‘Isotopic perspectives (δ 13C, δ 15N, δ 34 S) of diet, social complexity, and animal husbandry during the proto-Shang period (ca. 2000–1600 bc) of China’, American Journal of Physical Anthropology, 160: 433–45. McGovern, P.E., Zhang, J., Tang, J., Zhang, Z., Hall, G.R., Moreau, R.A., Nunez, A., Butrym, E.D., Richards, M.P., Wang, C.-S., Cheng, G., Zhao, Z., and Wang, C. (2004) ‘Fermented beverages of pre- and protohistoric China’, Proceedings of the National Academy of Sciences of the United States of America, 101: 17593–8. Minagawa, M., and Wada, E. (1984) ‘Stepwise enrichment of δ15N along food chains: Further evidence and the relation between δ15N and animal age’, Geochim Cosmochim Ac, 48: 1135–40. O’Leary, M. H. (1988) Carbon isotopes in photosynthesis: fractionation techniques may reveal new aspects of carbon dynamics in plants. Bioscience. 38:328–36. Pechenkina, E., Ambrose, S., Ma, X., and Benfer, R.J. (2005) ‘Reconstructing northern Chinese Neolithic subsistence practices by isotopic analysis’, Journal of Archaeological Science, 32: 1176–89. Pechenkina, E., Benfer, R., and Ma, X. (2007) ‘Diet and health in the Neolithic of the Wei and middle Yellow river basins, Northern China’, in: M. Cohen and G. Crane-Kramer (eds.) Ancient Health: Skeletal Indicators of Agricultural and Economic Intensification, Gainesville: University Press of Florida. Pechenkina, E.A., Benfer, R.A., Jr., and Wang, Z.J. (2002) ‘Diet and health changes at the end of the Chinese Neolithic: the Yangshao/Longshan transition in Shaanxi province’, American Journal of Physical Anthropology, 117: 15–36. Pechenkina, K., Ma, X., and Fan, W. (2013) ‘Trajectories of health in early farming communities of East Asia’, in: K. Pechenkina and M. Oxenham (eds.) Bioarchaeology of East Asia: Movement, Contact, Health, Gainesville: University Press of Florida. Peng,Y., Xiao, J., Nakamura, T., Toshio, Liu, Baolin, and Inouchi,Y. (2005) ‘Holocene East Asian monsoonal precipitation pattern revealed by grain-size distribution of core sediments of Daihai lake in Inner Mongolia of north-central China’, Earth and Planetary Science Letters, 233: 467–79. Ren S, Wu Y. 1999. Fifty years of Neolithic archaeology in China. Kaogu (考古) 9:11–22. Schoeninger, M.J., and Deniro, M.J. (1984) ‘Nitrogen and carbon isotopic composition of bone collagen from marine and terrestrial animals’, Geochim Cosmochim Ac, 48: 625–39. Schoeninger, M.J., and Moore, K.M. (1992) ‘Bone stable isotope studies in archaeology’, Journal of World Prehistory, 6: 247–96. Shelach, G. (2004) ‘Marxist and post-Marxist paradigm for the Neolithic’, in: K. M. Linduff and Y. Sun (eds.) Gender and Chinese Archaeology, Walnut Creek, CA: AltaMira Press. Shelach, G. (2006) ‘Economic adaptation, community structure, and sharing strategies of households at early sedentary communities in northeast China’, Journal of Anthropological Archaeology, 25: 318–45. Tayles, N., Domett. K., and Nelsen K. (2000) ‘Agriculture and dental Caries? The case of rice in prehistoric Southeast Asia’, World Archaeology, 32: 68–83. van der Merwe, N.J. (1982) ‘Carbon isotopes, photosynthesis, and archaeology: different pathways of photosynthesis cause characteristic changes in carbon isotope ratios that make possible the study of prehistoric human diets’, American Scientist, 70: 596–606. Walker, P.L., Bathurst, R.R., Richman, R., Gjerdrum,T., and Andrushko,V.A. (2009) ‘The causes of porotic hyperostosis and cribra orbitalia: a reappraisal of the iron-deficiency-anemia hypothesis’, American Journal of Physical Anthropology, 139: 109–25. Wang, X. (1993) ‘On the Shijia type of Yangshao culture (论仰韶文化史家类型)’, Acta Archaeologica Sinica (考古学报), 4: 415–34. Wang,Y., Nan, P.,Wei, D., Hu,Y.,Wang, X., and Wang, C. (2014) ‘Dietary differences in humans with similar social hierarchies: example from the Niedian site, Shanxi (相近社会等级先民的食物结构差异--以山 西聂店遗址为例)’, Acta Anthropologica Sinica (人类学学报), 33: 82–9. Wei, X. (2014) ‘The discoveries and preliminary research of plant remains in the region of the western Henan and the southwestern Shaanxi in the Neolithic age’, Dongfang Kaogu (东方考古), 11: 343–64. Wu, X., Xiao, H., Wei, C., Pan,Y., Huang,Y., Zhao, C., Xu, X., and Ogrinc, N. (2007) ‘The stable isotopic evidences of paleodiet of human and pigs, agriculture practices, and pig domestication at Xinzhai site, Henan (河南新砦遗址人、猪食物结构与农业形态和家猪驯养的稳定同位素证据)’, Science for Archaeology (科技考古), 2: 49–58. Xi’an Banpo Museum. (1975) The Matrilineal Clan Commune from the Cultural Remains of Banpo Type Yangshao Culture (从仰韶文化半坡类型文化遗存看母系氏族公社). Wenwu (文物) 12: 12:72–78.

59

Kate Pechenkina Xi’an Banpo Museum. (1978) ‘A Neolithic site at Shijia in Weinan county, Shaanxi province (陕西渭南史 家新石器时代遗址)’, Kaogu (考古), 1: 41–53. Xi’an Banpo Museum, Shaanxi Institute of Archaeology, and Lintong County Museum. (1988) Jiang Zhai: Report on the Excavation of the Neolithic at Jiangzhai by the Excavation Group Site (姜寨: 新石器时代遗址 发掘报告), Beijing: Cultural Relics Publishing House. Xiang, H., Gao, J., Yu, B., Zhou, H., Cai, D., Zhang, Y., Chen, X., Wang, X., Hofreiter, M., and Zhao, X. (2014) ‘Early Holocene chicken domestication in northern China’, Proceedings of the National Academy of Sciences of the U S A, 111: 17564–9. Xiao, J., Xu, Q., Nakamura, T.,Yang, X., Liang, W., and Inouchi,Y. (2004) ‘Holocene vegetation variation in the Daihai lake region of north-central China: a direct indication of the Asian monsoon climatic history’, Quaternary Science Reviews, 23: 1669–79. Yan W. 1992. Origins of agriculture and animal husbandry in China. In: Aikens CM, Rhee SN, editors. Pacific Northeast Asia in prehistory. Pullman: Washington State University Press. pp. 113–123. Yang, X., Wan, Z., Perry, L., Lu, H., Wang, Q., Zhao, C., Li, J., Xie, F.,Yu, J., Cui, T., Wang, T., Li, M., and Ge, Q. (2012) ‘Early millet use in northern China’, Proceedings of the National Academy of Sciences of the U S A, 109: 3726–30. Yu, Y. (1977) ‘Han’, in: K. C. Chang (ed.) Food in Chinese Culture: Anthropological and Historic Perspectives, New Haven:Yale University Press. Yuan, J. (2001) ‘The most recent studies of early Chinese domestication (中国新时期时代家畜起源的问 题)’, Wenwu (文物), 5: 51–8. Yuan, J., and An, J. (1997) ‘Two questions of Chinese zooarchaeology (中国动 物考古学研究的两个问 题)’, Zhongguo Wenwu Bao (中国文物报), 4.27: 3. Zhang, G., and Lu, L. (2004) ‘Archaeological study on civilization forming process in Haidai district: the chronology, character and social property of Neolithic culture in Lubei district’, Guanzi Xuekan, 1: 70–80. Zhang, X., and Guo,Y. (2003) ‘New discoveries of archaeological sites near Xi’an [西安再次发现大型史 前环壕聚落遗址] ‘, Zhongguo Wenwu Bao (中国文物报) 8.29: 1.). Zhang, X., Qiu, S., Bo, G., Wang, J., and Zhong, J. (2007) ‘Carbon 13 and nitrogen 15 analysis of human remains from Erlitou and Taosi sites (二里头遗址、陶寺遗址部分人骨碳十三、氮十五分析)’, Science for Archaeology (科技考古), 2: 41–8. Zhang, X., Zhao, X., and Cheng, L. (2011) ‘Human diets of Yangshao culture in the Central Plains’, Chinese Archaeology, 11: 188–96. Zhang, Z-P. (1985) ‘The social structure reflected in the Yuanjunmiao cemetery’, Journal of Anthropological Archaeology, 4: 19–33. Zhao, Z. (2011) ‘New archaeobotanic data for the study of the origins of agriculture in China’, Current Anthropology, 52: S295–S306. Zhao, Z., and He, N. (2006) ‘Results and analysis of 2002 soil sample flotation from Taosi (陶寺城址2002 年度浮选结果及分析)’, Kaogu (考古), 5: 77–90. Zhou, L. (2016) “From state to empire: human dietary change on the central plains of China from 770bc to 220 ad”, unpublished Ph.D. thesis, University of Alberta. Zhou, L., Garvie-Lok, S., Fan, W., and Chu, X. (2017) ‘Human diets during the social transition from territorial states to empire: stable isotope analysis of human and animal remains from 770 bce to 220 ce on the Central Plains of China’, Journal of Archaeological Science: Reports, 11: 211–23.

60

3 THE BRONZE AGE BEFORE THE ZHOU DYNASTY ROBERT BAGLEYTHE BRONZE AGE BEFORE THE ZHOU DYNASTY

Robert Bagley

Preliminaries: scope, aims, and sources In East Asia the earliest state-level societies that we know much about are those of the Yellow and Yangtze river valleys in the second millennium bc. In calling them states, we are diagnosing social organization from material remains: we take city walls, imposing building foundations, large-scale metal production, elite burials, and widely distributed artifact types to be material residues of highly stratified societies. Some of these features, city walls for example, have an earlier history, and a case for earlier states could be made.The third millennium Liangzhu culture of the Yangtze delta region is an obvious candidate. Toward the middle of the second millennium, however, the rise of a distinctive metal industry, and with it the characteristic artifacts of the Chinese Bronze Age, cast bronze bells and ritual vessels, was a new and consequential development. Writing may have been invented at about the same time, though we have no trace of first stages. The civilized societies to which metallurgy and writing direct our attention arose in the middle Yellow River valley, but by 1200 bc they had flourishing offspring throughout the Yangtze valley as well. These societies, whose achievements were inherited by the first millennium Zhou civilization, are our subject. The period of concern to us, roughly 1800–1000 bc, will for convenience be called the Early Bronze Age (EBA). For the earlier part of the period the dating of archaeological sites depends on radiocarbon measurements, which give absolute dates – calendar dates – but have uncertainties on the order of one or two centuries.Toward the end of the period we begin to rely on information taken from later texts. We try to fix the date of the Zhou conquest of Shang, an event that figures prominently in the texts, and then count generations backward from it.The conquest date endorsed by the state-sponsored Xia Shang Zhou Chronology Project is 1046 bc.1 “Ca. eleventh century bc” might be more realistic, but whichever we prefer, a date for the Zhou conquest is, in the material record, a date for an event at one city. It is a date for the supposedly punctual end of the Anyang settlement and hence for the end of the pottery sequence that archaeologists have constructed there. Other sites can be connected with the conquest date and the pottery sequence it terminates only by correlating their material culture with the material culture of Anyang. For an important tomb discovered a few decades ago at Xingan in Jiangxi, for example, neither radiocarbon nor written evidence is available. A date for the Xingan tomb can only be estimated by comparing its contents with the contents of Anyang tombs. As the best-dated and 61

Robert Bagley

best-explored EBA site, Anyang is our reference point: we judge other sites and artifacts to be earlier than, later than, or similar to Anyang and its artifacts.2 Absolute dates are indispensable for some purposes.They enable us to compare unrelated or widely separated cultures.We depend on them, for example, to say that chariots in the Caucasus pre-date chariots at Anyang. But they do not have the resolution to sort out developments within the EBA.They cannot put the tombs of a royal cemetery in order or align a newly discovered site with a level in the Anyang stratigraphic record. All of archaeology’s detailed reasoning rests either on artifact comparisons or, within a single physically continuous site, on stratigraphy. These give relative dates. The time of concern to us is the end of prehistory. At present the Chinese written record begins with the first Anyang oracle inscriptions, for which a date around 1200 bc has been shakily inferred from mentions of lunar eclipses in some of them. Because the oracle inscription corpus is small3 and restricted in provenance and content – one city mainly, and matters that its king divined about – we might suppose that the study of EBA states would not differ significantly from prehistoric archaeology. But this is far from being the case. Because the inscriptions mention kings’ names known from transmitted texts, the archaeology of the second millennium was from the start motivated and guided by late first millennium texts, and it has at times aspired to narrative history of a kind beyond the reach of the prehistorian. If this is now beginning to change, the reason is that the archaeological record has proved absorbing in itself. Our mental picture of the EBA is increasingly dominated by material culture, and scholars are increasingly preoccupied with matters that can be investigated through it – the process of state formation and the history of technology, to mention only two. The sources relevant to our subject are both material and written. All have biases of several kinds. The material record has both a preservation bias (the soft parts of history, we might say, do not fossilize) and a sample bias (accidents of discovery, agendas of archaeological exploration, the practicalities of salvage archaeology). Whether contemporary or later, the written sources too have a preservation bias: the bulk of what was written has not survived, and what does survive does so partly by accident, partly (in the case of transmitted texts known only in Han recensions) by the active intervention of editors and scribes. The written record also has innate biases: authors have reasons for writing; editors interpret. Any inscription or text is shaped by its author’s purposes, knowledge, and perspective.4 Our inferences should be informed by awareness of all these biases. The written sources for our period are second millennium inscriptions on oracle bones and bronzes and first millennium bronze inscriptions and transmitted texts. The oracle inscriptions, almost the only documents that survive from the EBA, have a very narrow bias. They see the world through the Anyang king’s eyes, and only that part of it that he divined about. Though they touch on many other matters, interactions with enemies and trading partners for instance, their principal concern is sacrifices to the king’s ancestors. Because the names of the people and places they mention can seldom be connected with archaeological finds and sites, they give us only a vague idea of the king’s view of the world and no idea at all of how his neighbors saw him. They do not answer our most basic questions about China in the last two centuries of the second millennium or even about everyday life at Anyang. What territory did the Anyang king rule? What other polities, large or small, near or distant, existed in his time? Did those neighbors view him as having special authority or standing, or was it later writers who for their own purposes represented him as a divinely sanctioned universal ruler? The Anyang king called himself “I, the one man.” How many of his contemporaries called him that? How many instead called themselves “the one man”? The answers scholars have given to questions like these have always owed less to second millennium evidence than to preconceptions absorbed from transmitted texts. 62

The Bronze Age before the Zhou dynasty

When we study a period for which both material and written evidence are available, we need a reasoned way of combining bodies of evidence that are sometimes complementary, sometimes contradictory, sometimes incommensurable. Ever since the birth of scientific archaeology in the nineteenth century, students of the biblical and Homeric worlds have wrestled with this problem without ever arriving at a clear set of rules.5 Archaeology can contradict texts, for instance by showing that at the time Joshua is supposed to have destroyed Jericho’s walls, ancient Jericho had no walls. But its ability to verify narratives of human action is limited. Proving that an event described in a text could have happened is not the same as proving that it did happen. If ruined walls of suitable date were discovered at Jericho, could archaeologists confirm that they were destroyed by Joshua? Most ancient walls are now in ruins. Hasty correlations between texts and archaeological finds foreclose options, blinding both reader and writer to alternatives. The naming of sites and cultures can be especially insidious. The identification of a certain deeply stratified mound in Anatolia as the site of an ancient city called Troy may rest on good evidence. The moment we call the site Homer’s Troy, however, we import a host of beliefs about it, for instance that the city was sacked in Mycenaean times by invaders from Greece, and including, all too often, assumptions that we are not conscious of making.6 Best practice accordingly keeps the two lines of investigation separate to the extent possible, with the double aim of keeping interpretative options open and making transparent the basis on which any conclusion rests.7 Clear epigraphic evidence justifies connecting the Anyang site with a ruling family that later texts call Shang. When a site does not yield such evidence, it is advisable to refer to its material culture by an arbitrary modern name.The Liangzhu culture, for example, is named after the modern village where its distinctive artifact assemblage was first found and described.8 The Anyang oracle inscriptions were discovered around 1900, and western scientific archaeology came to China in the 1920s. Before 1900, the only basis on which an ancient history for China could be constructed was texts transmitted from the Han period. A chapter title that invokes the Bronze Age rather than peoples or polities named in transmitted texts reflects a conviction that as a guide to times before the Zhou period, the texts are less trustworthy than the material evidence unearthed by archaeologists. But as the material evidence has been gathered and interpreted under the guidance of the texts, the two are not easy to separate.To suggest what archaeology conducted in ignorance of the texts might have told us is beyond the ambitions of the present chapter. Instead, by sketching the history of EBA archaeology, the chapter tries to show how a picture of the past derived from transmitted texts is giving way to a picture inferred from material remains – a picture less schematic, more complicated, and much more spacious. Archaeology in China began as a search for more oracle inscriptions and then as an exploration of the city whose kings produced the inscriptions. As it continued, however, it began to uncover peoples and places unmentioned in any text. In the process it has slowly come to be driven less by a textual agenda and more by questions arising from its own discoveries. The chapter concludes by examining some key points of contact between archaeological findings and the image of pre-Zhou times that prevailed before the advent of archaeology.

A short history of early Bronze Age archaeology The Anyang oracle inscriptions connect the traditional written record with material remains In 1898 or 1899, inscribed ox scapulas and turtle plastrons, material we now call the Anyang oracle bones, came to the attention of the Beijing antiquarian Wang Yirong (1845–1900), who recognized the writing on them as an archaic form of the Chinese script. Scholars immediately 63

Robert Bagley

began collecting and studying the bones, which at first were supplied to them by antique dealers. In 1908 Luo Zhenyu (1866–1940) discovered the dealers’ source.The bones were being dug up in northern Henan at a village on the Huan River near Anyang called Xiaotun. In an essay published two years later Luo noted the occurrence in the inscriptions of a dozen or so names that historians already knew from a king list given in Sima Qian’s Shiji (ca. 100 bc) and there ascribed to a dynasty called Shang or Yin. Taken together, Luo’s discoveries connected a dynasty previously known only from transmitted texts with tangible objects used and inscribed at its court. Most of the inscriptions are questions asked on behalf of the king about sacrifices to his predecessors. The king who asks the question is never named, but the kings who are to receive sacrifice are named and sometimes also addressed as “father/uncle” or “grandfather/ancestor.” By 1917 Wang Guowei (1877–1927) had reconstructed the genealogy of the kings from this information and had shown it to agree almost exactly with the Shang king list given by Sima Qian. The reconstructed genealogy contains twenty-nine kings. All but the twenty-ninth are mentioned in the inscriptions as recipients of sacrifice, but only the last nine ask questions. The oracle inscriptions therefore date from the reigns of the last nine Shang kings, and those nine kings lived at Anyang.9

Excavations at Anyang, 1928–37, 1950–present The oracle inscriptions made the Anyang site an obvious target for the first government-sponsored archaeological excavation in China. Work began in 1928 with a reconnaissance led by an oracle-bone scholar, Dong Zuobin (1895–1963). Systematic excavations began the next year under the direction of Li Ji (1896–1979), a Harvard-trained anthropologist, and continued until the Sino-Japanese war halted them in 1937. Work resumed in 1950 under a new government and goes on today. Archaeologists call the site Yinxu, the Waste of Yin, from an old name for the deserted capital, but in the oracle inscriptions it is called Da Yi Shang, Great City Shang.10 At Xiaotun, source of the oracle bones, Li Ji’s team found an area of palaces and temples. Little survived of the buildings besides rammed-earth foundations, but there were more than a hundred of these, many very large, some incorporating human sacrifices made during construction. In 1934, alerted by reports of tomb robbing across the river, the archaeologists found a cemetery of more than a dozen enormous shaft tombs.Though they had been stripped by looters ancient and modern, the tombs were surely royal. What little remained of their furnishings did not identify the occupants, but it was enough to connect artifacts of types long known to antiquarians, bronze ritual vessels above all, with the Shang period. At the royal cemetery the excavators found human sacrifice on a frightening scale. In and near the great shaft tombs were sacrifices of two kinds that we might distinguish as specific individuals or servants and anonymous human cattle. The former were victims buried in coffins or with grave goods of their own or with artifacts indicating the function they served in the tomb, for example guards with weapons or chariot drivers with their chariots. Of these a tomb might have several dozen. The anonymous victims were beheaded as the tomb was filled. Rows of headless bodies and rows of heads were laid out on stairs or ramps leading down into the tomb. Further victims were deposited in pits around the mouth of the tomb. Sacrifices in and near the tomb were made during the burial ceremony. Victims were also offered later. In 1934–35 Li Ji’s team excavated more than a thousand sacrificial pits laid out neatly in east-west rows in distinct groups, all the pits of a group containing only complete skeletons or headless skeletons or skulls, as though each group was a single sacrifice performed in a single way. In 1976 careful excavation of another 191 pits was able to distinguish twenty-two groups, the average group containing fifty victims, the largest consisting of forty-seven pits with 64

The Bronze Age before the Zhou dynasty

more than 339 victims. Mutilation of victims was common. Some were cut in half at the waist or cut into as many as ten pieces, and a few children were bound and buried alive. In the oracle inscriptions the king asks constantly about offerings to his ancestors, sometimes about offerings to several ancestors jointly. When the offering they required was human sacrifice, he evidently made it at the royal cemetery, even if the ancestor was buried elsewhere (the 339 victims in forty-seven pits were offered by the first Anyang king). Royal religion in the Shang period revolved around human sacrifice to the king’s ancestors. At Anyang Li Ji found a spectacular Bronze Age civilization that seemed to have no local precursor. It had a writing system and a bronze industry, both highly sophisticated, and it had features that seemed to point to western connections: horse-drawn chariots, knives and daggers of types native to the Eurasian steppe, and human sacrifice that called to mind Leonard Woolley’s discoveries a few years earlier at the Ur royal cemetery in Mesopotamia. Li’s western training had taught him that civilization had one birthplace, the ancient Near East, and that civilizations elsewhere were inspired by cultural contact, so he speculated that the Anyang civilization sprang into being full-grown when prehistoric cultures indigenous to East Asia received a fertilizing stimulus from outside.11 In the 1950s, when Anyang archaeology resumed after the founding of the People’s Republic, it did so in a new political and intellectual climate in which outside influence was no longer an acceptable explanatory mechanism. For several decades Chinese archaeology was to be confined both practically and imaginatively within the modern political boundaries of China. Within those boundaries it was immensely productive. The Yinxu site (meaning everything at Anyang that dates from the last nine Shang reigns) is still very unevenly explored, but a great deal is known about it. Remains are scattered over about 30 sq km. Besides the Xiaotun palace district and the royal cemetery across the river, excavations have revealed lesser settlements and cemeteries and workshops for pottery, stone tools, bronze, jade, and carved bone. Studies of pottery typology have divided the occupation of the site into four stages,Yinxu 1–4. Any tomb or stratum that contains pottery can be assigned to one of these stages. The first major postwar discovery at Anyang was an intact royal tomb found in 1976 not in the royal cemetery but at Xiaotun, near the palaces. Inscribed bronzes identify the occupant as Fu Hao, Lady Hao, a consort of Wu Ding, the first Anyang king. The immense wealth of her tomb, above all in jades and bronze vessels, came as a surprise to specialists. Spectacular bronzes of a sophistication that had been thought possible only at the end of the dynasty turned out to belong to the first Anyang reign. Clearly the bronze art had a long pre-Anyang history. Another startling discovery was made in 1999 across the river from Xiaotun at a place the excavators call Huanbei.Yinxu had no city wall. Huanbei is a very large city with an unfinished square outer wall 2 km on a side. An inner wall about 500 by 800 m encloses the remains of sixty buildings, two of them enormous. To judge from pottery typology and a few bronzes, Huanbei was occupied for only a brief time just before the Yinxu occupation. The buildings of its inner city burned down. The excavators believe that after a fire destroyed its palaces, Huanbei was abandoned and its inhabitants moved across the river to build the city at Yinxu. If this is correct, the king who oversaw the move may well have been Lady Hao’s husband Wu Ding, the first king whose presence at Anyang is attested by oracle inscriptions. Major construction at Yinxu, both in the palace district and at the royal cemetery, seems to have begun in his time. He also seems to have instituted the practice of carving inscriptions on divination bones and to have been the most enthusiastic diviner among the Anyang kings. More than half the known oracle inscription corpus comes from his reign (and more than half the human sacrificial victims). It is to Wu Ding, therefore, that we owe our first knowledge of a writing system that in his time was already fully developed, in the sense that a scribe at his court could 65

Robert Bagley

probably have written anything he wished. As in later periods, everyday writing at Anyang was done with brush and ink on slips of wood or bamboo, but while slips are depicted in the oraclebone character that means “document,” and brush-written ink inscriptions are known on a few durable objects (jades, potsherds), no actual slips survive from the second millennium bc. The earlier history of writing, the range of functions writing served at Anyang, and the geographical extent of literacy are correspondingly uncertain. The oracle inscriptions obviously draw at times on other written records, however, and inscribed oracle bones found at an Anyang colony in Shandong show that there were scribes there who could have communicated with scribes at Anyang in writing. It seems also to have been in Wu Ding’s time that bronze vessels began to be cast with inscriptions. The oracle inscriptions, the bulk of the Anyang epigraphic corpus, are written from the king’s point of view. Only the inscriptions on bronzes see the world through other eyes.Though the first examples are brief – little more than the name of the vessel’s owner or of the ancestor it was dedicated to – they confirm that the vessels were used by aristocrats for offering food and drink to their ancestors. Offerings were put into the tomb and presumably also, like the king’s sacrifices, made above ground at intervals after the funeral. Half a dozen vessels cast late in the dynasty have longer inscriptions announcing that the owner made the bronze to commemorate an award received from the king in acknowledgment of loyal service. The announcement, like the food or drink offered in the vessel, was no doubt addressed to the ancestor named at the end of the inscription, informing him that his descendant was doing his duty and maintaining the honor of the family. This inscription type, the written report to a superior of a transaction, was probably modelled on the administrative documents of a well-developed bureaucracy.12 The success of Li Ji’s excavations shaped the future of Chinese archaeology in two important ways. First, as Li recognized, the sophistication of the Anyang civilization posed a problem of origins. Li was willing to make stimulus from outside a part of the answer, but for a younger generation of scholars who were not, the need to find local antecedents was urgent. Second, the confirmation of Sima Qian’s list of Shang kings and the discovery of a Shang capital seemed to vindicate a whole tradition of classical learning. If the Shang dynasty was real, then the Xia dynasty must also be real, and its capitals too must be found.The Shang dynasty was confidently taken to be what Sima Qian believed it to be, one of a series of dynasties that each in turn ruled the whole of civilized China. This was an assumption with immediate implications for archaeological finds in places other than Anyang.

A local antecedent for the Anyang civilization: excavations at Zhengzhou, 1950-present Zhengzhou, the capital of Henan province, is located in the middle Yellow River valley 160 km south of Anyang. In 1950 a find of potsherds there drew the attention of archaeologists to a mound called Erligang. In 1952–53 they excavated the mound and distinguished two occupation levels, Lower and Upper Erligang, which are now thought to belong to the fifteenth and fourteenth centuries bc. Typological study suggested that Erligang pottery was related to Yinxu pottery, but earlier. In 1955 four modest graves containing bronze vessels were excavated, and the bronzes too are predecessors of Anyang types. In the same year investigation of a rammedearth city wall 500 m from the mound showed the wall to be a Lower Erligang construction: it contains Lower Erligang potsherds and its sloping base is overlapped by deposits containing Upper Erligang sherds. In the rammed-earth technique, invented in Neolithic times and used for walls and foundations throughout the Bronze Age, a thin layer of earth is poured between wooden forms and hammered until it rings, the operation being repeated until the desired 66

The Bronze Age before the Zhou dynasty

height is attained. City walls built in this way are major constructions. The Zhengzhou wall is 22 m thick at the base and in places survives to a height of 9 m. It is 7 km long and encloses an area of about 400 hectares (100 hectares = 1 sq km). The outer wall at Huanbei encloses 470 hectares. Zhengzhou and Huanbei are the two largest walled cities yet known from the EBA. Though the modern city of Zhengzhou sits atop the ancient one, making systematic excavation impossible, twenty rammed-earth building foundations are known inside the ancient wall, and workshops, habitation areas, cemeteries, and a second stretch of wall have been found outside it. Human sacrifice is in evidence, though not on the scale of the Anyang royal cemetery. No burial of an Erligang king has been found.Three major caches of bronze vessels have turned up at intervals over the last half-century but only a few bronze-bearing tombs, none approaching royal scale. The importance of the ancient city is declared chiefly by its size. Though it was surely the major city of the Erligang civilization, its material culture is better known from finds made elsewhere.13 To Li Ji the Chinese Bronze Age seemed to begin so abruptly that outside stimulus was required to explain it. The Erligang finds dispelled the appearance of abruptness by supplying local antecedents for Anyang building technology, human sacrifice, bronze casting, and burial forms, though not for the chariot or the Chinese writing system. The only writing yet found at Zhengzhou is ten graphs on a divination bone fragment that may be no earlier than the Anyang oracle bones (its archaeological context is unclear).There is good reason to believe that the writing system was an Erligang invention, however, and in the 1990s a few potsherds, each bearing a brush-written character or two, were found at Xiaoshuangqiao, a site 20 km from Zhengzhou that may slightly postdate Upper Erligang. As for the chariot, Li Ji was surely right to connect it with the steppe.14 The Erligang finds were a clear demonstration that the Anyang civilization had a sophisticated predecessor. This established, archaeologists sought to give the city at Zhengzhou a historical identity, that is, to fit it into the received history that was taken to have been verified by the royal names in the Anyang oracle inscriptions. Since Anyang was the last capital of the Shang dynasty, Zhengzhou must be an earlier capital. But which? Received texts name several. Counting on pottery typology to help them choose, but with no inscription to confirm an identification, scholars have never been able to agree. They have also never considered the possibility that Zhengzhou was not Shang at all. Certainly it is possible that the city was ruled by pre-Anyang Shang kings whose names we know from the Shang king list, but it is also possible that the early Shang kings lived elsewhere and that they were rivals of the kings of Erligang. Or perhaps the Shang dynasty was founded by a courtier of the Erligang king who usurped the throne and by way of justification invented the Shang king list – a list of ancestors who were not kings, who might even have been fictions, perhaps gods claimed as ultimate ancestors of the new royal house. These speculations are not idle. They highlight the hidden dangers of committing ourselves to a storyline that does not emerge from archaeology. When we declare a site to be an early Shang capital, we declare that it was ruled by kings whose descendants at some point moved to Anyang.This seemingly innocent statement inserts archaeological sites into a narrative of successive capitals in which the rise of one city coincides with the decline or abandonment of another and in which the Shang kings had no rival capable of building a major city. It assumes continuity from the uppermost stratum at one site to the lowermost at another, and it imposes uncontested universal rule on an archaeological record that shows nothing of the kind. Archaeologists at first supposed that Zhengzhou was abandoned after Upper Erligang, its population having moved to Anyang. As finds accumulated, however, it became increasingly clear that Upper Erligang was not continuous with early Yinxu. Bronze vessels intermediate between Erligang and Yinxu types have been found at Zhengzhou and elsewhere. Eventually the 67

Robert Bagley

difficulty was dealt with by defining a “transition period” (this term is always a sign of confusion) to which Xiaoshuangqiao and Huanbei were later assigned. Zhengzhou, Xiaoshuangqiao, and Huanbei are today taken to form a sequence of Shang capitals.Yet pottery typology cannot establish the family names of the rulers of these places. In the absence of epigraphic evidence we should forgo political identifications of sites – identifications with names found only in far later texts – and speak in strictly archaeological terms of cultures defined by material artifacts. To do otherwise is to remove the ambiguity of the archaeological record by imposing a Han picture of the past on it.

Bronze Age beginnings: excavations at Yanshi Erlitou, 1959-present In 1959 an archaeological survey searching for capitals of the Xia dynasty began exploring a site near Luoyang called Erlitou, which had yielded pottery of a type known also from early strata at Zhengzhou 85 km to the east. Excavations that continue to the present have revealed the largest settlement yet known from its time, about 1800 to 1500 bc according to current radiocarbon evidence. Pottery typology distinguishes four phases. By the end of the first, the settlement already covered 100 hectares, but metallurgy did not yet go beyond the manufacture of small implements such as knives. From this phase comes the only pre-Anyang evidence for a wheeled vehicle in China, wheel tracks a meter apart, half the axle length of an Anyang chariot. There is as yet no evidence for domesticated horses before Anyang times. During the second phase the settlement grew to 300 hectares. Remains include building compounds with courtyards and, within the courtyards, rich (royal?) burials containing jades, small bronze clapper‑bells, cowry shells, and objects inlaid with turquoise. All or most of these luxury materials were obtained by long‑distance trade. A bronze ax likely to be an import from the northern steppe belongs to this phase, as does a large turquoise workshop. A bronze knife of northern type comes from the third phase, during which additional large buildings surrounded by a rammed-earth wall were constructed. The cast bronze vessels that appear in graves of this stage are primitive ancestors of Zhengzhou and Anyang ritual vessels in shape and probably also in function. A few more vessels have been found in the fourth phase, but despite foundry remains said to cover one hectare, the current total from Erlitou is only eighteen, thirteen of them tripod cups of the type jue. Puzzlingly, a new city with palatial buildings and a wall around them was built just 6 km from Erlitou during the fourth phase. The excavators call this Yanshi Shangcheng, the Shang city at Yanshi, because its pottery resembles pottery from the Zhengzhou site. Some scholars take it to be the capital of the founder of the Shang dynasty, others interpret it as a fortress planted in the newly conquered Xia heartland by a king who resided at Zhengzhou. After phase four the Erlitou settlement shrank, and by the end of the Erligang period it had been abandoned.15 Pottery more or less resembling that of Erlitou has been found throughout the middle reaches of the Yellow River valley. In the sites from which it comes some archaeologists detect a three- or four-tiered settlement hierarchy that they take to signify a unified polity. Outside that area Erlitou artifacts and artifact types reached places as remote as Sichuan in the west (the Sanxingdui site), Liaoning in the northeast (the Lower Xiajiadian culture), and Wuhan in the middle Yangtze region (the Panlongcheng site). Such artifacts are likely to have spread by trade or exchange, followed sometimes by local imitation, but some kind of colonization might in some cases have played a role. In material culture Erlitou is clearly a precursor of Erligang. Its excavators at first called Erlitou “Early Shang” and Erligang “Middle Shang,” but these identifications did not long go undisputed. For a time some observers argued that Erlitou was a Xia capital, others that the first 68

The Bronze Age before the Zhou dynasty

two strata are Xia and the last two Shang. It is now universally regarded as Xia – it is “the capital of China’s first dynasty” – because radiocarbon dates have shown it to be too old to fall within the six centuries that tradition assigns to the Shang dynasty. Erligang has returned to being “Early Shang,” and “Middle Shang” now means the Huanbei site. These political identifications shift back and forth because none of them has any support from epigraphic evidence. A more obvious illustration of the methodological urgency of keeping texts and archaeology separate would be hard to find.16 Calling the Erlitou culture “Xia” attaches to it a set of preconceptions about the early second millennium political order in north China. We are invited to think of the formation of an Erlitou state as a process that brought a disordered Neolithic China under the unified rule of the first in a series of paramount dynasties. This amounts to adopting a center-periphery model with an active king at the center and passive subject peoples on the margins.The material record is mute, scanty, and ambiguous enough for such an interpretation to be easily read into it, and once it has been, the narrative has an appeal that makes discrepancies easy to overlook. But the credibility of the model has depended chiefly on ignorance of the periphery, an ignorance that archaeology in recent decades has begun to remedy.

Erligang was an empire: Panlongcheng, excavated in 1963 and 1974 In 1963 and 1974, alerted by repeated chance finds of potsherds, metal artifacts, and sections of an ancient wall, archaeologists excavated portions of a site called Panlongcheng near modern Wuhan in the middle Yangtze region. They found a small city whose elite material culture – a rammed-earth city wall 1 km long enclosing 7 hectares, foundations of three large buildings inside the wall, and thirty-eight tombs outside – is indistinguishable from that of the Zhengzhou site 450 km to the north. The richest of the Panlongcheng tombs were identical to Zhengzhou burials in form but larger and more lavishly furnished, containing bronzes and jades of the highest quality. They reveal that the Erlitou inventory of four or five bronze vessel types had by the end of Erligang grown to more than twenty. One of the tombs, the richest Erligang burial yet known, contained three sacrificial victims, twenty-three bronze vessels, forty bronze weapons and tools, jades, pottery, and glazed stoneware. It is significant that, unlike the bronzes and other luxury items, the pottery found at Panlongcheng is not all of Erligang type; some of it is local. The combination of Erligang elite culture and local pottery argues for an intrusive elite ruling indigenous commoners, in other words, for colonization or conquest by Zhengzhou. Like invaders of other times and places, the Erligang settlers brought with them all the experts they needed, from builders to bronze casters, to replicate their lives back home. Panlongcheng is the first place outside the Yellow River valley where Erligang remains were recognized and the first hint of what is today known as the Erligang expansion. More than sixty sites with Erligang artifacts are now on record, and while about twenty cluster within 100 km of Zhengzhou, the remainder are scattered west into the Wei valley, east and northeast to Shandong, Hebei, and Beijing, southeast to Anhui, and south to Panlongcheng and other places in Hubei. Most of these sites have not been excavated, and at most of them we can only guess what motivated the Erligang presence. A thirst for exotic raw materials is a regular symptom of the rise of civilization, however, and in the south the most obvious resource is metal. The Yangtze region has rich copper deposits, several of which were mined in antiquity, and it is possible that Panlongcheng and other Erligang settlements in Hubei and southern Anhui were way-stations in a network that Zhengzhou depended on for copper. 69

Robert Bagley

The Erligang expansion seems to be a phenomenon of the Upper Erligang period. It was sudden and short-lived, lasting perhaps only a few generations, but its impact was colossal. It disseminated the new civilization of the middle Yellow River valley to an enormous area of north and central China. It was, as one scholar puts it, the mechanism by which the early Bronze Age changed from a local to a regional phenomenon.17 As indigenous societies reacted to the encounter, new powers arose on the Erligang frontiers. Perhaps it was pressure from them that caused Erligang to retreat or collapse. The lands it briefly controlled were afterwards home to not one but many bronze-using cultures. At its peak the Erligang empire must have occupied a larger territory than the Anyang kings ruled.Western Shandong may be the only Erligang conquest that was also an Anyang possession. At the Shandong sites of Jinan Daxinzhuang, Teng Xian Qianzhangda, and Yidu Sufutun, the culture of Anyang – tomb forms, human sacrifice, bronze vessels, and inscribed oracle bones – was closely replicated.18 In the south, by contrast, the Anyang kings had culturally distinct rivals throughout the Yangtze valley. The major, slowly unfolding revelation of the last half century of EBA archaeology is the impact of the Erligang expansion in the middle and lower Yangtze region.19

Civilized aliens in the Sichuan Basin: the Sanxingdui pits discovered in 1986 In 1980 archaeologists began investigating an ancient city wall at Sanxingdui, a village 40 km north of Chengdu in Sichuan. Measuring 40 m thick at the base and enclosing 350 hectares, the wall proved to be roughly contemporary with the Zhengzhou city wall. An Erligang‑period city wall a thousand kilometers from Zhengzhou, in a region universally assumed to have been a cultural backwater until much later periods, was an astonishing discovery. But in 1986, while the archaeologists were at work inside the walled city, brickyard workers outside came upon two pit deposits that were still more astonishing. The pits, 30 m apart, date from the twelfth century bc. Close in time but somewhat different in contents, they are most easily interpreted as sacrifices of some sort, but both the artifacts themselves and the manner of their burial are very strange. Pit 1, slightly earlier than Pit 2, contained cowry shells, thirteen elephant tusks, 300 objects of bronze, jade, and gold, and three cubic meters of burnt animal bones and wood ash. Since the contents of the pit all showed signs of burning while the pit itself did not, the deposit looks like the product of a ceremony in which animals were sacrificed, bronzes and jades deliberately broken, and everything then burned and buried. The ceremony has no close parallel at Zhengzhou or Anyang, and many of the artifacts are of types never seen before, including life-sized bronze heads with facial features that look distinctly extraterrestrial. Pit 2 was much richer. Its contents were found in three layers: a hundred jades and other small items at the bottom, large bronzes in the middle layer, and sixty elephant tusks on top. Among the bronzes are forty-one heads and a life-sized statue on a pedestal. Perhaps the heads were fitted onto wooden bodies (dressed in silk robes?) to make statues like the bronze one. At Erligang and Anyang, ritual centered on sets of functionally distinct bronze vessels, but in Pit 2 the only bronze vessels were a dozen of a single vase-like type called zun or lei. Moreover, unlike the bronze heads, whose clay core material confirms that they were locally cast, the vessels are obvious imports, most from the middle Yangtze region. They attest to trade between Sanxingdui and its neighbors downriver, and they also help secure the twelfth century date of the pits. But the most extraordinary objects in Pit 2 were the fragments of three bronze trees, the largest of which has been restored and stands 4 m high. Birds perch on its flowering branches; other small bronzes and jades found in the pit may have been attached to it. If the 70

The Bronze Age before the Zhou dynasty

ritual in Pit 2 involved the sacrifice of bronze trees, perhaps the wood ash in Pit 1 was from the sacrifice of real ones. In the Anyang oracle inscriptions Wu Ding divines about attacking a people or place written with a graph that may be related to the modern character shu, an old name for Sichuan province. The first report on the Sanxingdui pits accordingly suggested that Wu Ding was meditating an attack on the city at Sanxingdui 1100 km to the southwest. Equating Sanxingdui with oraclebone Shu has the appeal that it incorporates a startling discovery at a remote location into an Anyang-centered picture of EBA China. Even if the equation were correct, however, the fact would remain that received texts left us wholly unprepared for a civilized city with a wildly distinctive material culture in twelfth century Sichuan. The discovery of Sanxingdui has made it impossible any longer to doubt that the bronze-using civilization born before 1500 bc in the middle Yellow River valley had by 1200 bc inspired local developments over a large region. It teaches us also that the textual record is never so misleading as when it is silent: its fictions are less dangerous than its omissions.The world of the texts is far smaller and far less varied than the world revealed by archaeology.20

The Xingan tomb (1989): bronze-using cultures arise beyond the Erligang frontiers In 1989 a rich tomb was found at a place called Xingan in central Jiangxi, 300 km farther south than Panlongcheng. Pottery connects the tomb with a small walled settlement 20 km away at Qingjiang (now Zhangshu) Wucheng, excavated in 1973. The Wucheng site yielded glazed stoneware, a specialty of the lower Yangtze region that was traded to the north, as well as a few pots with strings of incised signs that look like writing. As for the tomb, despite its location well south of the Yangtze it is the second richest EBA burial yet found, second only to the tomb of the Anyang royal consort Fu Hao. To judge by their contents the two tombs are close in date, about 1200 bc, but their occupants had very different ideas about funerary ritual. Fu Hao’s tomb contained 195 bronze vessels, 273 smaller bronzes, 756 jades, and 11 pieces of pottery. The Xingan tomb contained 48 bronze vessels, 4 large bronze bells, over 400 bronze tools and weapons, 150 jades, and 356 pieces of pottery. The numbers testify that Fu Hao was wealthier than the Xingan tomb’s occupant but also that pottery had more prestige in the south than in the north. Most revealing is the inventory of bronze vessel types. Fully 105 of Fu Hao’s vessels are wine containers of the types jia, jue, and gu. These three types had been indispensable components of northern ritual since Erligang times, but in the Xingan tomb they do not appear at all. Of its forty-eight vessels, thirty-five are tripods for cooking food (ding and li). This difference alone is enough to establish that the tomb’s occupant was not a northerner. Unlike Panlongcheng, Wucheng was not a southern outpost of northern culture. The point is underlined by the four bells in the tomb. They are larger and finer than any bell of similar date in the north, and as we will see in the next section, they connect Wucheng/Xingan with bronze-using cultures elsewhere in the middle and lower Yangtze region. The Xingan tomb is most important for the light its bronze vessels shed on the rise of civilization in the south. They have a clear time spread. A small number are standard Erligang types, giving us a starting date and source for the Xingan bronze industry. Most are a bit later, however, and they show Erligang types modified to suit local taste by the addition of such odd features as little tigers standing atop handles and surface patterns copied from local pottery.The latest of the bronzes are contemporary with Fu Hao’s tomb, and a few of them are obvious imports, suggesting that if contact with the north was disrupted at the time of the Erligang retreat, it had been resumed. But perhaps the most remarkable thing about the Xingan bronzes is the high quality 71

Robert Bagley

of the local castings. The best of them are both technically and artistically the equal of anything made in the north. The most efficient form of technology transfer is the transfer of experts, and Xingan must surely have acquired its bronze technology in the form of skilled casters from an Erligang settlement like Panlongcheng.21 Metal technology is the part of the Erligang impact that has survived best in the material record, but a southern ruler who used Erligang bronze vessels had also adopted something of Erligang ritual and could have adopted much else, writing for example. We must not suppose, however, that before its encounter with Erligang the Wucheng/Xingan culture was backward or primitive. Its ceramics prove otherwise, as do its ability and eagerness to adopt Erligang metal technology. Its embrace of Erligang culture was selective, however, as the Xingan inventory of vessel types attests, and the vessels themselves show that local taste quickly made itself felt in design. The same scenario must have repeated itself many times throughout the middle and lower Yangtze region. Contact with Anyang, occasional or continuous, did not prevent civilizations in the south from following very different trajectories of development.22

The Yangtze region: unplanned archaeology At major sites like Anyang and Erlitou, archaeology is planned and ongoing. The first Anyang excavators chose a site they already knew to be the capital of nine Shang kings. Erlitou was chosen for investigation by an archaeological survey seeking Xia capitals. But most archaeology in China today is unplanned. Often it is salvage archaeology aimed at learning as much as possible from a site found by a construction crew and about to be destroyed. Sometimes, as in the case of the Sanxingdui pits, an accidental discovery becomes the focus of ongoing excavation. But sometimes construction has destroyed a tomb before archaeologists are called in, and they must collect whatever objects and information they can from people who were present at the destruction. And sometimes we have only an artifact – an ancient bronze recognized by an alert worker at a metal recycling station, for example, or something looted from an unknown site and found on the art market.23 In all these forms unplanned archaeology is far from ideal, but it has been hugely important, for two reasons: it occurs on a vast scale, and it occurs all over China. It was chance finds accumulating over decades that gradually revealed civilized bronze-using societies in the south contemporary with the Anyang kings. Their existence had never been suspected by historians because transmitted texts do not mention them. Over the past century unusual bronzes have turned up at many places in the middle and lower Yangtze region. Among them are celebrated items in museum collections. From Hunan province, for example, come an elephant in the Musée Guimet in Paris, a boar in the Hunan Provincial Museum, a drum and a tiger-shaped you in the Sumitomo Collection in Kyoto, and a zun whose shape incorporates a quartet of rams in the National Museum in Beijing. All five are large and finely cast objects of extravagantly inventive design, and most depict animals very naturalistically. Vessels in animal shape are found less often downriver from Hunan, but one artifact type unites the entire region from Hunan to Zhejiang. This is the bronze nao, a clapperless bell mounted mouth-upward on a hollow stem and struck on the outside with a mallet or pole. Bells are more common and more imposing than bronze vessels throughout the middle and lower Yangtze region. Moreover, southerners were forming tuned sets of nao, and therefore using bells not for signalling purposes but for music, at least as early as the twelfth century. Bells rather than vessels must have been the dominant apparatus of southern ritual, a complete reversal of northern priorities. Nao are occasionally found in Anyang tombs, but at Anyang they are small, rare, and perfunctory in decoration. The only bells found in the Anyang tomb of Fu Hao are trifles by comparison with southern bells, and they are trifles also by comparison with her 72

The Bronze Age before the Zhou dynasty

ritual vessels. They cannot have been ritual necessities for her. The central importance of large nao to cultures in the middle and lower Yangtze region defines a vast province sharply distinct from the civilization of the north, and distinct also from Sanxingdui, which has only yielded a few sets of small clapper bells.24 Xingan and two modest Zhejiang finds are the only instances yet known in which southern bells have been found in burials or in company with bronze vessels. In Hunan, at least, nao are typically found buried in pits, by themselves, and they are often enormous. The largest so far known weighs 220 kg. Casting such a bell was not a small undertaking. Massive bells and tuned sets were the products of cities, not villages, but the cities have yet to be found. The occupant of the Xingan tomb probably came from the nearby Wucheng settlement, but his three nao show none of the local features of his other bronzes and may be imports from elsewhere in the lower Yangtze region. Lacking archaeological context, not to mention written evidence, we have no information about the southern societies that produced and used these bells. For many scholars this has made it necessary to attribute them to the north. Though no large nao has ever been unearthed at a northern site, in a 1972 article the director of the national Institute of Archaeology in Beijing explained nao unearthed in Hunan as possessions taken there by northern refugees, Anyang aristocrats fleeing the Zhou conquest.25 Two decades later the Xingan bronzes likewise were explained by a few scholars as northern imports despite their obvious connection with local pottery. The instinct to attach archaeological finds to a text-based picture of an Anyang-centered world is abetted by the careless habit of applying the label “Shang” – the name of a ruling family attested only at the Anyang site – to sites and artifacts all over China. Without so much as a name to attach to a southern site, the testimony of mute artifacts can be hard to accept. But material evidence, however haphazardly acquired, has revealed whole civilizations missing from the written record. By the twelfth century bc societies distinct from Anyang in culture but comparable in sophistication existed all the way from Sanxingdui to the sea.26 Students of early civilizations have often contrasted a “Mesopotamia model” (a land of many independent polities occasionally united by short-lived empires) with an “Egypt model” (a land normally under one rule with occasional short periods of disunity). Current opinion holds that Egypt is probably the only real instance of the Egyptian model, probably because the topography of the Nile valley below the cataracts uniquely favors political unity.27 The ancient China described in received texts, a land in which legitimate rule passed from Xia to Shang to Zhou without hiatus for more than a millennium, would be the ultimate instance of the Egyptian model if it were not fiction.The archaeological record shows China instead to have been a Mesopotamia of competing states united briefly by the Erligang empire and the early Zhou empire. But a history resembling Mesopotamia’s had little appeal for Warring States authors anxious to restore the perfect government of early Zhou. Their purposes were better served by a past in which unity was the norm and the disunity of their own time was the exception.

Predynastic Zhou: statesmen or barbarians? Traditional history says that the Anyang kings were overthrown by invaders from the Wei River valley. For Warring States and Han writers, the Zhou conquest was an event of magical significance (depending on the writer, a cataclysmic battle or a bloodless transfer of allegiance): it was the moment when Heaven shifted its support from the degenerate last Shang king to the virtuous first Zhou king. The archaeological record shows no trace of magic, but it does bear the imprint of the reasonably swift creation of a Zhou empire that incorporated former Shang territory. Impressive bronze vessels of consistent design are found all the way across north China, distributed over an area much larger than the Anyang kings had ruled. Many of them moreover 73

Robert Bagley

have finely written inscriptions mentioning honors, posts, or lands granted to Zhou lords by their king. A bronze cast by an officer named Li, for example, commemorates a gift bestowed on him by the first Zhou king in recognition of service rendered at the time of a Shang defeat. Li dates his inscription eight days after “King Wu attacked Shang” and concludes it by dedicating the vessel (and thus addressing its inscription) to the ancestor who will receive offerings of food in it. Such inscriptions have late Anyang precedents, as do the awards they commemorate, but the Zhou inscriptions are longer, and we have many more of them. Evidently the ceremony of award and/or appointment to office was a fixture of the Zhou king’s transactions with his subordinates. Perhaps it was even the inspiration for the mandate of Heaven ideology, according to which Heaven appoints the king to office. It is in bronze inscriptions of the third Zhou king that we first hear mention of Heaven’s mandate. The contexts are too limited to tell us whether the expression meant anything close to what Warring States writers understood it to mean, but it clearly had something to do with the king’s claim to the throne. In the pre-conquest Wei River valley we find no precedent for early Zhou inscribed bronzes. The precedents, in everything from calligraphy to court ceremony, and from ancestor ritual to casting technique, are at Anyang. Transmitted texts portray the founders of Zhou as wise statesmen, but the material record puts their forbears among the more backward of Anyang’s neighbors, in no way comparable to Sanxingdui, for example. An assessment written twentyfive years ago entitled “Statesmen or Barbarians?” came down firmly on the side of barbarians.28 The pre-conquest Wei valley acquired weapons, other small bronze items, and occasional bronze vessels from the Yangtze region and from nomadic neighbors to the north.29 From Anyang it acquired only two or three of the simplest and dreariest bronze vessel types (food vessels, not wine vessels). Soon after the conquest, however, the Zhou cast technically impressive inscribed bronzes in the full range of Anyang vessel types, a change that suggests not only the acquisition of Shang technical expertise but also the adoption of Shang rituals. Technically virtuosic, beautifully inscribed bronzes, the officer Li’s for example, argue that the Zhou transplanted Anyang founders and scribes to their homeland. No doubt they appropriated much more of the Anyang civilization. Their claim to inherit legitimacy from Shang suggests as much.Yet it was from the south that they obtained tuned sets of bells, and the adoption of bell music into Zhou ritual was a radical departure from Anyang ritual. Communication with the middle Yangtze region, probably along the Han River, is apparent before the conquest, but on the evidence of recent finds at Suizhou in Hubei it intensified immediately after. Traffic also went due south to Sanxingdui and the Sichuan Basin. Zhou connections with the south are very inadequately known, but they seem to have gone deep and to have been important for a very long time.

Some questions Let us conclude with a few questions that the discoveries of recent decades have raised or made more pressing.

Writing and literacy: functions and extent Though small and narrowly focused, the oracle inscription corpus exhibits a fully developed writing system that must have had an earlier history and wider uses. The evident completeness of the system, in the sense that an Anyang scribe could probably write anything he could say, implies a prior development in which a simple notational system invented to serve some restricted function (Erligang bookkeeping?) gradually expanded its linguistic capabilities as it 74

The Bronze Age before the Zhou dynasty

spread to new functions.The oracle inscriptions also testify abundantly that writing was not confined to the diviner’s office, for besides summarizing more detailed records of divinations they often draw on other records that must have been kept in writing – bookkeeping that tracked such things as incoming raw materials, enemy dead and captured, and agricultural lands.30 Nevertheless, the almost complete absence of earlier inscriptions and of writing from contexts other than divination and places other than Anyang leaves many questions unanswered. How many cities besides Anyang had writing? Did Erligang writing spread to all the places that Erligang bronze technology did? Inscribed oracle bones have been found at what looks like an Anyang colony in Shandong and in a late Shang or early Zhou hoard at Fengchu in the Wei valley. But it is easy to imagine Anyang functionaries communicating in writing with counterparts in many other places, or with their own agents in those places. Were there cities where languages other than Chinese were written? Was writing in use at Sanxingdui? Do the strings of signs on a few Wucheng pots belong to a writing system different from the one we know at Anyang? There is cultural variation in the material record of the late second millennium that we do not see in the written record. Chinese speakers have a monopoly of the known written record. The Egyptian script was hardly ever used to write any language but Egyptian, but Mesopotamian cuneiform wrote at least seventeen languages from several different language families. Did the Chinese language and script always have the monopoly that they enjoyed in the Han period, or was earlier diversity gradually eliminated by a dominant culture? The latter process certainly operated in the construction of ancient history.

Human sacrifice The sacrifice of people like cattle to a king list presents several problems.The first is its rationale in its own time. Royal religion normally serves a legitimating ideology. In Mesoamerica, where human victims were sacrificed on a scale that exceeded even the Anyang king’s, enough is known about Aztec cosmology to suggest how the king’s subjects understood sacrifice to sustain Aztec society. But Anyang sources give no clue to the beliefs of Anyang viewers, and later sources do not even mention the victims at the royal cemetery. Sacrifice may have helped advertise the king’s monopoly of violence at a time when the populace was still being habituated to coercive authority, but it must have been justified to them in other terms. A second problem posed by these sacrifices is their disappearance from memory. The Zhou did not copy them, as far as we know, but they cannot have been unaware of them. Early Spanish accounts give some idea of the terror Aztec sacrifice inspired. Sacrifice on the scale of the royal cemetery must have been ever‑present in the consciousness of Anyang’s neighbors, who probably supplied the victims. Yet it is not mentioned in any Zhou source. Nowhere do the Zhou express disapproval of it. When Zhou bronze inscriptions and transmitted texts that purport to be early Zhou explain the Shang dynasty’s loss of Heaven’s mandate, they reproach it with drunkenness, not human sacrifice. The inscription of the Da Yu ding, a bronze cast in the reign of the third Zhou king, seems to say that Yin (i.e. Shang) lost the mandate because its vassals and senior officers “became lax through wine-drinking. Therefore,Yin failed in discipline among its officials.”31 Perhaps the early Zhou do not mention human sacrifice because they did not see it as a reason for the Anyang king to lose the mandate. And perhaps references to excessive drinking allude to a debate over something that they did connect with legitimation, a debate between a party that favored adhering to pre‑conquest Zhou ritual offerings, which centered on food, and a party that favored wholesale adoption of Shang offerings, in which wine had a larger place. If so, ongoing change in the repertoire of bronze vessel types suggests that the debate continued 75

Robert Bagley

for much of the Western Zhou period. Meanwhile, Anyang human sacrifice fell from memory. Late Zhou and Han writers, including those who profess to recount the evil deeds of the last Anyang kings, are clearly unaware of it. Confucius, who deplored human sacrifice, assumed that the Shang kings were paragons of virtue but said that he followed the institutions of early Zhou because he did not know enough about Shang. His estimate of his knowledge was correct.

The south The Anyang-period Yangtze region is urgently in need of study. Anyang bronzes found in the south and southern bronzes found in the Wei valley give us glimpses of contact between north and south, but if the Yangtze region was the north’s main supplier of metals, as seems likely, the bronzes were incidental to an enormous trade in raw material and whatever the north exchanged for it. Traffic up and down the Yangtze must have been equally busy. A shared preoccupation with bells throughout the middle and lower Yangtze region argues for significant interaction among otherwise diverse local cultures. The Sanxingdui civilization of the upper Yangtze was sharply different from cultures downriver, but it did import bronzes from the middle Yangtze region, not bells but vessels of one particular type. Because our knowledge of the second millennium south is almost wholly dependent on chance finds, it is tantalizingly sketchy, but the sketch shows a landscape crisscrossed by interregional traffic spottily but durably recorded by bronze artifacts. Trading stations must have been everywhere. The Anyang monopoly of surviving documents should not lull us into supposing that Anyang had a monopoly of anything else. The south in the Western Zhou period is also poorly known. It clearly had relations of trade or exchange with the north, for it supplied early Zhou courts with sets of bells, no doubt accompanied by musicians who knew how to play them and the music the musicians knew.The evidence for a Zhou presence in the south is increasing but not always easy to interpret. Two hoards of spectacular early Zhou bronzes found decades ago in Sichuan were only a few kilometers from the Sanxingdui site. More recently two early Zhou cemeteries 25 km apart have been found at Suizhou in northern Hubei, one with inscribed bronze vessels naming marquises of Zeng and one with vessels naming marquises of E. The Zeng bronzes are mostly standard early Zhou types, but some of the E bronzes are bizarre local reinterpretations of Zhou types. Should we imagine a mixture of Zhou colonies and local polities in this region? And how much continuity was there between second millennium southern cultures – the offspring of Erligang – and the southern states that figure prominently in the Warring States textual tradition? Were such states as Chu,Wu, and Yue local developments from second millennium predecessors or Western Zhou colonies or a combination of the two? In transmitted texts, states invariably originate as vassals enfeoffed by the early Zhou kings, but this all-too-familiar narrative of active center and passive periphery is at best an oversimplification. The vassals were not planted in empty lands. Finally, why are the early civilizations of the Yangtze region missing from the transmitted texts? How did they disappear from memory? Did a northern ideology of cultural and dynastic legitimacy require writing the civilized south out of history and discarding texts that mentioned it or originated in it?

King lists and royal legitimation The consensus that second millennium archaeology has validated first millennium accounts of Xia and Shang rests solely on the agreement of Sima Qian’s Shang king list with the list

76

The Bronze Age before the Zhou dynasty

Figure 3.1 Stone relief in the Temple of Sety at Abydos, ca.1300 bc, showing Sety I and his son Prince Ramesse revering a list of their predecessors Source: Auguste Mariette, Abydos, description des fouilles exécutées sur l’emplacement de cette ville . . . (Paris: Imprimerie nationale, 1869–1880).

reconstructed from the Anyang oracle inscriptions.There may be something to be gained, therefore, by considering his sequence of Xia, Shang, and Zhou king lists in the light of the more amply documented history of king lists in Mesopotamia and Egypt.32 The list shown in Figure 3.1 was carved on the wall of an Egyptian temple around 1300 bc. We see Sety, second king of the nineteenth dynasty, making offerings to a list of his predecessors, seventy-six kings going back to the first king of the first dynasty. (In Mesopotamia too a ritual is known in which the king venerates a king list.) Several points deserve notice here. Sety does not confine his offerings to the kings of his own dynasty. His claim to the throne is a claim to inherit the land of Egypt from an unbroken sequence of kings stretching back to the beginning of history. (Other Egyptian lists begin even earlier, with the names of gods and spirits: earthly kings inherit the land from gods. Divine remote ancestors are a feature of king lists in many cultures.) As other sources reveal, however, Sety’s list has been edited, by him or by his predecessors, in several ways. It omits kings deemed illegitimate, the eighteenth dynasty “heretic king” Akhenaten for example, and it omits whole dynasties of foreign rulers. It also omits periods of disunity, when Upper Egypt was ruled by one royal house and Lower Egypt by another. (Other lists deal with periods of divided rule by making concurrent dynasties sequential. In Mesopotamia too, dynasties that in reality were contemporary were either omitted from the Sumerian King List or rearranged to make them sequential.) Sety’s list is the embodiment of a fictitious unity, an assertion of continuous legitimate rule of the whole land, and it was transmitted to later generations not as part of a fuller chronicle but by itself because it was important in itself. For the king at least, it was the one essential fact about the past. As the list on the wall of Sety’s temple shows, transmission of a king list need not entail the transmission of any other information.We possess the original of his list – countless tourists have seen it – and it is a self-contained document. From this it follows that the agreement of Sima Qian’s Shang king list with the list used by Anyang diviners has no bearing on the credibility of anything else he says about Shang. We cannot take his possession of a king list to guarantee that he possessed other information from the second millennium – or that the list itself is an accurate list of real kings.33 In Mesopotamia a text called the Sumerian King List (SKL), composed around 2000 bc, is known from multiple copies made by schoolboys. The list begins when kingship was handed

77

Robert Bagley

down from heaven. Kingship rotates among cities, each in turn made the seat of the sole legitimate dynasty by the gods. Although it is simply a list, the SKL . . . locates the present in a long mundane stream of hegemony and power. As a historical source it is of little value, at least as far as early periods are concerned, because this is a text that is firmly rooted in the fictional notion of Mesopotamia as a single, unified polity that was always ruled from one city by one king who belonged to a specific dynasty. . . .The SKL describes an imperial ideal rather than an accurate state of affairs in the land, and therefore constitutes a perfect example of the use of history for the purposes of legitimating politics of the present rather than as a disinterested depiction of the past. . . . The notion of a single unified hegemonic state is projected into the past, erasing the history of small independent contemporary polities.34 The parallels with ancient China are evident. An ideology of rule needs a history that corroborates it, a history that shows “this is indeed the way the world works.” In China, late formulations of both the ideology and the history are familiar from transmitted texts.The ideology goes under the name “mandate of Heaven,” and the history that supports it is the sequence of dynasties Xia, Shang, and Zhou, each in turn legitimate ruler of civilized China. Legitimacy comes from Heaven, which judges a dynasty and transfers universal rule from a royal line that has declined in virtue to one that is worthier. A linear sequence of king lists, which shows that things have always been so, is a claim to inherit exclusive legitimacy from equally exclusive predecessors. But no Shang or Zhou king ruled the whole of civilized China, nor was political change in ancient China driven by Heaven. On the contrary, both Heaven and its mandate were created by humans with human motives.To the extent that received history is a story edited or invented to substantiate an ideology, the story, like the ideology, has a history of composition. What preceded the formulations we know from transmitted texts? Key terms occur in early Zhou bronze inscriptions, but in passages too brief to explain how their authors understood them. In these inscriptions we encounter the first mentions of Heaven (tian, a deity not mentioned in the Anyang oracle inscriptions, and an anthropomorphic one to judge by the graph used to write the word), its mandate or command or appointment (ming), and the title “Heaven’s son” (tian zi). A Zhou defeat of Shang is mentioned, and we read also that Heaven withdrew its support from Shang because of the drunkenness of Shang officials (incorrect performance of rituals?). Despite the relationship suggested by the title “Heaven’s son,” Heaven seems not to support a dynasty unconditionally. Apart from sober rituals, however, there is no mention of what it (or should we be saying “he”?) looks for in a king. To judge by the oracle inscriptions, the Anyang king’s success in all the affairs of state depended on his sacrifices; his ancestors do not seem to have required any other virtue from him. What Heaven required from the Zhou king may have been similar. Two key elements of later formulations are not visible in the bronze inscriptions. One is the Xia dynasty. No Shang or Zhou inscription speaks of a dynasty before Shang.35 The earliest mentions of Xia are in texts written or edited more than a thousand years after the time it is supposed to have existed, and they mention it only to give the Zhou conquest a moral precedent: “Zhou overthrew Shang because Shang declined in virtue, just as Shang overthrew Xia when Xia declined in virtue.” The other element missing from the inscriptions is an ethical concept of royal virtue. The inscriptions give no hint that royal legitimation was in the hands of moral philosophers. In 78

The Bronze Age before the Zhou dynasty

transmitted texts, by contrast, continued possession of the mandate is contingent on the king’s virtue (de), and virtue is defined in moral terms, for instance as benevolent rule or concern for the well-being of the common people. In this form the ideology seems designed not for the king but for ministers exhorting kings. Was an ideology that originally required nothing from the king but service to the god at some point hijacked by ministers? When was an early Zhou claim to inherit from Shang elaborated into a cyclic theory of history? When and how were the complexities of the second millennium world reduced to a moralizing fable of good first kings and bad last ones, the “consensus history” of the transmitted texts?36 Such questions were not easy to ask before archaeology gave us glimpses of an alternative past. Archaeology in China began as an exploration of a traditional narrative, which it sought to confirm and amplify, but what has happened over the last half-century is an almost imperceptible transformation of its mission.37 From exploring the world described in the texts it has shifted to exploring the wider world that produced the texts. It is this rather than verification that is archaeology’s gift to the historian of ancient China.

Notes 1 Something of the flavor of this project, which announced its conclusions in 2000, can be got from six articles devoted to it in the Journal of East Asian Archaeology vol. 4 (2002). See also Thorp 2006: 23–5. 2 In this practice there lurks the danger that using Anyang as our standard for dating tempts us to think of Anyang as the source of the things being dated. It is a slippery slope from making correlations with Anyang to assuming derivation from Anyang. 3 By comparison, for example, with the early cuneiform corpus from Mesopotamia, which is also more diverse in content and authorship. The entire Anyang oracle-bone corpus can be transcribed in one volume of moderate size. In Mesopotamia more than 80,000 administrative documents have been published from the one century of the Ur III dynasty alone (2112–2004 bc). 4 This is a truism that historians embarrassingly often forget. Osborne (2003: 623) reminds us with an example: “Such accounts as we have of [Greek] religion as a ‘system’ almost all stem from philosophers keen to argue for a particular theological or philosophical position. Notoriously, for example, our fullest account of the rationale of animal sacrifice comes from a treatise advocating vegetarianism.” For all its shortcomings, the material record is free of such agendas. As a distinguished biblical scholar put it: “The Bible presents historical events in the light of a very specific religious interpretation, which archaeological situations do not possess” (H.J. Franken, quoted in Moorey 1991: 134). As for received texts transmitted to us from the Han period, they have been altered by Han editors to an extent that we have no way of estimating. Suppose that we had a Han editor’s transcription of an early Western Zhou inscription, the inscription of a now lost bronze vessel for instance. Even if we make the assumption – for which we have no evidence – that the editor’s intention was exact fidelity, could we trust him to have understood the inscription perfectly? Could we trust his choice of character forms? When Han editors added determinatives to the characters of pre-Han texts, they were interpreting. Can we be sure that our anonymous Han editor understood the inscription on the bronze as well as, say, Chen Mengjia would have? 5 P.R.S. Moorey’s A Century of Biblical Archaeology (Moorey 1991) offers much food for thought. See also Schaberg 2001 and, on the archaeologist’s longing for human narratives, Wang Haicheng 2013. 6 See M. I. Finley and the scholars responding to him in Finley et al. 1964. Most ancient cities burned to the ground more than once; archaeologists are seldom able to determine the cause of a fire. After more than a century of fervid claims for the truth of Homer, the cool agnosticism of the most recent scholarship is sobering (Jablonka 2011). Ancient texts that purport to be factual can, like modern novels, be fictions set in places that are real and elaborated around the names of people who really existed (Baines 2011 is instructive here, especially pp. 68–9). Though textual scholars often claim to be able to separate the “kernel of fact” from the “fictional embellishments,” they cannot give us a rule for doing so. Of course, for the reader who believes the text, archaeological proof is not really required. The believer accepts the narrative and goes to archaeology only for illustrations (a picture of a ruined wall). 7 Moorey (1991: 93–4) gives a spectacular example of complete fusion, and confusion, of biblical text and archaeology: no summary or sample can give the flavor of the paragraphs he quotes. Parallels in the

79

Robert Bagley literature of Chinese archaeology – passages in which it is impossible to guess what evidence, or even what form of evidence, lies behind particular statements – would be easy to find. 8 What archaeologists call a culture or an assemblage is generally defined as “a constellation of material traits that occur together consistently at different sites.” On the archaeological culture, its relation to pottery, and its interpretation in social terms, for example as the material expression of an ethnic group or a polity, see Trigger 1974; Wang Haicheng 2014b; Li Yung-ti 2014: 139–41. In archaeological parlance, Liangzhu is the type site of the Liangzhu culture, the site whose artifactual assemblage is taken to define the culture. In principle an assemblage can include everything from jades to buildings to burial forms, but in practice pottery is the archaeologist’s most reliable type fossil, for it is ubiquitous, indestructible, and variable enough in form and fabric to be a sensitive register of time and place.When archaeologists speak of “cultural remains” they often mean no more than “potsherds.” If pottery resembling that of the Liangzhu site is found at another site, the new site is said to be a site of the Liangzhu culture. 9 For the early history of oracle-bone studies see Li Chi 1977: chapters 1–2. For a general introduction to the inscriptions see Keightley 1999. 10 Li Ji describes the 1928–37 excavations in Li Chi 1977. For a summary account of those and later excavations see Bagley 1999: 180–208. See also Jing Zhichun et al. 2013 (especially for Huanbei); Thorp 2006; Liu & Chen 2012: chapter 10; Wang Haicheng 2015. 11 Cultural contact was under discussion already while the Anyang excavations were going on. When painted pottery that resembled Neolithic pottery from the Near East was found in northwest China in the early 1920s, it was widely taken to be of western origin; some equated it with the Xia dynasty. When distinctly different pottery was found at a site on the east coast in 1931, Li Ji and others took it to represent an indigenous contribution to the formation of Chinese civilization. See Bagley 1999: 127–30 and Bagley 2014. 12 Wang Haicheng 2015: 154. In Egypt and Mesopotamia the use of administrative forms as patterns for display inscriptions is well known. 13 On the Erligang site and civilization see Bagley 1999: 165–71; Steinke 2014a; Yuan Guangkuo 2013; Liu and Chen 2012: chapter 8; Thorp 2006: chapter 2. 14 On the origin of the Chinese writing system see Bagley 2004 and Bagley 2014: 43–5; on chariots and domesticated horses, Bagley 1999. 15 On the Erlitou site and civilization see Bagley 1999: 158–65; Liu and Chen 2012: chapter 8; Xu Hong 2013; Thorp 2006: chapter 1. On the Erlitou bronze industry and its relation to Erligang see Bagley 2014: 38–40. On Yanshi Shangcheng see Yuan Guangkuo 2013: 325–6, 328–9; Liu and Chen 2012: 278–80; Thorp 2006: 22–3, 67–73. 16 Liu and Chen (2012: 271) have recently advocated detaching Erlitou archaeology from Xia, but this remains a minority view; the Xia Shang Zhou Chronology Project has given state sanction to the three dynasties of textual tradition. On the dynastic model in second millennium archaeology see Li Yung-ti 2014. 17 Wang Haicheng 2014b: 68. 18 On Shandong see Fang Hui 2013; Liu and Chen 2012: 363–7; Bagley 1999: 219–21. 19 On the Panlongcheng site and the Erligang expansion see Bagley 1999: 168–71; Steinke 2014a, especially the chapters by Zhang Changping, Wang Haicheng, and John Baines (and, for copper mines, pp. 166–8 in the chapter by Steinke); and Liu and Chen 2012: 284–90. On Erligang in the north see Lin Yün 1986. 20 On Sanxingdui see Bagley 1999: 212–19; Bagley 2001; Sun Hua 2013. The city is believed to have declined around the eleventh century bc. A site called Jinsha in modern Chengdu is thought to be a successor. 21 Some scholars have insisted that the Zhengzhou king kept bronze casting a royal monopoly, prohibiting its spread beyond the capital to places like Panlongcheng, but this seems to be an a priori conviction, not an inference from evidence. At no point in the Bronze Age is a monopoly of metal technology visible in the archaeological record. Some of the finest Xingan castings are indisputably local. If the mastery of Erligang art and technology they display did not come from a place like Panlongcheng, 300 km to the north, where did it come from? 22 On the Xingan find and its implications see Bagley 1999: 171–5; Peng Shifan 2004; Steinke 2014b. 23 Notice that all these mechanisms are biased in favor of spectacular discoveries: construction crews do not halt work because an exceptionally observant worker has noticed some clay mold fragments. The oft-repeated claim that bronzes cannot have been cast in the south because clay molds for them have

80

The Bronze Age before the Zhou dynasty not been found in the south is the most dubious of arguments ex silentio. It could be countered by observing that molds for southern bronzes have not been unearthed in the north either. Does this mean that they were cast nowhere? 24 On the lower Yangtze region see Steinke 2014b; on tuned sets, Bagley 2005: 79–83 and Bagley 2015: 67–73. 25 As late as 1997, a senior bronze specialist assigned Hunan nao to the Chunqiu period – i.e. eighth to fifth century bc – because he could see no place for them in the second millennium. When the same scholar exhibited the contents of the Xingan tomb at the Shanghai Museum in 1992, he left the three nao out of the exhibition. Even scholars who believe southern nao to be southern castings have usually assumed that they descend from Anyang ancestors and that they postdate those ancestors by whatever time seems necessary to account for a dramatic increase in size. 26 The recognition of civilized Anyang-period cultures in the south is owed to Virginia Kane (1974), who drew particular attention to the large bells that connect them all and distinguish them from the civilization of the north. Her work is updated in Bagley 1987, introduction, section 1.12, and Bagley 1999: 208–12; see also Steinke 2014b. 27 Baines 2014. 28 Rawson 1989. See also Bagley 1999: 226–31. 29 For connections between the Wei valley, the Hanzhong region of southern Shaanxi, and Xingan in Jiangxi, see Bagley 1999: 178–80. 30 On the functions of Anyang writing see Wang Haicheng 2014a, especially chapters 4 and 6, and Wang Haicheng 2015. On the role of functional context in the invention of writing see Damerow 1999; Postgate 1994: chapter 3; Bagley 2004. Scholars inattentive to the preservation bias of the archaeological record and unaware of what is involved in the development of full writing have sometimes suggested that our sample of early writing is complete – that writing was invented in the reign of Wu Ding and used only at Anyang and only for divination. For reasons suggested earlier and laid out in detail in the works just cited, this cannot be so. 31 Wang Haicheng 2014a: 50, Text 2.6. 32 On king lists in Mesopotamia and Egypt see Michalowski 2011 and Baines 2011. On king lists in general see Wang Haicheng 2014a: chapters 1 and 2. 33 In other words, Sima Qian’s possession of a king list does not tell us whether other chronicles ever existed. Shaughnessy (2011: 391) speculates that Western Zhou royal archives might have contained royal speeches and “sagas,” including “a year-by-year annalistic history with entries similar to the greatevent notations found in bronze inscriptions” (on lists of year names compare Bagley 2004: 223, and Baines 2011: 57–9). But the ancient history recounted in transmitted texts does not seem to me to have the flavor of genuinely early annals or great‑event year names (cf. Postgate 1994: 40). It reads more like a king list embroidered with invented anecdotes of the didactic/argumentative kind discussed in Schaberg 2011. 34 Michalowski 2011: 15. 35 One middle Western Zhou bronze inscription, that of the Bin gong xu, mentions the controller of floods Yu, who in transmitted texts figures as the founder of the Xia dynasty. If the Xia story was elaborated around an already established culture hero, this would fit a familiar pattern in Chinese mythologizing. 36 See Wang Haicheng 2014a: chapter 2. Knoblock 1990 uses the term “consensus history” for the story of the past that was taken for granted by late Zhou and Han writers. His handy reconstruction of that history is marred by his enthusiasm for declaring parts of it to be vindicated by archaeology (and even by astronomy: claims that transfers of the mandate coincided with celestial events seem to accept that the heavens really do intervene in history – though only in Chinese history). The usefulness of the story Knoblock reconstructs lies in the fact that it was widely believed by late Zhou and Han writers. Whether we too should believe it is a question best kept separate. 37 On changes in the thinking of archaeologists see Li Yung-ti 2014.

Works cited Bagley, Robert 1987. Shang Ritual Bronzes in the Arthur M. Sackler Collections. Cambridge, MA: Harvard University Press. Bagley, Robert 1999. “Shang Archaeology.” Chapter 3 (pp. 124–231) in Michael Loewe and Edward L. Shaughnessy, eds., The Cambridge History of Ancient China (Cambridge: Cambridge University Press).

81

Robert Bagley Bagley, Robert (ed.) 2001. Ancient Sichuan, Treasures from a Lost Civilization. Seattle and Princeton: Seattle Art Museum and Princeton University Press. Bagley, Robert 2004. “Anyang Writing.” Chapter 7 (pp. 190–249) in Stephen Houston, ed., The First Writing (Cambridge: Cambridge University Press). Bagley, Robert 2005. “The Prehistory of Chinese Music Theory.” Proceedings of the British Academy 131 (2005): 41–90. Bagley, Robert 2014. “Erligang Bronzes and the Discovery of the Erligang Culture.” Chapter 1 (pp. 19–48) in Steinke 2014a. Bagley, Robert 2015. “Ancient Chinese Bells and the Origin of the Chromatic Scale.” Zhejiang University Journal of Art and Archaeology 2 (2015): 55–102. Baines, John 2011. “Ancient Egypt.” Chapter 3 (pp. 53–75) in Feldherr & Hardy 2011. Baines, John 2014. “Civilizations and Empires, A Perspective on Erligang from Early Egypt.” Chapter 4 (pp. 99–119) in Steinke 2014a. Damerow, Peter 1999. The Origins of Writing as a Problem of Historical Epistemology. Preprint 114. Berlin: Max-Planck-Institut für Wissenschaftsgeschichte. Fang, Hui 2013. “The Eastern Territories of the Shang and Western Zhou: Military Expansion and Cultural Assimilation.” Chapter 23 (pp. 473–93) in Underhill 2013. Feldherr, Andrew and Grant Hardy (eds.) 2011. The Oxford History of Historical Writing,Volume I: Beginnings to ad 600. Oxford: Oxford University Press. Finley, M. I., J. L. Caskey, G. S. Kirk, and D. L. Page 1964. “The Trojan War.” The Journal of Hellenic Studies 84 (1964): 1–20. Jablonka, Peter 2011. “Troy in Regional and International Context.” In Sharon R. Steadman and Gregory McMahon, eds., The Oxford Handbook of Ancient Anatolia, 10,000–323 b.c.e. (Oxford University Press, 2011), pp. 717–33. Jing, Zhichun, Tang Jigen, George Rapp, and James Stoltman 2013. “Recent Discoveries and Some Thoughts on Early Urbanization at Anyang.” Chapter 17 (pp. 343–66) in Underhill 2013. Kane, Virginia C. 1974–75. “The Independent Bronze Industries in the South of China Contemporary with the Shang and Western Zhou Dynasties.” Archives of Asian Art 28 (1974–75): 77–107. Keightley, David N. 1999.“The Shang: China’s First Historical Dynasty.” Chapter 4 (pp. 232–91) in Michael Loewe and Edward L. Shaughnessy, eds., The Cambridge History of Ancient China (Cambridge: Cambridge University Press, 1999). Knoblock, John 1990. Xunzi, A Translation and Study of the Complete Works, Volume II (Stanford: Stanford University Press), Chapter 1 (“The Lessons of History”). Li, Chi [Li Ji] 1977. Anyang. Seattle: University of Washington Press. Li,Yung-ti 2014. “The Politics of Maps, Pottery, and Archaeology: Hidden Assumptions in Chinese Bronze Age Archaeology.” Chapter 6 (pp. 137–46) in Steinke 2014a. Lin,Yün 1986. “A Reexamination of the Relationship between Bronzes of the Shang Culture and of the Northern Zone.” Chapter 10 (pp. 237–73) in K. C. Chang, ed., Studies of Shang Archaeology (New Haven: Yale University Press, 1986). Liu, Li and Xingcan Chen 2012. The Archaeology of China. Cambridge: Cambridge University Press. Michalowski, Piotr 2011. “Early Mesopotamia.” Chapter 1 (pp. 5–28) in Feldherr & Hardy 2011. Moorey, P.R.S. 1991. A Century of Biblical Archaeology. Louisville, KY: Westminster/John Knox. Osborne, Robin 2003.“Getting History from Greek Archaeology – ‘Some Way to Go’.” Antiquity 77(2003): 612–16. Peng, Shifan 2004. “A Study of the Dayangzhou Discovery.” Chapter 9 (pp. 216–33) in Yang Xiaoneng, ed., Chinese Archaeology, New Perspectives on China’s Past in the Twentieth Century, Volume 1 (New Haven: Yale University Press). Postgate, J. N. 1994. Early Mesopotamia: Society and Economy at the Dawn of History. Rev. ed. London: Routledge. Rawson, Jessica 1989. “Statesmen or Barbarians? The Western Zhou as Seen Through Their Bronzes.” Proceedings of the British Academy 75 (1989): 71–95. Schaberg, David 2001. “Texts and Artifacts: A Review of The Cambridge History of Ancient China.” Monumenta Serica 49 (2001): 463–515. Schaberg, David 2011. “Chinese History and Philosophy.” Chapter 16 (pp. 394–414) in Feldherr & Hardy 2011. Shaughnessy, Edward L. 2011. “History and Inscriptions, China.” Chapter 15 (pp. 371–93) in Feldherr & Hardy 2011.

82

The Bronze Age before the Zhou dynasty Steinke, Kyle (ed.) 2014a. Art and Archaeology of the Erligang Civilization. Princeton: Princeton University Press. Steinke, Kyle 2014b. “Erligang and the Southern Bronze Industries.” Chapter 7 (pp. 151–70) in Steinke 2014a. Sun, Hua 2013. “The Sanxingdui Culture of the Sichuan Basin.” Chapter 8 (pp. 147–68) in Underhill 2013. Thorp, Robert 2006. China in the Early Bronze Age: Shang Civilization. Philadelphia: University of Pennsylvania Press. Trigger, Bruce 1974. “The Archaeology of Government.” World Archaeology 6 (1974): 95‑106. Reprinted in Trigger, Time and Traditions (Edinburgh: Edinburgh University Press, 1978), chapter 10 (pp. 153–66). Underhill, Anne P. (ed.) 2013. A Companion to Chinese Archaeology. Chichester: Wiley-Blackwell. Wang, Haicheng 2013. “Inscriptions from Zhongshan: Chinese Texts and the Archaeology of Agency.” Chapter 9 (pp. 209–30) in Joshua Englehardt, ed., Agency in Ancient Writing (Boulder: University Press of Colorado, 2013). Wang, Haicheng 2014a. Writing and the Ancient State, Early China in Comparative Perspective. Cambridge: Cambridge University Press. Wang, Haicheng 2014b. “China’s First Empire? Interpreting the Material Record of the Erligang Expansion.” Chapter 3 (pp. 67–97) in Steinke 2014a. Wang, Haicheng 2015. “Writing and the City in Early China.” Chapter 7 (pp. 131–57) in Norman Yoffee, ed., The Cambridge World History, Volume III, Early Cities in Comparative Perspective, 4000 bce – 1200 ce (Cambridge: Cambridge University Press, 2015). Xu, Hong 2013. “The Erlitou Culture.” Chapter 15 (pp. 300–322) in Underhill 2013. Yuan, Guangkuo 2013. “The Discovery and Study of the Early Shang Culture.” Chapter 16 (pp. 323–42) in Underhill 2013.

83

4 THE WESTERN ZHOU STATE

LI FENGTHE WESTERN ZHOU STATE

Li Feng

If one were to point to a dynastic house that had the longest duration in Chinese history, that has to be that of Zhou (1045–256 bc). Even after the pages of Zhou’s glory were already turned over, the name “Zhou” still carried considerable prestige and was subsequently revived as the dynastic title of five regimes, making a total length of 857 years during which the whole or a large part of China was under “Zhou” rule. For Confucius and his disciples the Western Zhou (1045–771 bc) period was certainly the golden age of civilization. Notwithstanding the extremely long duration of its venerated name, the Western Zhou state suffered very early decline and was thereafter constantly troubled by political tensions built in or from outside. It was the cultural complex created by the Zhou under the guidance of a set of unique political and ritual institutions which were adopted by the amalgamation of diverse populations that helped penetrate the Zhou king’s ceremonial role as the “Son of Heaven” even centuries after the political power of the Zhou house had already waned. In global history, the Zhou rose to dominance in a time that paralleled the so-called Dark Age (ca. 1100–900 bc; Van De Mieroop 2004, 189–194) in the Near East which anticipated the rise of large-scale empires, namely the Neo-Assyrian Empire (911–609 bc), followed by Persia. It also paralleled the Greek “Dark Age” (ca. 1100–776 bc; Hornblower et al, 2012, 628) in the history of the Mediterranean World.The Zhou set up initial conditions for a long historical process that, through important modifications, eventually led to the rise of the Qin Empire (221–207 bc). But the Zhou also created a model of state whose structural and organizational logic was meaningfully different from that of the Assyrian Empire and the Greek “city states.”We are only now beginning to understand the true nature of the Zhou polity on the basis of an expanding pool of new data helped by new methodological tools, and we are yet to seriously address its comparative value for the study of the world’s early civilizations.

Time and space Much of what we know about early Western Zhou dates was tied to a single most important incident, the conjunction of the five major planets of the solar system brightly visible in the northern sky of the Zhou homeland in central Shaanxi, and this is said to have happened in the thirty-second year of the last Shang king, Di Xin 帝辛 (Jinben zhushu jinian, 34). Modern historians with the help of scientific methods were able to fix this incident in the fifth month of 1059 84

The Western Zhou state

bc, and there is solid ground to believe that this was the actual astronomical-religious basis for the Zhou belief in the “Mandate of Heaven” that was said to have been bestowed on King Wen 文, who thereupon declared his kingship (Pankenier 1995, 121–176).The historical record then calculates down sequentially for twelve years (Shiji: 4, 117–121) to when, in winter 1046 bc, the Zhou conquered Shang and inaugurated the Western Zhou dynasty (1045–771 bc). The Western Zhou dynasty had a total of 274 years, ending in 771 bc when the Zhou capitals were sacked by groups of enemies that descended from the northwestern highlands. Although the length of the period is firmly known, to establish dates for each of the twelve royal reigns has been a traditional and yet new controversy that is open to question. But there are at least a few “fixed points” that are generally accepted among scholars, who also agree to a tripartite division of the Western Zhou into three periods: early, middle, and late. In 842 bc, a rebellion stormed the Zhou capital and ended the reign of King Li 厲, after which the dates are traditionally known for the last two kings, Xuan 宣 (forty-six years) and You 幽 (eleven years); in 899 bc, a solar eclipse occurred shortly after sunrise, observed at the city of Zheng 鄭 in western Shaanxi, and this was in the first year of King Yih 懿 (eighth king), who ruled during the mid-Western Zhou (Shaughnessy 1991, 256). How then to partition the period of 1045–899 bc among the six early and mid-Western Zhou kings (Wu 武, Cheng 成, Kang 康, Zhao 昭, Mu 穆, and Gong 恭) or the period of 899–842 bc among the four later kings (Yih 懿, Xiao 孝, Yi 夷, and Li 厲) rests on scattered and sometimes conflicting textual references, combined with year numbers recorded in the bronze inscriptions that indicate a king ruled for at least a certain number of years (but usually we do not know which king). Many inscriptions indeed also record months, day numbers, and even lunar phases, and if they can be matched to a calendar counting months and days back from 842 bc, there is hope to determine the lengths of the relevant royal reigns. Systematic effort to achieve this goal was of course attempted by many scholars, and Shaughnessy’s dates (Table 4.1) Table 4.1  Dates of Western Zhou Kings Kings

Dates

Periodization

King Wen King Wu Duke of Zhou King Cheng King Kang King Zhao King Mu King Gong King Yih King Xiao King Yi King Li Gong He King Xuan King You

1099/56–1050 B.C. 1049/45*–1043 1042–1036 1042/35–1006 1005/3–978 977/75–957 956–918 917/15–900 899/97–873 872?-866 865–858 857/53–842/28 841–828 827/25–782 781–771 B.C.

Pre-conquest Early Western Zhou

Middle Western Zhou

Late Western Zhou

Dates of Western Zhou kings follow Shaughnessy 1991, xix. *  That some of the kings are assigned two first years is based on Nivison’s proposal that a king started his reign the next year after his father’s death, and then, after the completion of the mourning period, he calibrated the year and reinstituted another calendar (Nivison 1983, 481–580).

85

Li Feng

represent perhaps the most consistent reading of this evidence (Shaughnessy 1991, 217–287). However, it should be noted that we are still unable to come to a conclusion on many of these dates. Nevertheless, we know at least that King Wu died in 1043 bc, two years after the conquest;1 King Zhao died in his nineteenth year on his ill-fated southern campaign. Inscriptions also show that King Kang ruled for at least twenty-five years, and a new inscription (Yaogong gui 公簋) tells that King Cheng ruled for at least twenty-eight years. During the mid-Western Zhou, we know that King Mu ruled for a minimum of thirty-four years (Xian gui 鲜簋) and King Gong fifteen years. Finally, from King Yih to King Li, the average length of the reigns was fourteen years. In the Zhou case, the condition of sources (see below) allows us to establish convincing links between historical processes and actual geographical locations, based largely on the archaeologically excavated written evidence. The Zhou state-building process seems to have begun on the western highlands of eastern Gansu and western Shaanxi. Archaeological evidence combined with the transmitted Zhou poetic tradition suggest that the pre-dynastic Zhou center was located in present-day Zhouyuan 周原, on top of the loess highland south of the Qi Mountain and north of the Wei River (Map 4.1). This “Settlement of Qi” (Qiyi 岐邑) served as the locus of Zhou power in the pre-conquest century after the Zhou’s historical move to the Wei River valley, led by the Ancient Duke (Gugong 古公). Before it, Zhou activities were associated with a site called Bin 豳, most likely to have been located in the present-day Binxian 彬縣-Changwu 長武 area in the upper Jing River valley to the north of the Qi Mountain. Although the specific location of Bin has yet to be determined by archaeology, the ceramic traditions of the two valleys in

Map 4.1  The Zhou central region in the Wei River Valley, Shaanxi

86

The Western Zhou state

the pre-conquest era were clearly homologous, suggesting a material cultural setting from which the Zhou rose to power (Li Feng 1991, 265–284). Two generations later, with the accession of King Wen, the Zhou’s political sway had clearly extended to reach back to the upper Jing River region as far as the edge of the Long-Liupan Mountain ranges in the west; in the east, it reached the mountainous region of western Henan that separated the Wei River valley from the heartland of Shang in northeastern Henan. The conquest campaign led by King Wu in 1046 bc thus played out as the showdown between chiefdoms and tribes that inhabited the plateaus and valleys of western China, united in various degrees with the rising power of Zhou, versus the hegemonic power of the Shang king at the head of the states and communities located on the eastern China plains. The political map that was created as the result of the Zhou’s two conquests (see below) typically incorporated these two largely separate geographical zones (Map 4.2). By the closing of the early Western Zhou in the middle of the tenth century bc, the Zhou had carved out a large network of political-military presence of varying degrees that reached the entire area of Henan, Hebei, western and northern Shandong, central Shaanxi, eastern Gansu, southern Shanxi, the southwestern corner of Liaoning, northern Anhui, and the area to the east of Han River in the middle Yangtze valley of Hubei

Map 4.2  Geography of the Western Zhou state 1. Zhougongmiao, 2. Huxizhuang, 3. Zhaojialai, 4. Gaojiabu, 5. Dingjiagou, 6. Yangjiacun, 7. Shigushan, 8. Rujiazhuang, 9. Zhuyuangou, 10. Nanpo. 11. Baicaopo, 12. Nianzipo, 13. Zhengjiawa, 14.Yujiawan, 15. Yucun, 16. Pingliang, 17. Sunjiazhuang, 18. Lingkou, 19. Wangchuan, 20. Liangdaicun, 21. Tianshui, 22. Maojiaping, 23. Dabuzishan, 24. Sanmenxia, 25. Qucun, 26. Yongningbu, 27. Cangtou, 28. Tianzhuang, 29. Beipinggao, 30. Xincun, 31. Jieduanying, 32. Xiapanwang, 33. Guitaisi, 34. Gezhuang, 35. Xizhuangcun, 36. Liangshan, 37. Zhuanglixi, 38. Xue Gucheng, 39. Liutaizi, 40. Shouguang, 41. Yujia, 42. Hexi, 43. Dongqucheng, 44. Lujiaggou, 45. Zhuangtou, 46. Guicheng, 47. Xi’an, 48. Guheya, 49. Linzi, 50. Xinye, 51. Guojiamiao, 52.Yangzishan, 53.Yejiashan, 54. Lutaishan, 55. Xiaogan, 56. Sujialong, 57. Jiangling, 58. Baifu, 59. Niulanshan, 60. Shanwanzi, 61. Machanggou, 62. Beidong, 63. Weiyingzi.

87

Li Feng

(Li Feng 2006, 341). Given the nature of the Western Zhou as a settlement-based state (see later), this does not entail continuous Zhou control of every piece of land within the areas that the Zhou’s political power was able to reach (Li Feng 2010–2011, 291–292). However, combined evidence from archaeology and texts suggests that numerous settlements inhabited by the Zhou elites and their subjects, or by a mixed population that submitted to the Zhou, must have been located within this maximum composition of spaces through which the political network of the Western Zhou state extended. There is little doubt that the Zhou king’s ability to pursue political goals and to carry out coordinated military actions within or even beyond this composition was much higher than what was possible for the Shang king.There is also good evidence that the Zhou were able to achieve this within a relatively short period after their conquests of Shang. However, the political development following the disastrous southern campaign that ended the reign of King Zhao made the Zhou’s wide political network increasingly elusive, and efforts to maintain the same degree of control became difficult if not impossible. There were times the Zhou were clearly unable to defend even settlements that were located close to the core region of the state in the face of foreign invasions, especially by the Huaiyi 淮夷 from the southeast, who constantly challenged Zhou’s security. In the northeast, the Zhou seem to have retreated for a long stretch and therefore had to fight enemies in southern Hebei during the mid-Western Zhou. But in other times, even during the late Western Zhou, we still find that the Zhou kings were conducting coordinated military campaigns very far from the Zhou center, in regions such as northern Anhui and southeastern Shandong, involving also some of Zhou’s regional states (see below) located far away from the battlefields. Evidence also shows that in the reign of King Xuan, the Zhou made a systematic effort to recover control in the middle Yangtze region by trans locating a number of key polities to the Nanyang basin in southern Henan. But this seems to have then contributed to weakening Zhou’s defense in regions to the northwest of the Zhou royal domain in the Wei River valley.

Sources The Western Zhou is the first period of Chinese history about which our study can benefit substantially from the joint operation of history and archaeology on the basis of impartial consideration of three types of evidence: archaeology, inscriptions, and received texts. Since as early as the 1930s, sites dating to the Western Zhou period have been under continuous survey and excavation by archaeologists, which have sufficiently clarified the sequence of the Zhou material culture. While these early works were concentrated in the core regions of the Western Zhou, namely the Wei River valley and the Luoyang plain, the strategic move of research focus to the outlying regions in the 1980s led to a dramatic expansion of our knowledge about the Western Zhou state at large.The fieldwork has turned out a large number of sites with remains of palace and house foundations, storage pits, workshops, and burial remains that offer us direct access to the social life of the period. From such structures archaeologists extracted huge quantities of portable materials such as pottery, jades, lacquers, and bronze objects, alongside human and animal remains, sometimes also accompanied by the burial of horse-drawn chariots. This archaeological information not only allows us to attain an ever-deepening understanding of the material wealth of the Western Zhou, particularly of its regional modifications, but also offers a firm base for interpreting the history of the period. However, a special note must be made that, although the material evidence provides us with direct access to a scene of lives in the past, they are both fragmentary and accidental, and never in themselves constitute a systematical arrangement of information. For the purpose of historical analysis, they need to be used in connection with other types of evidence according to carefully designed research plans. 88

The Western Zhou state

Fortunately, the material evidence is not all silent for the existence of a special kind of archaeological evidence, the inscribed bronzes. The bronze inscriptions are texts of varying length and complexity that were cast on the bronze objects, most importantly vessels. It is hard to imagine today that the study of the Western Zhou period could move forward without these metal texts, and the latest compilation gives the total number of inscriptions on vessels and musical instruments dating from Shang to the Eastern Zhou as 15,989 (the vast majority being from the Western Zhou), excluding very new and still unpublished pieces (Wu Zhenfeng 2012, 1: 2). As the inscribed bronzes are both a material and textual presence, they offer the best link between archaeology and history and, when scientifically excavated from specific locales, help us put historical processes on the real ground. In the past a rather narrow understanding of the bronze inscriptions as religious documents addressed to the ancestors (Falkenhausen 1993, 145–152) suffered from both methodological deficiencies and defeat by a fuller consideration of more inscriptions. Recent scholarship instead has alerted scholars to the multiple purposes for which the inscribed bronzes were manufactured and multiple social contexts in which they served the needs of the Zhou elites (Li Feng 2008, 11–20; 2011, 293–300). Of particular importance are more than one hundred “appointment inscriptions” that record court ceremonies in which the Zhou king personally assigned administrative duties to his officials.These inscriptions, often presenting historical details, are direct evidence of Zhou governance and of the Western Zhou state. For the very fact that they emerged from various occasions almost invariably related to the life stories of those who commissioned their casting and hence owned them, the bronze inscriptions are critical testimonies for Zhou society in almost every aspect. The received texts on the Western Zhou period constitute different strata, the earliest being created in the contemporaneous Western Zhou period. This includes a dozen chapters in the Book of Documents (Shangshu 尚書) and a considerable number of poems from the Book of Poetry (Shijing 詩經). While the five “Announcement” chapters of the Shangshu were known to have been transmitted from the early Western Zhou period, other chapters together with the majority of Western Zhou poems in the Shijing are probably from the late Western Zhou. The second layer is composed of textual references from the Warring States period, from both transmitted and excavated texts. In the transmitted group the most important are Zhushu jinian 竹書紀年 (Bamboo Annals), the Guoyu 國語 (Speeches of the States), and the Zuozhuan 左 傳 (Zuo Commentary). Particularly the first (originally excavated from a Warring States tomb in the third century ad) offers chronicles that cover the entire Western Zhou and Spring and Autumn periods (Shaughnessy 2006, 185–256), and the second includes speeches supposed to have been given in royal and regional courts from the mid-Western Zhou onward. The third, though detailing historical developments in the Spring and Autumn period, gives frequent reference to early history. The excavated group from Warring States tombs includes mostly philosophical writings, but the recently published bamboo manuscripts from Qinghua University include about ten essays that purport to speak about Western Zhou history, together with a compilation of miscellaneous histories of obscure origins that is called Xinian 繫年 (Chronicles) (Allan 2012, 547–557; Li and Liu 2010, 6–15). The third stratum of information is provided by Han dynasty texts, most importantly the Shiji 史記 (Grand Scribe’s Records), which offer what is by nature a retrospective Han dynasty account of  Western Zhou history. The proper use of textual information involves methodological challenges, as in all regional histories in the world. Naturally earlier layers of information are preferred to later layers. But due to the very complex nature of text transmission, the chronological order of the texts cannot be simply transferred into the value order of the information they contain; there may be cases where information in a later text might have had an earlier origin. So to accept or to reject an early or late source requires the professional judgment of the historian, who must be familiar 89

Li Feng

with strategies of textual analysis in order to be able to weigh the value of the various sources. To offset this problem, it is always important that the textual information is used on grounds that have already been established from archaeological data and the contemporaneous bronze inscriptions (Li Feng 2010–2011, 306). On the other hand, the proper use of textual information allows for a fuller and often better contextualized understanding of archaeological data as well as inscriptions. Taking both textual and inscriptional sources together with regard to what archaeology also can tell us, an outline narrative of Western Zhou history, though still with some gaps that need to be filled in the future, can indeed be offered in the interest of revealing long-term trends of historical development.

Founding and expansion The poetic tradition of the Zhou credits the Ancient Duke for the founding of the Zhou state. The poem “Mian” 綿 says that after the Zhou’s relocation in the Settlement of Qi in the Wei River valley, the Ancient Duke laid out foundations for houses, demarcated the fields, constructed temples and city gates, and established administrative offices of the “Supervisor of Construction” and “Supervisor of Multitudes,” two key titles that characterized Zhou bureaucracy, and which were unknown to the Shang. Arguably this was a new state on the horizon. But it was still not a kingdom, and the title “Grand King” (Taiwang 太王) was almost certainly posthumously bestowed on the Ancient Duke after King Wen’s assertion of kingship in 1059 bc, if not after the Zhou conquest of Shang. By the same token, King Wen’s father, who was called Ji Li 季歷 or “King Ji,” was not a king in his lifetime either. The transition from the Ancient Duke to King Wen was a critical period in the rise of Zhou power but probably also created one of the political scars in early Zhou history. It is very likely that the issue of succession had led the three sons of the Ancient Duke to open conflict which was resolved only after the two elder brothers fled to the adjacent mountains, leaving Ji Li the only successor to their father. But this history seems to have been completely reworked after the establishment of King Wen as king. With his father promoted to “King Ji” and grandfather to “Grand King,” representing the legitimate line of succession to Zhou power, his two politically defeated unfavorable uncles were then honored as hermits who willingly yielded the position to their younger brother in order for it to pass onto the future King Wen, the founder of Zhou hegemony. This was a piece of mystery in early Zhou history that led to a millennium-long belief in the peaceful abdication of King Wen’s two uncles. Only now can we see it in the light of an alternative interpretation that may better capture the real dynamics of early Zhou politics. Ji Li is said to have been killed by the Shang king, an incident that reflects the Shang overlord no longer tolerating the growing Zhou power. But in any event, the Zhou seem to have been able to maneuver the situation to their best interests, as it is traditionally undisputed that King Wen himself married a princess from Shang and thus remained a nominal subject of the Shang king. This close relationship with Shang is actually well corroborated by the oracle-bone inscriptions found in the pre-dynastic Zhou capital Zhouyuan that show clearly that the Zhou worshiped the deceased Shang kings in a more prominent position than the Zhou’s own ancestors. On the other hand, the material culture, particularly bronzes in the Zhou homeland of the Wei River valley, also shared a great deal with bronzes found in the Shang center Anyang both in terms of their technical features and of their artistic standards (Figure 4.1). However, at the point of King Wen’s death in 1050 bc, the Zhou leadership must have grown impatient with the Shang overlord and confident about their own power; therefore, they had run quickly into a showdown with the Shang soon after the next king, Wu, completed his ritual obligation – three years of mourning for his father. 90

The Western Zhou state

Figure 4.1 Bronze deer from Shigushan 石鼓山 in Baoji (Tomb no. 4: 212, 214), dating slightly before the Zhou conquest: Images provided by Wang Zhankui and Ding Yan, Shaanxi Academy of Archaeology.

According to Shaughnessy’s reconstruction of the conquest campaign, the Zhou army set out to the east late in the tenth month of 1046 bc. After more than two months of marching, the Zhou army reached a place called Muye 牧野 in the vicinity of the Shang capital, where a very bloody battle indeed took place on the Jiazi day in the first month of 1045 bc (Shaughnessy 1981–1982, 66–67), a date that is corroborated well by the inscription on the Li gui 利簋 cast some seven days after the Zhou conquest.The Shang army was crushed, and the last Shang king, Zhou 紂, committed suicide back in his palace. But the Zhou seem to have been unprepared for a quick victory, as the post-conquest historical development would tell us. After a short period of occupation King Wu led the main body of the Zhou army westward towards home, leaving two of his brothers, Guanshu 管叔 and Caishu 蔡叔, each in command of a garrison force near the Shang capital, where the subjugated Shang population were left to the rule of Wu Geng 武庚, a son of the last Shang king. What happened next concerns a traditional debate about the alleged “kingship” of the “Duke of Zhou” (Zhougong 周公), who became the actual leader at the Zhou court upon the death of King Wu in 1043 bc. He seems to speak in the voice of a Zhou king in the “Kang Gao” 康 誥 chapter of the Shangshu; in order to promote his historical role in the founding of the Zhou regime, Confucian scholars had done much to downplay the age and significance of King Cheng, who was thus said to have still been held in his mother’s arms. In any event, this new 91

Li Feng

political change at the Zhou court seems to have provoked the rebellion of the Shang prince Wu Geng and the former Shang allies, who killed the Zhou brothers and the troops under their command.2 The bronze inscriptions suggest, however, that King Cheng was not only old enough to lead military campaigns to deal with the crisis in the east, he was dispatching another uncle, the Duke of Shao (Shaogong 召公) or the “Grand Protector,” to carry out putative attacks on five allied states of Shang after the Zhou regained control of the formal Shang capital.3 This “Second Conquest” that lasted for at least three years was a critical transition in Zhou power because it not only established Zhou’s firm control of the Shang center but helped them to expand their political sway to the peripheries of the eastern plains, in areas including northern Hebei, western Shandong, southern Henan, and possibly northern Anhui, thus paving the foundation for a new state. The consolidation of the fruit of the “Second Conquest” took two important steps: (1) a new capital named “Chengzhou 成周” (Accomplishment of Zhou) was constructed in present-day Luoyang, which then served as the political and administrative center for the Western Zhou state close to the eastern plain. (2) Settlements and populations were granted to the royal kinsmen and some Zhou allies to form more permanent structures of Zhou governance on the eastern plain and its peripheral regions – the “regional states.” This practice, which came to be referred to as the “Fengjian” 封建 institution, thus gave the Zhou state a new shape and became one of its most defining characteristics (see below).Although information is not available to determine the actual number of these regional states, twenty-six are reported in the Zuozhuan as having been founded by brothers or sons of the Duke of Zhou, while archaeological excavations have already confirmed the locations of centers belonging to some nine states including Jin 晉,Wey 衛, Guo 虢, Xing 邢,Yan 燕, Qi 齊, Lu 魯,Teng 滕,Ying 應, and two other states that were established or relocated during the mid- to late Western Zhou – Qin 秦 and Southern Shen 申 (Li Feng 2006, 66–76). It is true that the Zhou in the early Western Zhou period were very militant and had indeed constructed the most capable Bronze Age military machine on the surface of the earth in its contemporary time, easily rivaling the mighty Assyrian Empire, the first Iron Age state in Mesopotamia (Lloyd 1978, 188), which reached the height of its military power only about one hundred years after Zhou began to decline in the late tenth century bc. In terms of both the speed of military growth and of the combined scale of the diverse lands and populations it brought under control, the early Western Zhou period (for a least total of seventy-four years) is qualified by any historical standard as a period of “great expansion.” In the east, frequent military campaigns were launched to conquer the indigenous people of the eastern Shandong peninsula, reaching as far as the Yantai region. In the northeast, the Zhou regional state Yan may have joined Zhou royal forces to reach the Liao River region in southern Manchuria. In the northwest, inscriptions (e.g. Xiao Yu ding 小盂鼎) truthfully document Zhou’s conquest of the Guifang 鬼方, Bronze Age groups that had inhabited the highlands of northern Shaanxi and Shanxi since the late Shang. The capturing of new lands and peoples beyond the traditional core of agricultural societies in the middle Yellow River region was the priority of the royal court, and the continuing inflow of materials and people into the Zhou centers helped foster and spread the Zhou elite culture that was centered on the casting of inscribed bronzes of high literary value. Military glory and the royal awards for it were a goal easily attainable for the Zhou elites and a topic of high honor that they cast hundreds of inscriptions to commemorate.

Political development from the mid- to late Western Zhou However, the Zhou were not successful on all fronts. In the south, in the middle Yangtze River region, the Zhou encountered formidable resistance by groups that are at least partly related to 92

The Western Zhou state

the historical Chu 楚 polity. The conflict mounted in two campaigns organized by King Zhao, the last of the early Western Zhou kings. The first campaign, which took place in the sixteenth year of the king, was clearly aimed at Chu, and it seems to have all gone well as military commanders coming back from the campaign cast many bronzes with inscriptions to commemorate the capture of metal in the south, known to have been the main source of copper since Shang times. But the second campaign, which happened in the nineteenth year of the king, went disastrously wrong; not only did the royal Six Armies, nearly one half of the Zhou’s standing military force, vanish in the Han River, but King Zhao himself was killed in a sunken boat. Although there is still a debate about the identity of the main target of this second campaign,4 there seems little doubt that its failure put a sudden stop to early Western Zhou expansion. The closing of the early Western Zhou evidently set the Zhou state on a trend of adjustment and transition starting in the reign of King Mu, which determined the course of Chinese history for the next few centuries. On the borders, Zhou’s position gradually changed from that of an invader who pursued military goals in the time and space that they chose to one of the victim of frequent foreign invasions. For instance, invasion by the Huaiyi 淮夷 groups based in the Huai River region in the southeast already began in the thirteenth year of King Mu (Jinben zhushu jinian, 9). Given the long front the Zhou had to defend over the relatively flat land surface in southern Henan and northern Anhui, the war must have drained the Zhou royal court considerably and hence opened opportunities for tensions to grow along other parts of the Zhou border, particularly in the northwest, which turned out to be fatal to Zhou survival later. Even within the Western Zhou state, signs appeared that the Zhou king was no longer able to effectively meet challenges by ambitious regional rulers, and the inscriptions record a campaign sent out perhaps by King Yi against Zhou’s long-time subject, the regional state Qi in Shandong, provoked by a recent dispute between the Zhou court and Qi. As will be discussed later, the mid-Western Zhou is also the time during which the Zhou royal government was gradually bureaucratized, and a recent study shows that the Zhou ritual system had also undergone significant changes as the practice of combining multiple rites to enhance the political-religious role of the Zhou king gave way to a new ritual system intended to create and maintain internal differentiations of the Zhou elites (Vogt 2012, 350–356). On the societal level, the end of the great expansion also blocked the exit for an ever-growing population of the Zhou elites in the royal domain that could otherwise be translocated to the newly conquered peripheral lands. This change must have underlain some of the major social problems, most obviously dispute over land resources in the Wei River valley, frequently recorded in inscriptions cast by members of lineages involved in such disputes. As I have argued elsewhere, the land-award policy pursued by the Zhou court itself weakened the economic power of the Zhou regime in the long run (Li Feng 2006, 122–126). Of all the changes that occurred during the mid-Western Zhou, one is politically particularly important, but its cause remains a mystery. For unknown reasons, for the first time since the beginning of the dynasty, the normative rule of royal succession was interrupted when King Xiao, a son of King Mu and brother of King Gong, succeeded his nephew King Yih to become the eighth Zhou king. But after King Xiao died, power went back to the line of King Yih, whose son was established as King Yi, the ninth king of the dynasty. It had been hoped that the bronze inscriptions could throw some light on the circumstances surrounding this strange succession. But this is not the case, and we observe no abnormalities in the account of King Xiao’s reign in such long inscriptions as the Lai pan 逨盤, discovered in Meixian in 2003 (Shaanxi sheng kaogu yanjiusuo et al., 2003, 32–33). On the contrary, the evidence seems to suggest that King Xiao’s rule was accepted as perhaps legitimate if not inevitable in the mainstream political thought of the Western Zhou elites (Figure 4.2). 93

Figure 4.2 The Lai pan 逨盤, vessel and inscription. The inscription gives a history of the Shan 單 lineage juxtaposed with the history of the Zhou royal house from King Wen to King Li, the bronze dating to the reign of King Xuan: Provided by Zhang Tianen, Shaanxi Academy of Archaeology.

The Western Zhou state

The late Western Zhou period began with the reign of King Li, deeply troubled in the sense that the continuing decline of royal power in the previous century had brought the Zhou regime to a point of total crisis. In the southeast, the Huaiyi groups renewed their threat to Zhou’s security, seemingly in an immediate response to the rebellion of the ruler of E 鄂 in northern Hubei, as shown by the inscriptions. A recently discovered inscription, the Zhabo ding 柞伯鼎, records a campaign led by Guozhong 虢仲, who also enlisted regional troops led by the ruler of Cai 蔡 in a joint effort to capture a Huaiyi town in southern Henan. The information is corroborated by records in other inscriptions, including the Jinhou Su bianzhong 晉侯蘇編 鐘, as well as in the transmitted texts that all together show that the war with the Huaiyi and related groups must have been a prolonged and very difficult experience for the Zhou court.To the northwest, the “Xianyun” 玁狁 launched frequent invasions of the Zhou settlements in the upper Jing River valley, posing an immediate threat to the Zhou centers on the Wei River plain (Li Feng 2006, 141–192). Facing the new pressures, perhaps in an attempt for the royal house to regain economic strength that had long been weakened by the court’s land-granting policy, King Li ran into major conflicts with the aristocratic lineages in the royal domain. This eventually led to a rebellion in the Zhou capital that suddenly ended the reign of King Li in 842 bc. This was followed by a period of political transition during which Bo Hefu 伯龢父 (known in the transmitted texts as Gongbo He 共伯和) took actual control of the Zhou court until 827 bc, when King Li’s son was established as King Xuan, who brought to the Western Zhou the longest royal reign, forty-six years. The royal house was partly able to regain its strength under King Xuan by translocating a number of polities from the west to the Nanyang basin, hence reopening Zhou’s roads to the south; the most important was what became the state of Southern Shen 申, whose inscribed bronzes were discovered in the 1990s. On the other hand, the Zhou court also strove to improve relations with distant states such as Yan in northern Hebei and Qi and Lu in Shandong. In the northwest, the Zhou were able to hold the highland groups, the Xianyun, back for at least some time, through supporting the rise and expansion of the new state, Qin 秦, in the upper Wei and Xihan River valleys in southeastern Gansu. The length of the reign of King Xuan was itself an indication of sustained political stability in the Western Zhou state, though much weakened and vulnerable in comparison to the early Western Zhou.

Organization of the state At the fundamental level, the state of Western Zhou can be seen as a congregation of thousands of village and town settlements (called Yi 邑 in inscriptions) distributed in the river valleys of northern China and along the edges of the eastern China plain woven together by the political power of the Zhou royal house. These Yi settlements of various sizes, standing in complex relationship with one another, were essential building blocks of Western Zhou society and the loci of Zhou’s state power. Originally, they represented homogenous kin groups which each occupied the residential core of a Yi settlement and cultivated the fields surrounding it, although modifications did occur that inevitably increased the diversity of the Yi settlements’ population. Larger Yi settlements that came to encompass multiple kin groups might become regional centers, usually functioning as the capital sites of the regional states.Very small settlements were hamlets that housed agricultural laborers or slaves that belonged to the hereditary lineages of the Zhou elites.The bronze inscriptions offer examples that small Yi settlements were frequently objects of sale or exchange, or commodities paid for a lost lawsuit or compensation for war between the lineages.

95

Li Feng

For a very long time the Western Zhou state was interpreted as a “feudal” institution and the political ties that bound the Zhou polities together as “feudal” relationships. This suggested an unguaranteed parallel with medieval European “feudalism” (Maspero 1950, 111–146; Creel 1970, 317–387), which was based on the core value of contractual loyalty owed by the vassal to the lord and mutual obligations between the two. Recent studies, however, have invalidated this comparison, which was misconceived by early sinologists in the first place. Instead, we now see a more stable structure of state organization in the Western Zhou in which the thousands of Yi settlements were distributed through the kinship structure of the Zhou royal clan and its collaborating lineages, and this gave rise to a different set of relationships characterized by the unconditional submission of those who received the settlements to the Zhou king as both the head of the state and the head of the royal Ji clan (Li Feng 2003, 115–144). In the western half of the state, namely the Wei River plain and valleys along its tributaries in central Shaanxi and eastern Gansu, old lineages such as Guo 虢, Nangong 南宮, San 散, and Shan 單 already received multiple settlements from the Zhou house even before the Zhou conquest. As time passed, more recent descendants of the royal lineage were added to this list of lineages, such as Zhou 周, Shao 召, Rong 榮, and Jing 井, each in control of multiple Yi settlements from which they derived revenues. On the eastern plain, but also in the Fen River valley of southern Shanxi and the Nanyang basin in southern Henan, the regional states in a large number, mostly granted to brothers or sons of King Wu and the Duke of Zhou, functioned as the regional powers that organized the thousands of Yi settlements into large clusters. There were also non-Ji states, founded by members who were not descendants of the Zhou kings or dukes but might well have been former allies of Shang, incorporated into this state structure through marriage to the Zhou royal house or to its regional branches. Further down south in the Sui-Zao corridor of northern Hubei, a series of archaeological discoveries made since 2010 have revealed the material wealth of the state of Zeng 曾, which claimed its ancestry to the Nangong family in the royal domain in Shaanxi and continued to exist in the region from the early Western Zhou to the beginning of the Warring States period (Figure 4.3). As far as we can tell, no such regional states existed in the western core region, though we do know of a few states in the western periphery such as Qin 秦, which was founded later due to changing political situations in the west. This “bifurcation” was the essential feature of the Western Zhou state. Due to the varying circumstances surrounding their founding and changes over time, these regional states (the elite lineages in the west were similar) did not exist as continuing land blocks but as merely clusters of Yi settlements in which the state existed. There was no linear border that demarcated the state, as in the later “territorial state” (see later). Beyond the royal domain in the Wei River valley in Shaanxi and the small area surrounding the eastern capital Chengzhou in Luoyang, there was no such “territory” that can be called Zhou; instead, there were clusters of settlements that belonged to the regional states, separated from each other by forests, wetlands, and perhaps also some non-Zhou communities that resided among the states, especially when we move towards the periphery of the Western Zhou state. There could well be a case in which a settlement that belonged to a state was located closer to the center of another. The regional centers were linked by roads and pathways to each other that formed a network and were linked altogether to the royal capitals both in the west and in the east. This condition of state organization can be best characterized as a “settlement state,” and it was ordered by and largely through the kinship structure of the royal Ji clan. In this sense, the Yi settlements were both the building blocks and the bedrock of the political infrastructure of the Western Zhou state. The “regional rulers” (Zhuhou 諸侯) ruled their states with political power delegated to them by the Zhou king.They had full rights to all aspects of administration as well as justice in their own

96

The Western Zhou state

Figure 4.3 Bronzes newly discovered in the Yejiashan 葉家山 cemetery, from tomb no. 28 (1–9: 157, 181, 164, 162, 153, 178, 159, 174, 166), occupied by a ruler of the state of Zeng, dating to the early Western Zhou: Provided by Fang Qin, Hubei Institute of Archaeology.

states, but they were not independent sovereign rulers as the Zhou king was, or as the Warring States kings. Being heads of the junior branches of the Zhou royal Ji lineage in Shaanxi that granted their very existence, they owed inborn allegiance and unconditional submission to the Zhou king and were responsible for the security of Zhou state in their regions. By the same token, the power that the Zhou king had to command them and exact their obedience had a legitimate basis that was both legally and religiously sanctioned.5 Although a regional ruler could on occasions choose not to fulfill his obligation to the Zhou state and its king, it was the Zhou king’s prerogative as the head

97

Li Feng

of both the state and the Ji clan to demand its fulfillment by means including the use of force as he saw fit. In fact, more and more new inscriptions suggest that the ties that bound the regional states to the Zhou court were much stronger than scholars used to admit under the “feudal” interpretation, or some other alternative models.6 The Shu Ze ding 叔夨鼎 and many other inscriptions show that regional rulers frequently attended the Zhou king in state ceremonies held in the eastern capital, Chengzhou, or were required to pay homage back in Shaanxi.The Jinhou Su bianzhong 晉侯蘇編 鐘 bells record that the ruler of Jin commanded armies on behalf of King Li and was upon return rewarded by the king as his subject. The recently published Zhabo ding 柞伯鼎, among others, records a campaign led by officials from the royal court in Shaanxi under whom the regional ruler of Cai 蔡 served as a subordinate military officer. These new inscriptions give a vivid sense of the political and military roles that the regional rulers played in the Western Zhou state.The inscriptions also tell us exactly that the Zhou royal court installed hereditary inspectors in many regional states in order to closely watch them on behalf of the Zhou king.

Government and bureaucracy Since very little is known about the governments of the regional states, any account of the Western Zhou government would inevitably take the royal domain in the Wei River valley as its focus. Quite unlike the Shang metropolis – Anyang, on which, to borrow David Keightley’s words, all Shang political-religious energy was focused (Keightley 2000, 58) – none of the Zhou royal centers came anywhere close to the magnitude of Anyang, both in terms of their spatial dimension and of their cultural richness. But they certainly rivaled the Shang capital by numbers – determined by the political and administrative need of the Western Zhou state. Zhou royal power was based on a network of prominent cities, referred to in the bronze inscriptions as the “Five Cities” (Wuyi 五邑), through which the Zhou king frequently traveled and carried out his duty of governance, including making critical administrative appointments. It is likely that, though the accurate counting is unknown, at least during much of the mid-Western Zhou, the “royal city” network must have included Hao 鎬, Feng 豐, Zhou 周, Pang , and probably Zheng 鄭 to the west. We find that sometimes the Zhou king appointed officials with specified responsibility for all “Five Cities” as a tier of Zhou administration. To this network we must also add Chengzhou, the eastern capital in present-day Luoyang that played a key role in Zhou’s control of eastern China. Certain administrative structures, for instance the “Ministry” (Qingshiliao 卿事寮), existed in very early times of Zhou governance, which was a routine bureau that combined key positions in Zhou government: “Supervisor of Land” (Situ 司土), “Supervisor of Construction” (Sigong 司工), and “Supervisor of Horses” (Sima 司馬), although our evidence for the term “Three Supervisors” (San yousi 三有司) dates no earlier than the mid-Western Zhou. This shows that from very early times the Zhou had embraced an approach to governance that was very different from that of the Shang,7 and it is not wrong to characterize this approach as a civil- and practical-administrative orientation. Since the largest group of bronze inscriptions relating to Zhou governance are the “appointment inscriptions” that preserve information about the various administrative appointments the Zhou king made, luckily we can speak about the Zhou government in much more detail than we can possibly do for any government in early China until many centuries later. The inscriptions suggest that by the mid-Western Zhou another prominent structure in the Zhou government was formed, called the “Secretariat” (Taishiliao 太 史寮), headed by the Grand Scribe and staffed by a large number of scribes who kept records and produced written orders for the Zhou central government.

98

The Western Zhou state

While this binary structure was central to the Zhou government, it is evident that during the mid-Western Zhou the royal household (including its various facilities located in the “Five Cities” and beyond) had become a separate body of management headed by the “Superintendent of the Royal Household” (Zai 宰) and was staffed by a large number of functionaries whom the Zhou king literarily called “my officers.” Embedded in the structure of the royal household administration were also multiple scribes who were subordinate to the “Interior Chief Scribe” (Neishi yin 内史尹). In the inscriptions, we not only see these titles but actually see officials with such titles conducting specific business.To add more to these structures were the organizations of the “Six Armies of the West” (Xi liushi 西六師), stationed near the royal capitals on the Wei River plain, and the “Eight Armies of Chengzhou” (Chengzhou bashi 成周八師) stationed outside the eastern capital in Luoyang. By the mid-Western Zhou each of these two army divisions had acquired considerable landed properties and civil population and therefore developed certain civil functions. The structural differentiation described here is based completely on contemporaneous evidence – the bronze inscriptions. The trend towards increased structural complexity and functional stability has been described as the “bureaucratization” of the Zhou government, which doubtless took place during the mid-Western Zhou and continued to intensify through the late Western Zhou. On the other hand, the power to command the Zhou government and the procedure by which it did so also underwent considerable change through the Western Zhou period. During the early Western Zhou, as the central government’s main goal was to mobilize resources to support war under an expansionist agenda, the politics at the Zhou court were characterized by conflicts and negotiation among a few eminent figures who were either uncles or brothers of the Zhou king, including, most famously, the Duke of Zhou and the “Duke of Shao.” During the mid-Western Zhou, for at least one or two royal reigns, many decisions were made by a group of high officials whose names appear in the inscriptions in a fixed order, which indicates the trend of increasing bureaucratic influence on the government’s decision-making process. More importantly, the procedure by which administrative offices were assigned had been typically bureaucratized and subject to the regular “appointment ceremony” (Ceming 冊命), conducted according to strict rules. During the ceremony, the Zhou king would personally meet with (and on a few occasions personally address) the candidates before the appointment letter was read by the Interior Chief Scribe and handed over to the candidates (Li Feng 2008, 107–111). We have more than one hundred such inscriptions from the mid- and late Western Zhou periods that tirelessly recount details of the ceremony and describe the administrative responsibilities of the appointees. Procedures also developed whereby young elites would first be appointed as “Assistants” to senior officials, and after a certain period of service, they would be promoted to senior positions (Li Feng 2008, 190–234). The characteristic of the Zhou model of state was to entrust the governance of regions beyond the royal domain to the hands of the regional rulers who conducted their own governments in their designated geographical settings (see above). But our information about these local governments is poor to say the least. Administrative titles such as the three “Supervisors” that we see on bronzes from the royal domain also appear on regional bronzes, but because they were either little contextualized or were gathered from different states, it is hard to see how they were constructed in a general structure. We have also to be aware that some of these titles might have had different functions in different states. However, recently discovered inscriptions do shed light on offices that are unique to the regional states, such as the “Inspector” (Jian 監), a key position for royal control of the regional states. We have grounds to believe that many states had

99

Li Feng

such officials, and in one case, Southern Shen, located in southern Henan, the Inspector was a brother of the Zhou king (Zhong Chengfu gui 仲爯父簋).

Social order Clan and lineage were two key social institutions of the Western Zhou. It has been suggested that the Zhou were the first people in China to have introduced the institution of clan names (Xìng 姓, or “surname”) as a way to regulate marriage relationships among the various ethnic groups in the Zhou commonwealth (Pulleyblank 2000, 1–27). As such, clans were usually kin groups that descended from a common ancestry, and clan names were derived from the maternal origin of a clan’s distant ancestor. For instance, the Zhou royal house was a member of the Ji 姬 clan, which had traditionally taken women from the Jiang 姜 clan as its primary marriage partner. But marriage between members of the same clan and therefore of the same ancestry was prohibited in principle, despite some rare cases of violations of this principle that we know from the transmitted texts.8 There are a half-dozen other clans regularly appearing in bronze inscriptions that participated in marriage with the royal Ji clan or with one another, including Jiang,Ying 嬴, Ji 姞,Yun 妘, Si 姒,Wei 媯, Ren 妊, Ji 妀, Man 嫚, Cao 曹, Zi 子, etc. Since these names which determined marriage outside the same clan continued in use until the late Spring and Autumn period, the surnames were the most important system of classification of the Zhou and non-Zhou elites. The lineages were social solidarities that held land estates and populations, and lineage names (Shi 氏) were in most cases derived from their central Yi settlement – a large natural village surrounded by fields possessed by the lineage (see above). Each lineage had further under its control numerous smaller settlements where their farmers lived and activities of subsistence were carried out. Rich lineages also held residential quarters in the royal capitals and supplied them with revenues derived from their lineage settlements. Certainly, the largest number of lineages on the Wei River plain were founded by descendants of the royal Ji 姬 clan. Thus, we can speak not only of the “Ji clan,” but also of the “Ji-surnamed lineages.” These Ji and non-Ji lineages competed for social and economic resources in the royal domain, sent their members to serve in the Zhou central government, and supported wars by providing auxiliary forces to the Zhou royal armies stationed near the Zhou capitals. By the late Western Zhou, their power grew to even overshadow that of the Zhou royal house, which had to negotiate policies with the prominent lineages such as Shao 召, Jing 井, Zheng 鄭, San 散, and Guo 虢, Rong 榮, and Shan 單. Lineage segmentation was a process that caused important social changes. Later Confucian texts describe this process in the way that in every five generations, minor sons of a lineage would be required to move away and found sublineages, so the lineage’s growing population could be kept at a manageable level (Falkenhausen 2006, 64–70). Naturally, there was a distinction between the primary lineage (Dazong 大宗) and the derivative ones (Xiaozong 小宗), and the sublineages by their position on the lineage’s genealogical tree owed allegiance to the primary lineage. Outlying settlements, or even the residential quarters in the royal capitals, previously owned by the lineage could easily become central bases of the new-born sublineages. The bronze inscriptions do not confirm the practice of a strict “five-generation” rule, but they offer sufficient evidence for the activity of sublineages with respect to the primary lineage.There is also evidence that the primary lineage represented its sublineages in legal cases that were brought for settlement at the Zhou court. Although social divides were constructed along lineage boundaries in Zhou society, each lineage was itself a social hierarchy internally. Under the practice of primogeniture, only the

100

The Western Zhou state

oldest son (called Bo 伯) in each generation had the chance to become the head of the lineage, while other sons (designated as Zhong 仲, Shu 叔, and Ji 季) formed their individual families and worshiped in the common ancestral temple of the lineage. Nuclear families, if they existed at all, for which we have little proof, must have been inactive socially and economically in a lineagedominated society. Furthermore, bronze inscriptions suggest that a lineage’s population also included two categories of people who were not descendants of the lineage ancestor.The first is called “retainers” (Fuyong 附庸), who were unfree farmers attached to the lineage, although the exact origin of Fuyong is unknown and might have varied from group to group. But it is likely that they included indigenous populations that previously lived in the Zhou periphery or were translocated to the Wei River plain from eastern China. The second category is called Renli 人 鬲, which probably meant “slaves.” We still do not know their exact status, but it is likely they were war prisoners brought into Zhou society during the great early Western Zhou expansion. Because of the important role of lineages in the Zhou sociopolitical system, the challenge was nevertheless how to maintain lineage identity in a situation in which the Zhou elite population continued to grow and amalgamate through intermarriage, and the lineages continued to segment. Recent scholarship uncovered two rules that were important for the continuing functioning of Zhou society which I term “Rules of Name Differentiation I and II” (Figure 4.4). “Rule I” determines the ways in which elite women were referred to in marriage. For instance, a husband called each of his wives by their different clan names identifying their ancestral origins. But as a father he would call his married daughters by each of their husbands’ lineage name plus his clan name, which is certainly also their clan name. In this way a woman always carried her original clan designation, but her social identity in the lineage-based society changed depending on who addressed her. When she referred to herself or was addressed by her sons, her name would be a combination of her husband’s lineage name and her original clan name. As far as we know, this rule was strictly observed in bronze inscriptions, and some of the women referred to by such terms were very important political figures. For instance, Wang Jiang 王姜 (King Kang’s spouse from the Jiang clan) played a prominent role in bronze inscriptions that were cast by officials who received awards from her in the reign of King Zhao. During the mid-Western Zhou, we know that a Superintendent of the Royal Household named Cai 蔡 was appointed with special responsibility to send out orders issued by a Consort Jiang. At the fundamental level, marriage between the Ji and non-Ji lineages (such as lineages of the Jiang clan), whose identity had to always be clearly recorded, was an important institution that bound the Western Zhou state together. “Rule II” concerns the problem of lineage segmentation. When a lineage segmented, it created sublineages whose elite members needed to be differentiated from that of the primary lineage in every generation. A possible way to limit membership in the primary lineage was to adopt the names of the sublineage founders such as “Guoji 虢季” or “Jingshu 井叔” as names of the relevant sublineages, but to what extent this was the rule needs further research to clarify. More often, elite members of the sublineage would (or perhaps were required to) take on names of their mothers’ original lineage as part of their own designation. Therefore, we see names like “Zhousheng 琱生,” “Guosheng 虢生,” and “Fansheng 番生,” which meant that the persons called by these names were nephews of the Zhou, Guo, and Fan lineages, and their own lineage names (in the first case, Shao 召) were therefore concealed. Some of these people, for instance Fansheng, were able to rise to the top of Zhou bureaucracy during the mid-Western Zhou, but they were still referred to in a way that would hide their lineage identity. The frequency of such names in Western Zhou inscriptions suggests that this was an important way to maintain or limit lineage identity in the kin-ordered Zhou society.

101

Figure 4.4 Rules of Name Differentiation I and II. A. Rule governing names of women; B. Rule governing names of elites of sublineages.

The Western Zhou state

Breakdown of royal order and transition to territorial states The problem that finally brought the Western Zhou dynasty to an end in 771 bc had its origin in the generational transition to the reign of King You, who ruled for the next eleven years. A fierce struggle took place between a group of senior officials led by the “August Father” (Huangfu 皇父), who had long served under King Xuan, and the newly rising King You and his supporters. This also brought the Zhou royal court into conflict with polities such as Western Shen 申 and Zeng 鄫, located on the northwestern borders. In 771 bc, the allied forces of the Xianyun, Western Shen, and Zeng marched down the Jing River valley and captured the Zhou capital, killing King You and ending the Western Zhou dynasty. When the royal prince Yijiu 宜 臼, who, persecuted by his father had previously taken asylum in Western Shen, was established as King Ping 平 (770–720 bc), he found it no longer possible to rule in the Wei River valley. This was not only because of the strategic vulnerability of the region, already opened to invaders from the west, but also because of the political rivalry with a nominal king, the King of Xie 攜, established by the lineage of Guo on the Wei River plain. By this time, Guo had already established its stronghold in the narrow pass of present-day Sanmenxia 三門峽 in western Henan, virtually cutting off the Wei River valley from any possible aid from the eastern states. Thus, King Ping decided, with support by the states of Jin 晉 and Zheng 鄭 in the main, to settle his court in what was once the eastern capital of the Western Zhou state, Chengzhou 成周, in present-day Luoyang, thus beginning the Eastern Zhou (770–221 bc) period of Chinese history. The transition to the Eastern Zhou established trends towards deep-running social changes that laid the foundation for the future Qin Empire (221–207 bc). But only recent scholarship has been able to logically explain this transformation to empire, on the basis of a new understanding of the Western Zhou state as its starting point. As clarified earlier, the most important feature of the “settlement state” like the Western Zhou lay in the fact that the state was defined not as an integral territory but as a cluster of settlements that belonged to it; as such the perceived spaces of different states might overlap each other. Such condition of the states was maintained through the Western Zhou period precisely because of the existence of the Zhou royal authority that mediated between them. Due to the complete loss of its economic base in the Wei River valley, the resettled Zhou court in Luoyang was powerless to continue in this role. It was the time for the regional rulers to compete for control of the Zhou court and to use it to pursue their own goals. The political-military conflict of the Eastern Zhou began precisely with the newcomer states like Zheng and Guo that had moved east with the Zhou court and struggled to consolidate their new bases by victimizing smaller polities near the royal capital. Gradually, states in the periphery of the former Western Zhou such as Qi, Lu,Yan, Jin, and Chu were pulled into this battle by their connections with those inner states, thus driving the birth of a new era of interstate warfare and regional conflict. As war continued to increase in scale and affected more areas in northern China, it was the settlements that were located closest to the center of an enemy state that were first conquered and incorporated into the settlement system of the latter, or the states would exchange distant territories because they were strategically vulnerable for the purpose of better geopolitical control. This process resulted in states as tightly packed settlement clusters that could be defined by systematic defenses, and eventually as integrated territorial masses that can be called “territorial states,” the prototype of “empire.” The position of this process in Chinese history is very similar to that of the Thirty Years’ War, the most destructive conflict in European history after the Roman Empire. The resultant Treaty of Westphalia cosigned by all powers in 1648 secured the territorial integrity of the German princes and guaranteed their religious freedoms within their own territorial states. In China, the process of territorial consolidation resulted in a more visible 103

Li Feng

geographical feature – the walls constructed by the states in the fifth to fourth centuries bc.Those which left clear marks on the landscape today included the Long Walls of Qi, Wei, and Qin, and the Square Wall of Chu, etc. (Li Feng 2014, 183–186). Within these states, whose territories became increasingly clear, the rulers were no longer willing to award newly conquered peripheral settlements to their brothers or sons, who by receiving such settlements might become heads of local lineages. Instead, for the purpose of better mobilizing the people and economic resources that these settlements could provide to support the states in war, the ruler would directly send his officials to manage such settlements, thus creating a new type of administrative units called Xian 縣, “county” – a key institution in the transition to “territorial states.” Past scholars have paid attention to the link between the word Xian, “county,” and the word Xuan 懸, “suspension,” which were homophones in the Zhou language, and suggested that the counties were units attached directly to the state capital in the sense that they were directly controlled by officials dispatched by the ruler. Instead, in the new interpretation, the counties were rather “special economic zones” that were suspended in the process of land redistribution of the traditional “settlement states,” to be managed directly by the ruler’s officials (Li Feng 2014, 166–170). The historical record suggests that between 740 and 690 bc King Wu of Chu conquered the small polity Quan 權 and turned it into a county, and this constituted the first example of a conquered polity becoming a county, followed by the former Zhou states of Southern Shen 申, Lü 呂, Deng 鄧, and Xi 息 in southern Henan, which were subsequently incorporated into Chu territory as counties. In the north, Qin 秦 shared the practice as early as the seventh century bc (Creel 1964, 155–183). By the fifth century bc, the practice of creating counties was widely established, as all major states of the early Warring States period held a couple dozen such counties. This is key to understanding a chain of social changes that occurred in Early China and together made the “territorial state” a social reality. In the counties, farmers lived in small nuclear families which included a couple and their children and parents and managed their agricultural production as independent economic units, as the traditional aristocratic lineages were either eliminated in the process of conquest or were purposely suppressed by the new state. Given the economic advantages of the free families, poor members of a lineage would opt to leave it to become free farmers of the new state. The state itself also took active measures to encourage the population migration into the counties and to open virgin lands for cultivation, and the relevant policies effectively served to drain the traditional lineages. As the traditional lineages lost both their economic power and political prestige, young descendants of the lineage who could no longer rely on their hereditary rights turned to seek a living elsewhere, relying on their education as well as noble spirit as warriors.This was the origin of the so-called Shi 士, “scholarwarriors”, who were frequently employed by the rulers to manage the affairs of their new states. Without mediation by the lineages, for the first time the state entered into direct relationship with a mass of people that were engaged in subsistence production. This new relationship between the state and the individual farmers was realized in three ways. The first was taxation. Free farmers cultivated packs of land allocated to them by the state and were obligated to hand in a portion of their gains to the state. The historical record suggests that in 594 bc for the first time the ruler of Lu imposed a tax on land, which the context suggests was a common tax. By the late Spring and Autumn period, as the practice of creating counties was widespread, taxation also became a well-established institution in all major states, and in some states the tax rate is said to have been as high as 20% of a family’s production gains (Li Feng 2014, 191–194). Second, the free farmers were not only subject to taxes, they also owed military service to the state, which awarded them privileges such as tax exemption in accordance with a system of universal ranking.

104

The Western Zhou state

In the state of Qin, seventeen ranks were established, and the entire population was subject to the same meritocratic system.Third, the state promulgated laws in order to regulate the conduct of the free farmers, who no longer lived under the supervision of the lineage heads. On the other hand, the cities, particularly state capitals, grew dramatically during the Eastern Zhou period as the result of population growth and industrial and commercial development, fully evident in archaeology. There were large numbers of free professional groups in all major cities employed in the various state-run or private workshops that produced weapons and tools as well as luxury goods for the new social elites, and the state needed to control them. Thus, we read in the historical record that in 536 bc, the reforming state Zheng first took action to cast legal codes on a bronze ding cauldron, and this was followed by the northern state of Jin in 513 bc. The changes discussed here are a chain of responses that took place in the transition from the “settlement state” to the “territorial state.”War was the initial cause of the changes.Through this transition, early Chinese society was reshaped from the bottom, and the foundation of the future empire was firmly constructed. Empire was the enlargement of the “territorial state” when it conquered its peers.The “territorial state” invented all skills that were needed for the construction of empire.

Notes 1 However, even this date is recently challenged by the Qinghua Manuscripts, which include a document titled “Zhou Wuwang you ji” 周武王有疾 which describes that King Wu became ill (and presumably died) in the third year after the conquest (Li Xueqin ed. 2010–2015, 1: 158). The Qinghua Manuscripts are a corpus of more than 2,000 bamboo strips stolen from a tomb and purchased by Qinghua University in Beijing and published in five volumes since 2001 (Allan 2012, 547–557). 2 The received textual records in Shiji describe Guanshu and Caishu as co-conspirators of Wu Geng in a joint rebellion against the Zhou court in the west, but the Xinian 繫年 chronicle in the Qinghua Manuscript gives them as immediate victims of the Shang insurrection. 3 The Bao you 保卣 inscription mentions the king giving campaign orders to the “Grand Protector.” This seems to have been a continuing military effort to pacify the eastern states upon an attack led by the king himself on Luzi Sheng 彔子聖 (Taibao gui 太保簋), identified as Wu Geng by most scholars, who is apparently the same person as Luzi Geng 彔子耿 mentioned in the Xinian chronicle. 4 Traditional scholars all believed that the target of both campaigns was Chu, but inscriptions suggest that the campaign of the nineteenth year of King Zhao was aimed instead at the Hufang 虎方, a group that might have been located somewhere to the south of the Yangtze River (Li Feng 2006, 93–94, especially note 6 on page 94). 5 In this regard, the nature of the Zhou king’s power is fundamentally different from the nature of kingship in the Shang state, which was organized on very different principles. 6 For instance, the “Segmentary State” model discussed by Barry B. Blakeley (Blakeley 1970, 59–62) and the “Potlatch” model discussed by Constance Cook (Cook 1997, 269–273). Problems with the “Segmentary State” model have been discussed in another place (Li Feng 2008, 278–280, 290–293). The validity of the “Potlatch” model is based on the understanding of the Western Zhou (at least the early Western Zhou) as a pre-state society, posed somewhere in the transition from a “complex chiefdom” to a “state” (Cook 1997, 290). However, this view is largely invalidated by new evidence. 7 The Shang had certain positions like the “Document Maker” (Zuoce 作冊) which the Zhou continued to have, but the Shang never had any routine government structure like the Zhou “Ministry” or its Three Supervisors. 8 For instance, the Mu tianzi zhuan 穆天子傳, excavated from a Warring States tomb in northern Henan in the third century bc, mentions that King Mu took as his consort a lady named Sheng Ji 盛姬, who by name rules (see above) was a woman from the Ji clan (Mu tianzi zhuan, 6: 74).

105

Li Feng

Works cited Allan, Sarah (2012). “On Shu 書 (“Documents”) and the Origin of the Shang shu 尚書 (“Ancient Documents”) in Light of Recently Discovered Bamboo Slip Manuscripts,” Bulletin of the School of Oriental and African Studies 75.3: 547–557. Blakeley, Barry B. (1970). “Regional Aspects of Chinese Socio-Political Development in the Spring and Autumn Period (722–464 B.C.): Clan Power in A Segmentary State,” Ph.D. Dissertation: University of Michigan. Cook, Constance (1997). “Wealth and the Western Zhou,” Bulletin of the School of Oriental and African Studies 60.2: 253–294. Creel, Herrlee (1970). The Origins of Statecraft in China, Vol. 1: The Western Chou Empire. Chicago: University of Chicago Press. Creel, Herrlee (1964). “The Beginning of Bureaucracy in China:The Origins of the Hsien,” Journal of Asian Studies 22: 155–183. Falkenhausen, Lothar von (1993). “Issues in Western Zhou Studies: A Review Article,” Early China 18: 139–226. Falkenhausen, Lothar von (2006). Chinese Society in the Age of Confucius (1000–250 bc): The Archaeological Evidence. Los Angeles: Cotsen Institute of Archaeology, UCLA. Hornblower, Simon, Antony Spawforth, and Esther Eidinow (eds.) (2012). The Oxford Classical Dictionary. Oxford: Oxford University Press. Jinben zhushu jinian (1920). 今本竹書紀年(Current Bamboo Annals). Shanghai: Shangwu yinshuguan 商務印書館. Keightley, David (2000). The Ancestral Landscape: Time, Space, and Community in Late Shang China. Berkeley, CA: Institute of East Asian Studies. Li, Feng 李峰 (1991). “Xian-Zhou wenhua de neihan jiqi yuanyuan tantao” 先周文化的内涵及其淵源 探討, Kaogu xuebao 考古學報 3: 265–284. Li, Feng (2003). “ ‘Feudalism’ and Western Zhou China: A Criticism,” Harvard Journal of Asiatic Studies 63.1: 115–144. Li, Feng (2006). Landscape and Power in Early China: The Crisis and Fall of the Western Zhou, 1045–771 bc. Cambridge: Cambridge University Press. Li, Feng (2008). Bureaucracy and the State in Early China: Governing the Western Zhou, 1045–771 bc. Cambridge: Cambridge University Press. Li, Feng (2010–2011). “The Study of Western Zhou History: A Response and a Methodological Explication,” Early China 33–34: 287–306. Li, Feng (2011). “Literacy and the Social Contexts of Writing in the Western Zhou,” in Writing and Literacy in Early China: Studies from the Columbia Early China Seminar, ed. Li Feng and David Prager Branner, pp. 271–302. Seattle: University of Washington Press. Li, Feng (2014). Early China: A Social and Cultural History. Cambridge: Cambridge University Press. Li, Xueqin 李學勤 (ed.) (2010–2015). Qinghua daxue cang Zhanguo zhujian 清華大學藏戰國竹簡, 5 volumes. Shanghai: Zhongxi shuju 中西書局. Li, Xueqin, and Liu Guozhong (2010). “The Tsinghua Bamboo Strips and Ancient Chinese Civilization,” Journal of Chinese Philosophy, supplement to volume 37: 6–15. Lloyd, Seton (1978). The Archaeology of Mesopotamia: From the Old Stone Age to the Persian Conquest. London: Thomas and Hudson Ltd. Maspero, Henri (1950). “Le régime féodal et la propriété foncière dans la Chine antique,” in Mélanges posthumes sur les religions et l’histoire de la Chine, III: Études historiques, pp. 111–146. Paris: Musée Guimet. Mu tianzi zhuan 穆天子傳 (Sibu congkan 四部叢刊 edition, 1922). Shanghai: Shangwu yinshuguan 商務印書館. Nivison, David (1983). “The Dates of Western Chou,” Harvard Journal of Asiatic Studies 43: 481–580. Pankenier, David (1995). “The Cosmo-Political Background of Heaven’s Mandate,” Early China 20: 121–176. Pulleyblank, Edwin (2000). “Ji and Jiang: the Role of Exogamic Clans in the Organization of the Zhou Polity,” Early China 25: 1–27. Shaanxi sheng kaogu yanjiusuo 陝西省考古研究所 et al (2003). “Shaanxi Meixian Yangjiacun Xi Zhou qingtongqi jiaocang fajue jianbao” 陝西眉縣楊家村西周青銅器窖藏發掘簡報, Wenwu 文物 6: 4–42. Shaughnessy, Edward L. (1981–1982). “ ‘New’ Evidence on the Zhou Conquest,” Early China 6: 57–79. Shaughnessy, Edward L. (1991). Sources of Western Zhou History: Inscribed Bronze Vessels, Berkeley, CA: University of California Press.

106

The Western Zhou state Shaughnessy, Edward L. (2006). Rewriting Early Chinese Texts. Albany: State University of New York Press. Shiji 史記 (1959). (The Grand Scribe’s Records) (by Sima Qian 司馬遷), 10 vols. Beijing: Zhonghua shuju 中華書局. Van De Mieroop, Marc (2004). A History of the Ancient Near East, ca. 3000–323 bc. Malden, MA: Blackwell Publishing. Vogt, Paul Nicholas (2012). “Between Kin and King: Social Aspects of Western Zhou Ritual,” Ph.D. Dissertation, Columbia University. Wu, Zhenfeng 吳鎮烽 (2012). Shang Zhou qingtongqi mingwen ji tuxiang jicheng 商周青銅器銘文暨圖像集 成, 35 volumes. Shanghai: Shanghai guji chubanshe. Zhushu jinian 竹書紀年 (Pingjinguan 平津舘 edition, 1885). Tianjin: Sunshi Pingjinguan 孫氏平津舘.

107

5 THE AGE OF TERRITORIAL LORDS CHEN SHENTHE AGE OF TERRITORIAL LORDS

Chen Shen

The age of transformation The Eastern Zhou period (770–221 bc) was an era when hundreds of aggressive but ambitious lords of Zhou’s vassals transformed themselves into kings of a few powerful states. These rulers were supposed to govern their territories under the kings of Zhou. But these lords of territorial states exercised their extreme powers for purposes of forming their formidable military, building their strong economic positions, and implementing politically strategic reforms, which would allow them to expand, to compete, to conquer, and to survive. During the span of 550 years, the numbers of territorial states tragically, as well as expectedly, were reduced to a handful of rivals, notably the Seven States (Qin 秦, Qi 齊, Wei 魏, Han 韓, Zhao 趙,Yan 燕, Chu 楚). Decades-long conflict would lead to eventual unification, establishing the first dynastic government of imperial China. It would be this governing structure that would remain in existence for the next two thousand years. This chapter will focus on how the Eastern Zhou period became fractured and divided into independent states, and the consequences of rapidly changing social structures. The Western Zhou (1045–771 bc) ruled for about three hundred years, until Zhou King You Wang was captured and killed by Xi Rong, a nomadic tribe who invaded from the northeast, and made alliances with Zhou vassals. King Ping Wang, the successor, the first king of the Eastern Zhou dynasty (770–256 bc), who had to relocate the court to Zhou’s secondary political center, which was about 500 kilometers to the east. The following 500-plus years were ruled by twenty-five kings, who would reside in present-day Luoyang. However, the second half of the Eastern Zhou dynasty saw a decline and eventual collapse of kings, which gave rise to several influential lords of the vassal states gaining political power. In competition to control existing territories and resources, states became semi-independent from Royal Zhou rulers, whose influence fell into decline, eventually leading to ambitious lords seeking further territorial expansion or alliances. It was the Eastern Zhou period that completely transformed Chinese societies (Hsu 1965, 1999; von Falkenhausen 1999, 2006; Lewis 1999). During the process of territorialization, the authority of Zhou kings was not only challenged but was purposefully ignored.Territorial states were able to increase their political influence by engaging in war and implementing persuasive

108

The age of territorial lords

economic reforms. However, Zhou bureaucratic and hierarchic ritual systems, which were supposedly primary governing mechanisms, were now completely broken down. As a result, the Eastern Zhou began to develop into a multi-faceted and multi-layered society with competitive states in play. Talented individuals and intellectuals were able to seek out new opportunities to serve the lords of these vassal states instead of serving kings of the Zhou Kingdom. At this time, social transformation was stimulated because of, but not limited to, the following factors: (1) urbanization and increased population through migrations of farmers who were freed from privatized landownership; (2) commercialization and trade in competing and compromising modes between state industries and private productions; (3) innovations in technology, such as iron production that increased productivity in the area of agriculture as well as efficiency in war engagement; and (4) freedom of philosophical thinking and persuasion that changed state rulers’ governing doctrines and raised the ambitions of territorial lords. In the age of transformation, parts or all of these efforts moved towards one single goal, that is, to re-establish the order of hierarchic systems and to rebuild a unified China that shared common cultural grounds. The State of Qin from West China unexpectedly achieved this goal, through stiff competitions and bloody conflicts against its rivals.

Sources The following chapter intends to distinguish itself from previous studies in new and exciting ways of interpreting Eastern Zhou society by integrating three components of evidence: classic texts, archaeology, and museum artworks. Substantial evidence is integral in providing better possibilities of storylines in cross-reference to each other, in order to arrive at supportable information on social transformations in Eastern Zhou. For example, one’s interest could lie with bronze vessels from many existing museum collections that can contribute to new interpretations. Bronze vessels, similar to the one at the Royal Ontario Museum shown in Figure 5.1, are often treated primarily as artworks, which hold many ritual symbolisms. However, they must be, and should be, studied for their archaeological context through surviving inscriptions and context (Zhu 1995; Ma 2002; Chen 2004, So 1995; Li 1985). Figure 5.1A shows one of a pair of hu wine vessels which were reportedly unearthed in the 1930s from noble tombs at Jinchun near Luoyang, once a royal city of Eastern Zhou. The two vessels are now housed at the Royal Ontario Museum in Canada and National Museum of China, respectively. The 25-characters inscribed on the neck (Figure 5.1B) reveals that the highly appraised ritual vessel belonged to the lord of Linghu 令狐 jun 君 monarch. Linghu was a fief of the State of Jin 晉 during the Spring and Autumn period. However, Linghu later fell within the State of Wei in the Warring States period, recorded in the texts of both Zou Zhuan 左傳 and Guo Yu 國語. Some scholars speculated that the ruler of Linghu jun might have served as a high officer in the royal court at Luoyang. Conventionally, the study of Eastern Zhou is usually divided into two sub-periods: the Spring and Autumn (Chunqiu) period and the Warring States (Zhanguo) period. A history of Eastern Zhou’s political and military events has always been a fascinating subject in Chinese scholarship. Over the decades, researchers have extensively explored over two thousand years of annotated records and evaluated preserved classic texts. Gu and Zhu (2001) and Yang (1998) list at least forty-two classic works as primary historic sources that range from the contemporaneous records in the form of essays or chronicles, military strategies, medical books, geographic, literature, and ritual works, etc., to thematic summaries of history prior to the third century ad. Li (2008: 19–41) noted sixty categories of preserved classic texts that existed prior to the

109

Chen Shen

Figure 5.1  A. Hu ritual wine vessel, Bronze, Eastern Zhou, sixth century bc, h. 27.4 cm. B.The inscription from the neck of the vessel. 933.12.76 The Bishop William C. White Collection, Royal Ontario Museum

Qin dynasty (221 bc – 206 bc). Some important ones are the following; Zuo Zhuan 左傳(Zuo’s Annotation to Chunqiu), Shiji 史記 (Records of a Grand Historian), Zhushu Jinian 竹書紀年 (Bamboo Annals), Guo Yu 國語 (Discourses of the States), Zhanguo Ce 戰國策(Stratagems of the Warring States), and Zhouli 周禮 (Rites of Zhou). These classic texts are probably among the most frequently cited (Lewis 1999: 588–593; Gu and Zhu 2001: 3–14; and Yang 1998: 20–30). Early twentieth-century scholarship surrounding reconstructions of Eastern Zhou society has been primarily based on the previously mentioned textual analyses (e.g., Tong 1946; Fang 1949). The rise of scientific archaeology in China and the study of unearthed antiquities in the first half of the twentieth century significantly changed traditional narratives. Some attempts have been made to reinterpret the Eastern Zhou based solely on new archaeological evidence (von Falkenhausen 1999, 2006; Yang 2004; Teng 2002); others have presented new perspectives combining historic records and archaeological data (Li 1985; Li 2014; Yang 1998; Gao and Shao 2005; Chen 2005). Eastern Zhou archaeology became a new subject only during the second half of the twentieth century, particularly after the 1980s.Two monographs summarizing

110

The age of territorial lords

Figure 5.1  (Continued)

comprehensive data sets of Eastern Zhou archaeology (Zhongguo 2004; Zhao and Guo 2004) were made available in the Chinese language only after the publication of the Cambridge History of Ancient China in 1999 (Loewe and Shaughnessy 1999). Archaeological data provides additional information about shifting societies. But it is a varying collection of data that makes scholars feel ineffectual, since information cannot be gained in some degree of accuracy from either archaeological data or historic texts alone. For example, in South China, a small state called Sui 随 was consistently mentioned in historic records. In 1976, the tomb of Zeng Houyi (State of Zeng Marquis of Yi) was discovered at Leigudun, Hubei province. The most famous discovery at the site was unearthing the largest complete set of bronze suspended bells – sixty-five pieces in total. However, this amazing find generated much debate around the question of whether the archaeological State of Zeng 曾 be recorded as the historic State of Sui (see Shi 1979). This territorial dispute remained unresolved for decades, and it was only until recently that this mystery would be resolved. In the beginning of the twenty-first century archaeological excavations at three cemetery sites, Yejiashan, Guojiamiao, and Yidigang, all in Hubei province, belonging to lords of Zeng, yielded substantial evidence from bronze inscriptions, confirming that the State of Zeng must have been renamed “Sui” in texts, thus leaving no mention of “Zeng” in the preserved classics. Together with the material from the previous findings from the Leigudun tomb of Zeng Lord Yi, these new archaeological sources revealed the unknown history of Zeng between the ninth century bc and fifth century bc, with a sequence of lord names that were associated with the Ji clan (Ji is the original clan of

111

Chen Shen

Royal Zhou). It further suggests that – archaeologically – Zeng, or “Sui” in text, was in fact a military stronghold state that could challenge the superpower State of Chu in the south (Fang and Wu 2015). Zhang Changping, a bronze specialist and a professor at Wuhan University, has systematically analyzed a few hundred bronze vessels that were historically collected and now reside in museum collections alongside ones recovered from archaeological sites. Zhang arrived at the conclusion that the territory of Zeng was larger than it was previously thought, covering the landscape in the north of Hubei province and the west of Henan province (Zhang 2007). Zhang further suggests that Zeng’s political center could have been at Anju of Hubei during the early Spring and Autumn period. Near Anju, a wall enclosure, 190 meters by 170 meters, was detected by remote sensor technology in 1997 (Zhang 1998). During the sixth–fifth century bc, the Zeng state progressively advanced to the East and possibly “faced-off ” with the State of Chu. Zhang (2007: 24) implies that the area near the Jiuliandun site, where a noble general of the Chu state and his wife were buried, which was excavated in 2002, must be within the borderline of the Zeng state. The question remains, why were the general and his wife living, and then eventually buried, in this area in the first place? It can only be that this was once part of Zeng’s territories which was conquered by the State of Chu in the fourth century bc. Zhang’s scholarship demonstrates the importance of re-interpreting inscribed bronzes housed in museum and private collections around the world as compared to archaeological materials. Thus, documentation represented by bamboo, silks, and bronzes where ancient scripts were preserved, in association with archaeological context, are of great significance. The preserved documents are best seen on inscribed bronzes that were exclusively commissioned and used by aristocrats; so were surviving bamboo books. Recent studies on bamboo strips (such as the Qinghua University Collection, the Shanghai Museum collection, and the Guodian and Zhangjianshan burial collections) have resulted in scholarly agreements on changes in perceptions on Eastern Zhou societies (Barbieri-Low and Yates 2015; Cook 2012; Chen 2013; Li 2008). What is noteworthy is a collection of bamboo strips, which is currently in the Shanghai Museum. The bamboo books were looted from a Chu state noble tomb and eventually turned up in a Hong Kong antique market in 1994. With state approval, the Shanghai Museum purchased the collection. The Shanghai Museum took over three years to fully conserve them, followed by another ten years of research and interpretation.The first volume, Warring State Bamboo Books Collection of Chu State from Shanghai Museum, was published in 2001, which included images and annotations. In 2012 the ninth volume was completed (Ma 2012). This collection consisted of over 1200 strips (both fragments or complete), including a total of over 30,000 words that were excerpts from missing Warring States books of literature, philosophy, history, and political treatises. The essays were sorted into 100 categories, but only fewer than a dozen can be matched to the classic books known today. Many documents are complementary to the storyline of existing texts and further have brought up unknown but significant historic events, figures, and places. For example, the essay entitled Jing Gong nue from the collection tells the story of Lord Jing Gong, who ruled the State of Qi from 547 bc to 490 bc. Lord Jing was ill for over a year; he was covered in scabs and at times ran a high fever. A few of his close ministers were debating who(s) should be blamed or take responsibility for the lord’s poor health, suggesting that supplicants and court astrologers be executed. But Lord Jing decided to change his deceitful ways and decided not sacrifice his astrologers. Once Lord Jing did this, within fifteen days he quickly recovered from his illness, going on to govern his realm. This event was well documented in texts of Zuo Zhuan (Zuo’s Annotation to Chunqiu) and Yanzi Chunqiu (Annals of Master Yan). However, for over two thousand years, scholars have argued about what this mysterious and strange disease 112

The age of territorial lords

was that possessed Lord Jing Gong. Now, with the discovery of Shanghai Museum bamboo collection, this essay solves the puzzle; the disease was called nue (Pu 2007: 159–162).

Dates One way to unfold facts and changes that occurred during this time in history is to verify sequences of dates that are crucial to this period. Most discussions on dates have centred on divisions between the Spring and Autumn period and the Warring States period. Different textbooks and studies have used one of the following four dates to divide the two periods: 481 bc, 475 bc, 453 bc, and 403 bc. Each date has its own specific merits and measures because of particular events in that year, which commemorated a historical change (see later). In fact, it was the fifth century bc itself that marked a time of a progressive transition to the total independence of a few strong territorial lords from the royal court. It was a time when they were free from ruling masters and authorities (e.g., establishment of a state’s own bureaucratic systems and/or use of local currency, etc.). When faced with rapidly changing political strongholds during the period of the Warring States, Zhou kings would eventually lose complete control of their once strong blood-bounded lordships. In 770 bc, King Ping Wang was rushed out of the old capital near present-day Baoji of Shaanxi province by a military escort from the four states Jin 晉, Zheng 鄭,Wei 衛, and Qin 秦.The court was re-established in present Luoyang, beginning the new era of the Eastern Zhou. For the next sixty years, both lords of the State of Jin and the State of Zheng gained more attention and authority at court, which allowed them to call upon alliances of states in the name of the kings.They also established strong political ties and powerful military alliances among the states. In 704 bc, the lord of the Chu State, at the time a small state the south of the Yangtze River, declared himself “king” in title, abandoning his Zhou official rank of Viscount for “King Wu Wang of Chu”. These self-entitled claims of “kings” of their land continued throughout the existence of this state. This proclamation signified total disrespect of royal authority by the far south. The next lord who made such a self-declaration to be a “king” was that of the Wei State in 369 bc. The year of 722 bc was the first year on the official records of the State of Lu 魯, namely Chunqiu (Spring and Autumn).These celebrated classic documents 242 years of Lu state chronicles, state affairs associated with other states’ lords and royal lineages. Chunqiu was said to be a primary source for the teachings of Confucius (also known as Kong Qiu 孔丘) (551–479 bc). The end of the Spring and Autumn period was also the end of the chronicles – 481 bc. In that year, according to Chunqiu, the Qi State Left Primary Minister Tian Chang (also known as Chen Heng) destroyed his opponent, the Right Primary Minister Jian Zhi, and later killed his lord, Jian Gong. From then on, the Clan Tian family began its control over the government of the Qi state. In 386 bc, the head of Tian Clan in the State of Qi was given the noble title of Marquis, thus officially replacing the blood-lines of the original nobles previously held by the Jiang Clan since the eleventh century bc. The year of 475 bc was the first year of the reign of the Zhou King Yuan Wang. Although the year had no major events documented, it would see the first new king ascend to the throne in Luoyang’s court after the end of the Chunqiu chronicle. Scholars who invested their studies heavily in royal reign divisions regarded this year as the beginning of Warring States period. In the State of Jin during the fifth century bc, six Qing-ranked (or Dafu; see later) ministers/ families began to share authority and power with their lords to govern the state. However, this in turn transformed into bloody conflicts and infighting over land ownership and supremacy. In 453 bc, three ministers out of the six, Zhao, Wei, and Han, formed an alliance to kill the most 113

Chen Shen

aggressive and powerful noble, Zhi, after the other two (Fan and Zhonghang) fled out of the state. The three nobles, Zhao, Wei, and Han, negotiated among themselves to divide the state territories and estates into three constituents.This event in the Eastern Zhou history is regarded as “Dividing the Jin by the Three Families 三家分晉”. For the next few decades, the actual power of the Jin State was held by the families of Zhao, Wei, and Han, whereas the lords of Jin were considered insignificant. Military commands were shared by the three ruling families. In 404 bc, since Qi’s court was in chaos, the Zhou King asked the three families to deploy military personnel to go against the State of Qi. The three took their massive armies and crashed into the defensive Long Wall of Qi (or Changcheng – an earlier form of the Great Wall), capturing the Qi state lord. It was a remarkable advancement that was recorded in classic texts. This event was also recorded in the inscriptions on a set of fourteen bronze bells that were commissioned by a general named Biao Qiang, who led the battle and celebrated the victory. The bell set, which is now housed in the Royal Ontario Museum in Canada and the Sumitomo collection of Sen-oku museum in Japan, hold great historical significance, revealing the proven storyline in finely carved calligraphy (Figure 5.2A–5.2B). Recent analyses on two chapters of Xi Nian (系年) from the Warring State Bamboo Strip Collection at Qinghua University verify this important event. The text also suggests that allied armies of the three lords (Zhao, Wei, and Han) had launched military assaults against the states of Qin and Chu that same year before attacking the state of Qi. The facts of this historical event have only survived through the inscription on the bronze bells and bamboo strips. In 403 bc, given as reward for their military services, or probably feared by their growing powers, the Zhou King was forced to knight the three ministers, elevating them to the status of Marquis of Zhao, Marquis of Wei, and Marquis of Han. The State of Jin continued to exist as a minor piece of territory until 376 bc, when the rest of the region was partitioned by the Marquises Han, Zhao, and Wei. With the power already in their hands, this was the ultimate rise of the three out of the Seven States before the unification. Entering the fourth century bc, the states with strong military power had to endure stiff competition. What began was a series of reforms to strengthen their economy and military, as well as establish more aggressive territorial advancements. In 356 bc, the State of Qin, at its weakest moment, accepted proposals from a statesman named Gongsun Yang (390–338 bc) to begin the reform.Yang began a campaign to reduce the privileges of noblemen and to provide new opportunities for commoners and slaves. It was now time for the average man to retain guaranteed rewards from their military service. By 341 bc, less than twenty years since Yang’s reform, the Qin army was stronger than ever. Yang led his men into a decisive battle against the State of Wei, Qin’s long-time rival. A year later,Yang was given the title of “Jun gentleman of Shang”, with a fief of fifteen towns, thus historically known as “Shang Yang” 商鞅. A kind of architectural tile, engraved with the character of “Shang” was probably used on Jun Shang’s official residences or temples, and was uncovered from archaeological sites in a mountainous area of Mount Qinling. Now this previously uncertain fiefdom has been confirmed as existing in the south of the Shaanxi province (Figure 5.3) (Shen 2010). In 337 bc, the year when Lord Xiao Gong of the Qin state passed away, his son was declared “King”, thus known as King Hui-wen Wang of Qin. This self-proclaimed designation was immediately acknowledged by the states of Chu, Zhao, and Han for their own benefit. Three years later, the lords of Qi and Wei met in Xuzhou and acknowledged each other using the selfpromoting status of “King”. By then, the Zhou king had completely lost his royal authority and was relegated to the status of a common man in a royal residence. In 256 bc, Qin’s army advanced to Zhou’s royal territory and took over thirty-six towns with a population of 300,000. It was the year when the last Zhou king died, and the lineage 114

Figure 5.2  One of a pair of Biao Qiang bells (A) and the rubbing of its inscription (B). Bronze, Eastern Zhou, fifth century bc, h. 26 cm and 21.6 cm.The two bells are part of a 14-piece set unearthed from Luoyang in the late 1920s. The inscription reveals that the owner of these bells was a general of the Han State who was involved in a battle that took place in 404 bc in which the State of Qi was invaded.The battle was also mentioned in the records of Guben Jinian (古本紀年) and Lüshi Chunqiu (呂氏春秋). 930.21.136 The Bishop William C. White Collection, Royal Ontario Museum

Chen Shen

Figure 5.3 Roof tile with character Shang. Earthenware, fourth century bc; diameter 15.5 cm. Displayed on the tile is the moulded character Shang, which indicates that it was used on the roofs of residences, temples, or official structures that belonged to Shang jun (d. 338 bc), a reformer and the most influential figure in the Qin state. Recovered in a mountainous region about 300 kilometres south of the Qin capital Xianyang, it is the first archaeological evidence verifying the historical record that Gongsun Yang was given the noble title jun and a territory of fifteen cities included in his fief by Duke Xiao in 340 bc. Shangluo Municipal Museum, China

was no longer needed. This officially ended the era of the Zhou government that had ruled since 1045 bc. In the following thirty-five years before the unification, the heartland of today’s China was a place covered with battlefields and smelling of blood due to the continuous wars among the Seven States (Qin, Qi,Yan, Chu, Zhao, Wei, and Han) and their alliances. In 230 bc, Qin terminated the first of the Seven States, Han, beginning the ultimate fight for unification. In 221 bc, the army of Qin conquered the last state, Qi. China began its unified Qin dynasty with its capital located at Xianyang, just north of today’s city of Xian, famous for its terracotta army world heritage site.

People and social hierarchy The social order of Western Zhou was based strictly on its societal hierarchy: the status of its people, bureaucratic, and clan systems (Shaughnessy 1991; Li 2014; Hsu 1965). Such hierarchical systems continued in the Eastern Zhou only to be challenged and reformed by rising and ambitious territorial lords and upper-middle-class aristocrats. Li (2003, 2014: 128–132) correctly pointed out that Zhou’s social-political system should not be simply addressed as “feudalism” in the Medieval European concept. Zhou’s political system, known as Fengjian, according to Li, refers to the establishment of vassal states under complete control of the Royal Zhou rulers and in accordance with noble ranking for size and scope of benefices. This is particularly important 116

The age of territorial lords

in understanding the ranking order of the Zhou nobility system and networks of lordships to be described later, not to be confused with titles literarily equivalent to medieval Europe. People in Zhou society were divided into ten classes, according to records in Zuo Zhuan (Zuo’s Annotations to Chunqiu) and Guo Yu (Discourses of the States). At the top was Wang 王, which literally was the King of Zhou, who owned all his people and all lands that were governed on behalf of Wang by nobility within the Fengjian system.Within 515 years under the Zhou ruling, there were twenty-five known kings. The nobility consisted of the next three classes: Gong, Dafu, and Shi. Gong 公 (lord, sometimes known as Hou 侯， or in general Zhuhou) are the highest ranked aristocrats, within which were five ranks roughly equal to a medieval European nobility: duke, marquis, count, viscount, and baron, although some details used in the terms translated for Chinese social ranks must be reinterpreted. The rulers of vassal states are Gong, but their ranks vary. For example, the State of Qin was established in the year 770 bc as a reward to the ruler of the Qin clan who escorted the Zhou king. The lord was then recorded in the text as Qin Xiang Gong, the title composing of a “Qin” referring to the state, “Gong” as his status, and “Xiang” as his respectful temple name. Lord Qin Xiang Gong was promoted from Dafu status, one class lower (see later) and entitled to the rank of count.The State of Jin was established in the eleventh century bc with the rank of marquis, and the first marquis was the younger brother of King Cheng Wang (r. 1132 bc–1083 bc). Aristocrats in the next two classes, Dafu 大夫 (Magnate) and Shi 士 (Gentlemen), received their allotment from the lords of the states they served. Like those in the Gong class above them, their titles were hereditary. Dafu sometimes were also called Qing 卿，depending on how it was used in different states and times (see earlier), or as Qing Dafu. People with Dafu status were likely descendants or siblings of Gong, serving higher officials in vassal state governments than those in Shi. But in the Eastern Zhou, its instability of aristocracy became increasingly possible given changes in power and lineages. Taking the famous family history of Confucius, for instance, he was born into an aristocratic family in decline. Confucius’ great-grandfather was Dafu, serving the State of Song. Due to conflicts in the Song Court, Dafu Kong was on the losing side, and he had to flee to the State of Lu. There, Confucius’ father descended to the rank of Shi, as he served only as a warrior associated with a higher status Dafu family. Although he inherited the Shi status in the State of Lu, Confucius could not convince the lords of various states to give him a position where he could advocate his governing philosophy or influence state policy. Shi is a special social group necessitating a combination of intellectuals and warriors who often served as assistants and secretaries to Dafu state ministers. Shi formed a fundamental base of Zhou’s aristocratic society. They were well educated in the Arts of Six (Liu Yi): rite, music, archery, equestrian, literacy, and math. However, Shi played an influential role in changing Eastern Zhou, because it was this group of intellectuals who voiced their opinions for reform, competed for strategic governance, advocated philosophic doctrines, and embraced innovations. In the age of transformation, lords of territorial states started appreciating and promoting the talents from this group and accepted their new ideas for surviving strategies. It is apparent that a substantial part of the population in Shi, who had opportunities to rise from the common classes, became a driving force in readjusting the disordered Eastern Zhou society. Between aristocracy at the top and slavery at the bottom, commoners emerged in Eastern Zhou with their varying skills during a fast-moving urbanization and commercialization process (detailed later). These classes, called Shuren 庶人 (citizens), Gong 工 (craftsmen), and Shang 商(traders), survived independently within society and were beneficiary to reform policies in various states. Individuals were allowed to enter the Shi class, or at least enjoy the freedom of increased private properties. However, with such strict social hierarchies in place, commoners 117

Chen Shen

who became affluent and wealthy would never be treated as respectable citizens.They employed physical skills and craftsmanship but were not allowed to enjoy educational freedom or free thinking. The hierarchical system seemed to work well in the Western Zhou. To restore such an ideal ordering system was one of Confucius’ teaching philosophies. However, in the Eastern Zhou, borderlines of social status became blurred as each of the states tried to exercise their own rites and classification. Imposed social ranks forced restrictions on individuals, as well as allowed privileges that were exclusive to each class. Even in this energetic society, with its general trends of classification rooted in spiritual burial rites, information can be clearly verified through archaeological data. For example, in the ideal and orderly world, burials having a set of seven ritual bronze ding-tripods as funerary goods are restricted for ownership by lords in Gong status, while those having a set of five must have been for Dafu, and a set of three or one for Shi. Over decades of fieldwork, about 10,000 Eastern Zhou burials belonging to individuals of the State of Chu have been excavated and reported (see Guo 1995). These burials are divided into five categories in rank (Zhongguo 2004), based on a combination of burial size, structural layout (e.g., where or not a ramp and/or mounts were present), employment of multiple coffins verses a single coffin, use of ritual objects, and size and scope of burial goods. In ideal circumstances, the distinction of cultural materials would signify correlations to the ranked members of society. Interestingly, von Falkenhausen (2006: 326–400) suggests that burials of the Chu State demonstrate no definite or clear distinction between each ranked elite member but rather two kinds of burials, one with “special assemblage” and the other with “ordinary assemblage” that could represent the separation of two groups – those who are rulers and those being ruled.Thus, current archaeological data could no longer reflect the realities of social status at this time. The fact that confusing indicators of social stratification constantly occurred in archaeological data suggests something in the way of challenging Zhou’s authority and points to changes about to come.

Territorial states and their lords It is difficult to assess exact numbers of vassal states that existed in the Eastern Zhou period because of the emergence and disappearance (by termination or relocation/renaming) of vassal states which occurred during the different periods of Zhou’s nearly 900 years of ruling governments. Fang (1949: 173) counted between 140 and 150 states during the Spring and Autumn period, while Bai (1994) noted 148 states that were reduced from about 800 to 1000 states which were established in the early part of Western Zhou. In Bai’s findings he believed that, moving forward within the Warring States period, only less than 20 states prevailed as noted in the classic texts. The first attempt at comprehensive analyses of state names and their lords was done by an eighteenth century scholar, Gu Donggao (1679–1759). Gu’s work, entitled “Chunqiu Dashi Biao” (Tables for Major Events of the Spring and Autumn) listed an annotated 209 names of states mentioned in records of Chunqiu and Zuo Zhuan. Based on his work, some scholars further analyzed the existence of states by exhaustively checking references to other classical texts available and further terminating unjustified polities. Among those studies two works are reliable and thorough; Gu and Zhu (2001: 27–37) tabled 154 states, while Chen (1997) commentated the existence of 156 states during the eighth to fifth century bc. If we compare and cross-reference both works, excluding those referring to peripheral polities such as Rong in the north and Man in the south, we feel comfortable confirming at least 122 known vassal states as well as their ranks and clans of lords. 118

The age of territorial lords

Only a few dozen states have adequate information and can be cross-referenced in various texts. And of course even fewer were major players fighting to the end. The State of Yan ruled the northeastern territory and fought against various Rong 戎 or Di 狄 polities in the north.The State of Qin was given territories to the west of the Yellow River, which was once the heartland of Western Zhou but was lost to Xi Rong nomads. The lords of Qin constantly battled to gain control over their entitled lands. To the East the State of Qi was a major state on the Shandong peninsula, and the State of Wu as well as the State of Yue competed for lands in the Lower Reach of the Yangtze River. In the south the State of Chu grew around the lands near the Middle Yangtze River, fostering ambitious lords who were the first to challenge the authority of the Royal Zhou. Ideally, these major states were meant to stabilize and control the territories for the Zhou in central China (called the Middle Kingdom by Zhou people who once lived in the west) by improving the rights of minorities and expanding their regions in Zhou’s peripheral zones. In the central lands, many states were established surrounding Zhou political centers and royal residence. These lords were linked by royal lineages such as brothers, sons, uncles, and cousins. And further to the center of these states was the royal homeland, Wangcheng. Wangcheng, which was identified during 1950s archaeological surveys, had a layout of about 3.7 kilometers north-south and 2.9 kilometers east-west, in the heart of today’s Luoyang city. During the latter part of the 1920s, a few tombs were dug up by unauthorized local farmers in a village called Jincun, just to the east outside the Royal City. The exquisite artefacts, assumed to be from important noble tombs, were sold to private collectors and dealers as well as museums outside China (White 1934; Irwin and Shen 2016) (Figure 5.4). However, recent archaeological excavations and surveys of 397 Eastern Zhou tombs confirm the fact that the royal cemetery was located on the east side inside Wangcheng, where no other residential remains were discovered (Luoyang 2009). Most of the tombs are large in size, some descending as much as 7 meters beneath the surface. Two signs indicative of ultimate royal status were two bronze vessels (li container and ding-tripod) that are inscribed with “Made for the King” and a horse-and-chariot pit (ZK5) that revealed a six-horse-drawn chariot. According to the Zhou rites, only Zhou Kings had the exclusive privilege to use a six-horse-drawn chariot. Lords of lower ranks were allowed to use chariots pulled by four horses, or less. It is interesting to note, a nobleman of Dafu status in the State of Chu in South China, during the fourth century bc, according to the excavators of Jiuliandun tombs at Zaoyang of Hubei province, was buried with a six-horse-drawn chariot. The nobleman also owned thirty-two four-horse- and two-horse-drawn chariots (Hubei 2007).This illustrates vividly how the local lord from the south was being disrespectful to the royal house, abusing the rite that no one had previously challenged. Lord Zheng Wu Gong, holding the rank of count, ruled the State of Zheng in 770 bc and helped the Zhou King move his royal residence to Wangcheng.This move empowered the lords of Zheng, establishing them as the mightiest figures in the eighth century bc. In 1923, tombs of Zheng lords were excavated within the ruin of the ancient city of Xinzheng (National 2001). One hundred and two bronze vessels were recovered from this site and are now separately housed in five museums, including Taiwan’s National History Museum in Taipei. A bronze basin that belonged to one of the Zheng lords during the second half of the eighth century bc is identified in the Shanghai Museum. The state of Zheng, whose founding lord was the brother of the Zhou King in the later part of the ninth century bc, became one of the most powerful of the vassal states during the eighth to sixth centuries bc, but it was inevitably terminated in 375 bc by a rising power, the State of Han. As mentioned earlier, the lords of the Han State grew in power out of the weakening Jin State. The State of Jin was the strongest among the vassal states, enjoying its upward moving 119

Chen Shen

Figure 5.4  A. Dragon pendant. Jade (nephrite). Height: 3.3 cm; length: 5.5 cm; thickness: 0.6 cm. B. Dragon pendant. Jade (nephrite). Height: 3.5 cm; length: 9.2 cm. Both from Warring States period, 475–221 bc. Both jade pendants are exquisite artworks that belong to high aristocrats in the Eastern Zhou court, unearthed possibly from royal tombs at Jincun in the late 1920s, near Luoyang, then known as Wangcheng, the royal city of Eastern Zhou. The Bishop William C. White Collection: A: 931.13.17; B. 931.13.18

status, like that of Zheng State in the seventh century bc. However, Lord Jin Jing Gong moved to Xintian in 585 bc, when the state became weak. Jin used this location as a political center until 376 bc. The last Jin lord was forced to move out of this territory by the three newly knighted lords of Zhao,Wei, and Han. In the 1950s this ruin was identified and then extensively surveyed, where it continues to be an active archaeological site to this day. And to everyone’s surprise, unlike other Eastern Zhou capitals, the ruin at Xintian consisted of eleven independent walled sites instead of a single enclosure. The foundations of palatial structures or temple remains can be identified within those walled sites. The unique distributions of segments were clearly signs of a weakening regime, and the State of Jin had to rely on a few powerful Dafu ministers to run the palace city. To the southeast of the ruin, a special site yielded about 5000 covenant tablets from 326 pits. Texts of covenants written in red on about six hundred tablets of stone and jade can still be recognized, most of which are related to the alliances loyalty pledges, and curses to harm those breaking oaths. Further studies of covenant tablets suggest that conflicts of interests over resources and properties among the Dafu ministers, namely Zhao, Zhonghang, and Fan, were severe (Shanxi 1996a).The stone and jade tablets were buried with sacrificial livestock such as cattle, sheep, and horses, suggesting multiple events in which serious juration took place at Houma, Shanxi province, and during the occupation of Xintian. 120

The age of territorial lords

Figure 5.4  (Continued)

Scholars tend to believe the date for the covenant events to be around 495 bc, although another date, 424 bc, also has some merit (Zhao and Guo 2004: 141). During the early fifth century bc, one of the six Jin Qing ministers, Zhao Yang (ca. 540–475 bc), acted as the regent governing state affairs. Zhao Yang was believed to be the primary convenor of juration, and actively lead the Jin lord’s army to attack other vassal states. In 1988, a tomb labelled M251 was excavated in the suburban Taiyuan City of the Shanxi province. The owner of the tomb lay within a set of three coffins, one over the other. He was a male between the age of 60 and 70 years old. The owner was identified as Jin’s magnate Zhao Yang, according to the scale of the tomb as well as the inscriptions on several weapons.The intact tomb revealed thousands of burial goods, including exotic objects and materials such as jade ornaments, musical instruments, imported glass beads, and horse-drawn chariots created exclusively for the highest level noblemen. A special set of seven lie-ding bronze tripods, symbolizing the status of the vassal state lord were found in the tomb. This symbolic bronze tripod set was indicates that the regent minister, regarded himself as the actual head of state and who ruled the state before his great-grandson was given the title of Marquis of Zhao in 403 bc. In the surrounding area of tomb M251 were a number of smaller burial sites dated slightly later. It is evident that the site was a Zhao clan cemetery. Confirmation can now be made that the residential palace and administrative center of the Zhao clan were located to the north of the Jinyang walled-site ruin. Eventually, the State of Zhao would move its own capital to Handan in the fourth century bc (Shanxi 1996b). Some states were established through blood-lines of the Royal Zhou clan, namely Ji, and other states were established as an honour given to other clans. For example, the State of Chen 121

Chen Shen

陳, which was located in the Henan province east of the State of Jin, was enfeoffed to the gui clan in 1045 bc. In both the Shanghai Museum and the Royal Ontario Museum, there are matching pairs (one at each museum) of bronze food containers – fu (Figure 5.5). The bronzes both have 25-character inscriptions inside the body and lid, respectively. The text reveals a set of ritual vessels given as wedding gifts to the second daughter by the lord of Marquis Chen in the seventh century bc. Bronze vessels with same inscriptions were archaeologically recovered in 1963 within the territory of the State of Lu (Gao and Shao 2005: 532). Such a wedding gift might have suggested that the marriage was likely part of an inter-state alliance that was well documented.The State of Chen was demolished by the State of Chu in 478 bc after Chen lords had ruled the state through twenty-five generations. The formation of vassal states under Zhou’s fengjian system (see earlier) allowed territorial lords to retain their independence while governing their lands and taxing their people for the Zhou kings.This was not just for political reasons or strong social identities but to maintain their indigenous cultures and customs. Those states tried to be independent from Zhou rites, while the Royal Zhou expected local governments to mainstream Zhou beliefs by moving away from “barbarian” traditions. Such a “pull-and-push” enviroment can be clearly demonstrated among vassal states in the Shandong peninsula. The region of East China was primarily occupied by “Yi” 夷 groups, who developed their own indigenous cultures during the Erlitou and Shang period (twentieth–eleventh century bc). In the Eastern Zhou, there were more than thirty vassal states on the peninsula. They represented three types of territorial lords: Zhou traditions, Yi traditions, and those in between. The founding lord of the State of Lu, Zhou Gong 周公, was the brother of the founding Zhou King Wu Wang (? – 1043 bc). Lord Zhou Gong was sent by the King to pacify the rebellions of Shang supporters in the peninsular region during the second half of the eleventh century bc. Yi people in the State of Lu retained their local traditions during the Western Zhou, evidence of which can be found through burial traditions from archaeological sites. However, the lords of Lu advocated for the rites of Zhou as governing principals to pressure Yi into Zhou traditions. During the Eastern Zhou, the State of Lu had tried every means possible to retain and to revive the declining rites of Zhou, which had fallen into out of favour. History regarded Lu as the “State of Rite”, where Confucius and his students were a primary force keeping societal principals under the guidelines of Zhou rites. Lords of Lu had become influential figures during the eighth–seventh century bc but were struggling to survive during the Warring States period. The State of Lu was eliminated by the State of Chu from the south in 256 bc, the same year the Royal Zhou was terminated by the State of Qin. In contrast, the State of Ju 莒, which was terrorizing most of the southeast part of the Shandong peninsula along the Si River, had already been a long-standing self-governing polity of the Yi people since the Shang period (ca. 1600–1045). Ju became Zhou’s vassal state at the rank of viscount in the eleventh century bc. It was a secondary state, but it represented one of those peripheral governments under the rulership of the Zhou. Like the State of Lu, it was eliminated by the State of Chu in 431 bc, when Chu expanded its territory from the middle valley of the Yangtze River to East China. But according to classic documents such as Chunqiu, the state played an important role in balancing the conflicts between the superpower states. This was done by creating alliances, resourcing, and pacifying, for the benefit of the indigenous Yi people. Walking a fine line between the superpowers, lords of Ju would have to maintain their own identity and traditions while accepting the rites and rulings of Zhou. Archaeological remains uncovered from tombs of Ju lords reveal a strong local material culture (Li 1985: 149–151; Shao 2013: 193). A tomb excavated at Huayuancun in Junan County suggested four points of material culture which departed from the main rites of the Zhou. First, a ramp leading up to the 122

Figure 5.5 Fu food-container (A), detail of the lid with inscription (B), and rubbings of inscriptions (C). Bronze, eighth–seventh century bc. L: 29.5 cm, Eastern Zhou dynasty. The inscription on the bottom of the container, as well as on the lid, indicate that one of the Chen State lords during the eighth to seventh century bc commissioned sets of ritual vessels as a wedding gift to his daughter, who possibly married a prince of the State of Lu. The Bishop C. White Collection, 932.16.82

Chen Shen

Figure 5.5  (Continued)

earthen-pit chamber is located at the south part of the east wall. Second, there is a dog-sacrifice pit (yaokeng) under the center of the main chamber, a tradition of the Shang period, but rarely seen in Zhou lord tombs. Thirdly, ten sacrificed human bodies were buried within a coffin that surrounded a main coffin. Fourthly, the main chamber was divided equally into two parts: the north part is for the owner’s resting place, while the south is for an enormous amount of exotic funeral goods. Of course, besides Yi in East China, there were peripheral states or polities that retained similar relationships with Zhou. Eastern Zhou documents refer to some political strongholds surrounding the central lands of Eastern Zhou, but recent archaeological discoveries have illustrated their presence and ways of life. A cemetery of 595 Eastern Zhou burials, which recovered 60,000 artefacts at Jundushan near present-day Beijing, revealed undoubtedly a northern tradition of Shan Rong state polity. Shan Rong was one of many Rong ethnic groups living in the north just outside of Zhou’s territory. The other one, Xi Rong, was responsible for invading the Zhou homeland at Zhouyuan from the west, capturing and killing King You Wang in 770 bc, ending the era of Western Zhou. Further in the west, excavations of the Majiayuan cemetery in Gansu during 2006–2011 also suggested a unique western ethnic polity of Rong people, who, it is believed by many scholars, were ruled by the State of Qin. In 2015, a report from an important archaeological discovery in Luoyang revealed the existence of the legendary Luhun 陸渾clan of Rong (Guojia 2016: 86–91). As noted in Zuo Zhuan, the Luhun Rong people were forced by the states of Qin and Jin to relocate their territory to the suburban area of today’s Luoyang in 638 bc and established a small State of Luhun with the entitlement of viscount by the Zhou Court. Falling into conflict between such superpowers like Jin and Chu, the State of Luhun was eliminated by Jin in 525 bc. The existence of Luhun State, with its 113-year history, is now unveiled by archaeological evidence uncovered in the Xuyang village. So far three years of excavations 2013–2015 have uncovered a city ruin (Nanliu site), two hundred plus burials, including two to three large scaled tombs likely belonging to the lords, eight pits of horses and chariots, and ten kilns within a territory of 20,000 square meters. This ongoing archaeological investigation is significant in its validation of this segment of Eastern Zhou’s history. 124

The age of territorial lords

In the south, the people of Eastern Zhou were referred to in Chinese history as Man 蠻. The Man people, just like Yi and Rong, were enormous clans who resided in fertile lands in the valleys of the Yangtze River and its tributaries (e.g., the Nanyang Basin) and the valley of the Huai River. During Zhou rule, lands south of the Yangtze River were controlled by the Royal Zhou through the Fengjian system (see earlier), establishing vassal states by enforcing jurisdiction of Zhou’s Ji and Jiang clan brothers and princes (such as the states of Chu,Ying, Cai, Zeng, and Tang, etc.) (namely Sui), as well as leading clans of local populations (such as the states of Jiang, Xi, Shen, Dao, Lü, Xu, etc.). There were also a number of states that were relocated from elsewhere, just like the aforementioned Luhun Rong. States of Hu and Xuan were Rong clans from the North, whereas states of Huang, Fang, and Yang were Yi clans from the East. Xu Shaohua, a professor from Wuhan University, focused his study on states of Zhou’s southern lands.Taking a historic-geographic perspective, Xu’s works (1994) identified the locations of at least sixty states, small and large, that were more or less known in the historic texts. He further analyzed in some details twenty-nine states (including Chu) for their origins, development, locations and relocations, environments, and their relationships with neighbouring states. In particular, Xu’s study was supported by recovered inscribed bronzes and surviving bamboo strips along with field archaeological investigations. One small state, namely the State of Zhongli, was not included in his twenty-nine-state case study due to lack of archaeological evidence, but this situation changed in 2006, twelve years after his dissertation. In 2006–2008, a large-scale tomb was excavated in several sessions at Shuangdun village in the valley of the Huai River in Anhui province. Its unique tomb mound and funeral structures stand outside Chu-style tradition in the south (Anhui 2013). Based on inscriptions from a series of bronzes, including nine bells, two fu container vessels, and one ji dagger, the owner of the tomb can be undoubtedly identified as Lord Bai of the State of Zhongli. The dental analysis of his remains suggests Lord Bai died at the age of 40. By comparison of materials from this tomb, scholars are now able to identify the owners of the other two tombs excavated earlier; they are Bai’s son Lord Kang, buried at Bianzhuang of Fengyang and Bai’s descendant, Lord Yu, buried at Jiulidun of Xucheng, respectively. The samples of C14 dating were taken from Bai’s tomb, which has resulted in the dates ranging from between 2600–2650 years ago (roughly 650–600 bc). Not far from the three burials of Zhongli lords was a small walled city ruin 360 meters by 380 meters in area known as the Zhongli ruin. Allegedly, this was the residence and administrative center of Zhongli lords. According to Zuo Zhuan, in 576 bc ministers and/or representatives from the states of Jin, Qi, Song, Wei, Zheng, and Zhu conveyed a meeting at the State of Zhongli 鐘離 to negotiate with representatives of the State of Wu. The date of this meeting was not far from the rule of Lord Bai and his son Lord Kang.These kinds of inter-state meetings frequently occurred during the Spring and Autumn period when northern superpowers tried to strategically extend their territories into the south using military force. Two bronze weapons, ji-ge and ge daggers, were buried in the tomb with Lord Bai and were identified as property of the State of Xu 許.The appearance of two inscribed weapons indicated that Lord Bai led a battle against the army of Xu and captured a large property, including these two weapons.

Urbanization and population in cities The Zhongli city ruin revealed important evidence for verifying the location and territory of the State of Zhongli. It was also one of the enormous cities built in Eastern Zhou. During the competitive evolution of vassal states at the time when they were expanding their territories, building new cities and conquering the cities of their enemies were the most effective ways of 125

Chen Shen

securing lands and claiming victories. Building large cities created a wave of immigration. With increased populations they adopted new ways of living in these cities. Urbanization took place in an unprecedented rate. leading to dramatic changes in Eastern Zhou societies (Shen 1994, 2004). Zuo Zhuan recorded sixty-eight events of city-building, including five events of the same cities being rebuilt. Zhushu Jinian (Annual of Bamboo) mentioned an additional twenty-eight events of city-building that had not been previously mentioned. The recorded names of these cities were dated only to the Spring and Autumn period. Combining the text records of Zhushu Jinian and Shiji (the Record of the Grand Historian), there were forty-six events of city building during the Warring States period between 469 bc and 391 bc.The latter numbers are underrepresented for the period because it referred to those in only nine states. Archaeologically, decades of field investigations have gradually verified some of these records and identified new locations of cities. By 2000, 448 urban sites of the Eastern Zhou have been published in survey reports, many of which are not yet been systematically excavated (Xu 2000).Within the territory of Chu State alone, archaeological surveys to date have identified more than fifty city ruins (Zhongguo 2004: 227). While there are just a dozen or so large urban sites identified as known state capitals, the number of county cities and military fortresses found has dramatically increased. This corresponded to the rapid changes in political situations during the fifth–fourth century bc. By analyzing the archaeological data of six capital city sites (Qufu 曲阜 of Lu state, Linzi 臨淄of Qi state, Xinzheng 新鄭 of Zheng and Han states, Handan 邯鄲 of Zhao state,Yanxiadu 燕下都 of Yan state, and Jinancheng 紀南城 of Chu state), Shen (1994) has offered several observations in reference to the urbanization process. First, the sizes of the state capitals increased rapidly over time. Among the six, Qufu was the earliest to appear, in middle Western Zhou (ca. ninth century bc), but in size it was the smallest (8.75 square kilometers).Yanxiadu, founded in the middle Warring States period (between fifth– fourth century bc), became the largest in size (22.68 square kilometres). The average size of the six cities under examination was 16 square kilometers, which was quite sizeable to accommodate a residence population of hundreds of thousands. Of course, increased urban populations were probably the driving force for expanding cities. Zhanguo Ce (Stratagems of the Warring States), a collection of essays originally written in the Warring States period and later edited in the Western Han period (ca. first century bc), revealed a story in which a famous Zhao State military general, Zhao She, who possibly lived between 324–245 bc, was questioned by his primary minister,Tian Shan, about the amount of troops he sent into a battle. Zhao She responded that “in the old days, we attacked a city [in square layout] about three-hundred zhang 丈 [on one side] with about three thousand household people; now, our troops are marching to a city with a thousand zhang and ten thousand household people” (zhang is ancient length unit). Second, cities that were built during this time were no longer following the principals of city planning in Zhou rites; instead cities were built for strategic purposes as well as for domestic living convenience. City walls conformed to topographic-hydrographic and demographic features, witnessed in the city ruins of Jinancheng, Linzi, and Xinzheng, among others. In Qufu, the road leading to the palace zone was significantly wider than the others, at least 15 meters in width rather than the average 10 meters. However, this regulation did not appear in Linzi of Qi State, where commercial streets were more magnificent than roads built in the palatial city. It clearly suggests that royal rites once again had been challenged by growing commerce, trade, and migration in the new style of urban life. Third, as for the internal layout of the city, the palatial city (with an inner wall) or a palatial zone (without an inner wall) within a city became more and more irregular and revealed a sense of informal construction through time. A cluster of royal palaces or temples were no longer 126

The age of territorial lords

positioned in the center of the city, with few exceptions. For example, Qufu, as the capital of Lu State, had its palatial zone in the center of the city. It had a royal gateway leading to the central south city-gate, a model that was set forth by the construction principle of Zhou rites. But other capital cities, especially ones built during the later period, had their palatial cities/ zones set aside in a separate enclosure, as found in Linzi and Handan, or separated by natural or man-made canals, such as in Yanxiadu. It is believed that during the seventh century bc,Yanxiadu in Yi county of Hebei province was formally established as a lower capital of the State of Yan. It was the largest state in northern China used to defend against the northern nomadic groups Di and Rong for the Royal Zhou. The site was enclosed by two rivers, one to the north and the other to the south, and moats were dug to protect the east and west sides (Figure 5.6). The west enclosure of the city was built during later expansions, which was a direct result of population growth, but it was never fully occupied by residences. Nor did it actively participate in commercial trade before the collapse of the state. The royal clusters, with ruins of large-scale

Figure 5.6 City layout of the Yanxiadu ruin, a lower capital of the State of Yan, now in the present Yixian county of Hubei province.The city was established in the seventh century bc when the center of the State of Yan was moved southward. The layout illustrates the locations and foundations of royal structures, workshop sites, occupational sites, and cemeteries.

127

Chen Shen

foundations, were located in the north portion of the east enclosure separated by a man-made canal at the south side and a dividing wall in the north. Outside of the royal court the area was surrounded by workshops and resident compounds. In the case of Linzi city, the capital State of Qi, the palace city was constructed later in the fifth century bc (the beginning of the Warring States period). The original large city was built and occupied by a variety of residents and production sites beginning in the ninth century bc. A recent report on the city ruin by field investigations revealed archaeological evidence which suggests that at least nine stages of constructions and reconstruction of the city (including expansion) over the Eastern Zhou period (Shandong 2013). The new survey report also indicated that original palaces and temples dated to the Spring and Autumn period were set in the center of the original Western Zhou-dated city before expansion. Adding a separate palatial city is regarded as a significant change and a direct result of urbanization. Fourth, another archaeological implication of urbanization at this time was expressed by changes in the sizes and locations of cemeteries that were associated with urban sites. Archaeological evidences found in six capital cities examined by Shen (1994) revealed that the majority of cemeteries recovered within city ruins are dated to the Spring and Autumn period, whereas the tombs of the Warring States period were found mostly outside the cities. It is reasonable to believe that this consistent change across large cities was probably a direct result of the pressures of increased populations on urban land uses. In the case of Yanxiadu city, during the Warring States period, only twenty-three tombs were recovered within the confines of the city site, compared to a total of 480 tombs that were clustered within an area of 550 meters by 300 meters southeast outside the city. Currently, studies on the Eastern Zhou demographic data have been merely speculative through the analyses of texts and literature. A frequent quotation used for interpreting population at Linzi came from Su Qin (ca. ? – 284 bc), a strategist of the Qi state, when he met with Lord Qi Xuan Wang (ca. 350–301 bc). Su Qin mentioned to the lord that there were about 70,000 households in the city of Linzi. The lord could easily recruit up to 210,000 soldiers, assuming that each household had at least three males residing in it. Some scholars have gone to some extent to verify the accuracy and reliability of this quotation (Han 1996). If we consider adding one female to each household at least, there would be 280,000 people in the city of Linzi during the second half of the fourth century bc. Archaeological surveys at Linzi suggest the total inhabited area to be about 3,000,000 square meters, thus giving a density of 11 square meters per person. Or according to Jiang (2002), there were 290 square meters per household at Linzi, as the total area of city was about 20 square kilometres. It is apparent that the dramatic population increase paralleled urbanization, as is being demonstrated by archaeological data in respect to the sizes of residential /inhabitanted areas within the city ruins at Qufu (Shen 1994). Archaeological data also revealed how Qufu's residential areas increased from 545,000 square meters in the Western Zhou, through 1,087,000 square meters in the Spring and Autumn period, to finally 1,386,000 square meters in the Warring States period. It is not surprising that areas identified as workshop sites or factory ruins at Qufu also consistently increased over time: 147,000 square meters in the Western Zhou, 239,000 square meters in the Spring and Autumn period, and 356,000 square meters in the Warring States period (Shen 1994).

Commercialization and production One of the key indicators of urbanization is the increase in land use for workshops or factories within the city ruins under archaeological investigations. This resulted in a social class or 128

The age of territorial lords

classes who were free of farming. New social groups emerged such as workers, tradesmen, and merchants. However, their social status, which was under the influence of Confucianism, was inferior to scholars, farmers, and even artisans.Yet such social constructions of the Eastern Zhou cities had stimulated mature economic prosperity, provided that the lords of state and their nobles were ready to accept their presence and actions in their city life (Shen 2004). A unique characteristic of the Eastern Zhou cities, which was distinctly different from previous political centers during the twentieth–tenth century bc, was the emergence of markets and private production workshops. The Kao Gong Ji, a chapter of Zhouli (Rites of Zhou), a rite system manuscript of Zhou supposedly re-edited in the Western Han dynasty (206 bc–ad 9), suggests a principle city plan: The artificers demarcated the [Royal Zhou] capital as a square with sides of nine li, each side having three gateways. Within the capital there were nine meridional and nine latitudinal avenues, each of the former being nine chariot-tracks in width. The ancestor-worship temples were on the left side of the city, while the god-worship temples were on the right. The administrative centre was in the front and the markets were in the back. This clearly documents “markets”, which had already played an important part in the city plan. In fact, the Chinese phrase Chengshi for “city” is consisted with Cheng as a walled enclosure of living and working and Shi literally as market. Such a concept was naturally derived from the Eastern Zhan urbanization process (Shen 2004: 291–292). The earliest dated “marketplace” site so far was discovered in the ruin of Yongcheng, a capital city of the State of Qin during the eight to fifth century bc. The site was located at the northeastern corner of the 3.4 by 3.1 kilometer city enclosure, meaning at the “back” of the city, whereas royal palaces and temples were facing south. This marketplace site was also enclosed by a walled enclosure running 180 meters east-west and 160 meters north-south. A gate structure was identified at the middle of each side, making the market a place that could be forcefully managed. Paved walkways with large bricks formed streets in a grid system. This regularity allowed the shops and vendors to be located accordingly. Based on artefact assemblages including coins, roof tiles, and pottery wares that were impressed with marks, archaeologists positively identified that this was truly an excellent example of city-managed urban markets identical to what was mentioned in the surviving ancient texts. In 1975, a cluster of bamboo strip books was recovered from a Western Han tomb at Yinqueshan, Shandong province (Shangdong 1975). One of the reconstructed books revealed an incomplete section entitled Shi Fa (Regulation of the Marketplace), written in the Warring States period. The surviving text describes that the city regularized the size and layout of the marketplace, placement of vendors and shops, the organization of the market administration, and market operation. It also listed strict state bye-laws to control market officers’ possible misconduct. In Xunzi, a collection of philosophical writings by a third-century bc philosopher and teacher, Xun Kuang, the chapter Regulation of the Marketplace categorized the duties of the market officer to include maintenance and cleaning, traffic control, security patrol, and price control. However, when marketplaces were large and fully developed in the later period, market officers had additional duties, notably vendor management, merchandise inspection, dispute settlement, and loan arrangement. In addition, one of the major tasks of market officers in the Warring States period was to collect sales taxes, property taxes, and import taxes. Yang (1998: 129) noted in his analysis that these commercial taxes became a very important part of state revenues in the Eastern Zhou period. A commercial city, namely Dingtao in the State of Song in present Henan province, was famously known for its vigorous and vibrant market activities. 129

Chen Shen

Some powerful neighbouring states attempted to conquer the State of Song in order just to collect the tax revenues from Dingtao (Bai 1994). There were designated names for each marketplace. According to various classic texts, we know there was Yang Shi marketplace in the State of Zheng, Zhou Shi and Xu Shi in the State of Lu, and Er Shi in the State of Chu, etc. Ten tombs in the State of Qin were excavated in the southern suburban area of today’s Xian city, near the ancient city of Xianyang, the Qin capital. Twenty pottery pieces were discovered with twenty-six marks; these pots were either stamped or engraved when the earthenware was half dried. The majority of marks are consistent and are composed of two characters, either Du Shi or Du Ting. Here Ting is similar to Shi, referring to a marketplace. Compelling archaeological evidence suggests the existence of a marketplace being managed in the city of Xianyang at one time during the fourth–third century bc, which was not mentioned anywhere in surviving texts (Xian 2006). Some scholars also believe that pottery wares with Shi or Ting marks must have been products from state-run factories (Yuan 1980). The development of marketplaces in the Eastern Zhou cities was closely related to the rapid growth of private production. Craft production factories were frequently identified in archaeological ruins of cities dating before the Eastern Zhou period, but products were mainly bronze, ceramic, bone, and ivory, items made exclusively for the lavish lifestyles of the royal family and upper-class members. Because of that, very few products were exchanged in the market. During the Eastern Zhou period, there were four types of productive modes in cities.The first type was household craft workshops. These evolved directly from farmer families with particular skills who migrated into the cities for permanent settlement and relied on exchanging goods. Their operations were rather small but self-sufficient, and their products might supply large operations or be sold in the markets. The second type was specialized craft production for goods made of ceramic, leather, metal, and wood. The craftsmen were referred to in the record as Bai Gong (skilful workers), indicating an unprecedented class of non-farming residents. Their products were mainly sold in markets. The third type was private cooperative production of merchandise on a large scale. These manufacturing establishments were mainly iron foundries and salt production operations, which required systemic production flows. Both kinds of production had previously been managed by states, but later some entrepreneurs were given permits of operation under certain (i.e. tax) conditions (Yang 1998: 108–109).The last type of production in the city was state production, managed by state representatives. The products, including weapons, ritual bronzes, and coins, were mainly for state or royal court use, not for sale. Most state foundries were administered by a three-level hierarchy: master craftsmen, who were responsible for casting; supervisors, who were in charge of the daily operation of the foundries; and inspectors, or lord designates, who inspected and accepted products. The name and title of officers were normally cast on the products, which have been identified from archaeological findings. Detailed analyses of archaeological materials from the workshop sites within the ruin of Yanxiadu city illustrate vividly the different production modes governed under the State of Yan. Based on the recovery of different types of raw materials and by-products, the excavators identified eleven workshops, including one iron foundry (G5 in Figure 5.6), one bronze-casting foundry (W21), four weapons manufacturing factories (LJ13; W23, W18, and LJ10), one mint (G4), one coin-mould workshop (LJ30), two pottery workshops (LJ11 and LJ29), and one bone workshop (W22) (Hebei 1996: 85–434). All of these workshop sites, except one, were distributed along river canals located in the central to northern parts of the city, suggesting a good use of water resources. Sites LJ13, W2I, W18, W23, and G4 may be interpreted as state-operated foundries (Shen 2004: 299–301).The interpretation of state-run production is not merely based on exclusive and exquisite products such as bronze weapons, bronze ritual vessels, and coins from these sites but also in accordance with scale and the structure of the sites themselves. At site W23, hundreds 130

The age of territorial lords

of weapons of a single type were found, a ge halberd, a long-shafted dagger; these were collected from an area measuring 50 square meters, suggesting mass production. Evidence of state involvement in manufacturing is confirmed by inscriptions on the weapons recovered. All ge halberds from the W23 sites had the lords’ names cast on them. Further archaeological evidence of state-controlled production comes from official seals recovered from these sites, which give the names and ranks of officers in charge of the production. These sites are also large in comparison to other identified workshops, and some of the more elaborate foundations could be the remains of administrative buildings associated with the foundry sites. At the LJ13 site, there are large numbers of decorated semicircular eaves-tiles with styles similar to those found in the royal complex (Figure 5.7).

Figure 5.7 Ridge tile (A) and detail (B). Earthenware, Warring States period; length 90 cm, height 20 cm, and width 30 cm. A large roof ridge tile, probably used on a grand official palace at Yanxiadu in the state of Yan, terminated in 222 bc by the State of Qin. Similar tiles have been discovered from sites within the First Emperor’s tomb complex. The George Crofts Collection, 922.20.630

131

Chen Shen

According to historic records, private producers were usually engaged in the general manufacture of daily utensils and farming tools; accordingly, archaeologists would expect these types of workshop sites to be limited in scale and the artefacts recovered to lack elaborate materials. Shen (2004: 301–303) further identified four workshop sites, W22, LJ10, LJ11, and LJ30, to be locales of private production in the State of Yan capital. Artefacts from these workshops were not made of elaborate materials, and the sites were relatively small. Artefact assemblages were dominated by pottery vessels and iron handicraft or agricultural tools. Site LJ10 may have been a household-level workshop specializing in stone working in the early phase of its use, but then it may have developed into a large-scale multifunctional workshop in the later period. Most of the artefacts from the early Warring States period are limited to pottery utensils and stone objects. A cluster of nearly 300 stone objects were unearthed from a single pit, H729, in 0.30 meters deep. All of the stone objects are ornamental in design, and more importantly standardized in production. These open-worked phoenix-design stone pendants, along with manufactured by-products that were surface finds at the site, are identical to those excavated from one of the lower-class tombs outside the city,YNXM2. This suggests that such objects must have been made for market exchanges and the grave owner must have obtained these items from markets to bring into his or her afterlife. Pottery vessels constituted the main component of artefact assemblages throughout the three phases at the LJI0 workshop site at Yanxiadu. Because no kilns were positively identified at the site, it is hard to assess whether the ceramic standardization was a result of on-site production. Rather, these vessels may have been personal items used by workshop craftsmen. In this case, it is suggested that these ceramic wares were merchandise from the market, indicated by impressed marks on the vessels. A majority of these vessels featured characters or symbols indicating some properties of pottery manufacturing. In general, there are three kinds of marks. The first one is a line of three characters indicating a potter’s name. The second kind of marks shows a single character, probably standing for the name of a “workshop” or “shop.” The third kind of marks consists of symbols of value, such as “three”, “five”, “seven”, “ten”, “twenty”, and so forth. The standardization of these goods is also evident from the location of the impressed marks on the vessel. Almost all of the jars have marks on the shoulder, while all of the basins have marks on the rim. Another type of ceramic ware, the dou stemmed cup, displays impressed marks on the stem. All of these pottery vessels bearing marks appeared after the beginning of the Warring States period, which clearly suggests the timing of the development of market exchanges.

Life in peace and war Numerous texts, poems, and literature written in the period or immediately after the era have illustrated many aspects of Eastern Zhou social life. The focus of this chapter has been on changes in the societal and political practices in the Eastern Zhou, and how these changes affected people’s lives. Thus, a full summary of the Eastern Zhou life is beyond the scope of this paper. Readers are welcome to explore other aspects of social-cultural life in the Eastern Zhou in the other chapters of this book. However, the interpretations of a bronze vessel housed at the Royal Ontario Museum (ROM) will be presented here in order to illustrate the fact that people who were fond of music and dance had to simultaneously live with war and violence. Life in Eastern Zhou was not only vibrant and sophisticated but equally romantic and realistic. The bronze vessel, called a hu in Chinese, is a type of ritual wine container.The ROM’s vessel was purchased by the museum from a Belgian private collector in 1992 through the generous

132

The age of territorial lords

bequest fund of Dr. Herman Herzog Levy (Figure  5.8A). The hu vessel has a round body of 22 cm in maximum diameter and 36.6 cm in height and is missing its original lid. The unique feature of this decorated vessel is its copper-inlaid engraved design nearly covering its entire body. The design is composed of more than 100 people grouped in six themed scenes that are evenly distributed in three horizontal registers. Each theme is repeated once in its respective register so as to make the front and back of the vessel nearly identical in pictorial presentations. The six themed motifs depict the following scenes: an archery contest and mulberry picking on the upper register; a music picnic and bird hunting in the middle register; and battles on the water and attacking a city in the lower register (Figure 5.9). If we use this pictorial scenery as illustration of Easter Zhou daily live, we need to first understand how the date of this vessel can be established. In addition to the fact that from a stylistic point of view the artwork gives away the fact that it is an Eastern Zhou bronze product, only four other pictorial hu vessels resembling the ROM’s were archaeologically uncovered and dated to the Warring States period: one from tomb M10 at Baihuatan in Sichuan province was unearthed with forty-seven other bronzes, while two others were recovered from the Gaowangsi tomb in Shaanxi province (Figure 5.8B).The fourth one was discovered in tomb M2 at Dazhang of Xiangfeng county in Shanxi province and was dated to between 430 and 420 bc (Fang and Shen 1999: 69–70). Therefore, we can comfortably place the ROM’s hu vessels in the range of the Warring States period. Archery (she 射) was one of the Six Arts in Eastern Zhou education, and archery training and contests must have frequently taken place as community activities. There were at least four categories of archery contests at that time: da she, bing she, yan she, and xiang she, known from surviving texts.The image of archery on the vessel in Figure 5.8A (shown in detail in Figure 5.10A) depicts a two floored structure with a roof; one person has shot an arrow, which has hit the target, and the person behind him is ready to aim. Five people are lined up waiting for their turn to shoot. It is interesting to note that all seven people are wearing similar clothes, probably uniforms, but their hairstyles differ, with some archers wearing hats. Alongside the archery scene is a depiction of mulberry picking (Figure 5.10B). A group of ten people, five males and five females, show signs of joy and pleasant encounters on and under two mulberry trees. The scene has been interpreted in many ways, the common one suggesting the action of mulberry picking for domestic silkworms. However, some scholars believe that the group of five couples are on an outing in the forest suggesting private dating and romantic moments (Liu 1990). The central body of the vessel illustrates two mass gatherings, one for camping with music and dance and the other for bird hunting in the field. Archery contests that were similar to the one discussed earlier were also a part of the camping scene, and music presentations and took place under a canopy (Figure 5.10C and 5.10D). A band consisting of ten people, with two people each playing suspended bronze bells zhong, overhanging stone chimes qing, reed-pipe sheng, vertical bamboo flute xiao, and drum percussion, while another six are illustrated dancing outside. The dancers are wearing short swords. There are four people outside the canopy who are either getting a drink or focusing on cooking. On the other side of the camping festival is a bird hunting scene (Figure 5.10E). Bird hunting, called yi she 弋射 was the most popular sport during the Eastern Zhou. Zou Zhuan (Zou’s Annotations to Chunqiu) tells a story that the State of Cao lord, Yang, was so fond of yi she that he improperly promoted a civilian, Gongsun Jiang, to an important ministry position, just because Gongsun was the best yi she archer in the state. Targets for bird hunting were primarily white goose and widgeon (a freshwater duck). In Eastern Zhou society, it was customary to present a goose as a betrothal gift, so hunting-goose using yi she technique was in much demand. The yi

133

Figure 5.8 A. Pictorial Hu wine vessel. Bronze inlaid with copper. H. 36.6 cm. The Eastern Zhou. B. Pictorial Hu wine vessel. Bronze inlaid with copper. H. 40 cm, diameter 12.3 cm. The Eastern Zhou. Numerous activities are represented in the engraved images on pictorial hu wine vessels, ranging from spectacular displays of full-scale battles on land and water to celebrating events and gatherings with dancing and music. Revealed in (A) are more than 100 human figures within three horizontal registers. A. Dr. Herman Herzog Levy Bequest Fund, 992.169.1 Royal Ontario Museum B. Fengxiang County Museum, China

The age of territorial lords

Figure 5.9 A full-scale drawing illustrating descriptive scenes from the pictorial bronze Hu wine vessel in Figure 5.8A

she technique is one where an arrow is attached to an extendable long string to shoot targets, and this allows archers to retrieve their prey by pulling back the string. In the illustration, we see four people in action; each wear different costumes from those on the camping ground. Birds, or geese, are shown in three poses: flying away, falling down, or being shot with a stringed arrow. Hunters are shown shooting arrows and pulling strings. However, these idyllic and picturesque times would come to an abrupt end when war broke out. Images on the lower register of the vessel vividly reconstruct two battle scenes that commonly occurred at this time. One image shows a battle on the water taking place on a boat (Figure 5.10F), and the other one displays soldiers attacking and defending a city. Two boats are shown to be sailing head-on suggesting that the navies from the two states were against each other; this is clearly indicated by the two distinct ensigns. The ships have two decks: four to six oars in the lower deck, while six soldiers on the upper deck face each other. The front ones use short swords; one already has fallen into the water, and the other is falling down. Soldiers in the back are using long-shafted weapons called a ji halberd and display skilled battle techniques. On the other side, the battle over the city wall reveals a bloody scene. The artist used a single 135

Figure 5.10  Detailed imagery from the pictorial bronze Hu wine vessel in Figure 5.8A

Figure 5.10  (Continued)

Figure 5.10  (Continued)

The age of territorial lords

horizontal line to symbolize the city wall, while two oblique lines ending on the city wall suggest battle ladders, where soldiers are portrayed climbing with weapons. Within such limited space, the artist depicted eleven soldiers on the city wall (above the horizontal line) fighting in groups, with a single soldier focusing on defending enemies from the bottom of the city. The bloody scene displays decapitated soldiers falling off the wall, as well as graphic images of stabbed and wounded fighters. The weapons illustrated on the images include short daggers, swords, long-shafted ji halberds, bows and arrows, shields, and crossbows. Fang and Shen (1999) carefully analyzed clothing and uniforms as well as the detailed hairstyles of the soldiers that appear on four similar pictorial vessels. They were able to categorize these soldiers, both on land and on the water, and divide them into four groups: (a) short hair, (b) high hair, (c) open hairdo, and (d) wearing caps or helmets. In the scene on the water, b-soldiers are depicted winning the battle over a-army. On land and on the wall, b-soldiers had lost the fight to d-soldiers, but won over c-soldiers. It may be speculated that (a) and (b) represent armies from the States of Wu and Chu, respectively, as both were in the south and commanded strong naval forces. Since both the states of Wu and Chu had penetrated their forces into central China, it is possible that both (c) and (d) represented states like Zheng and Jin. Such an assumption is based on the frequent battles recorded among those states. According to Shiji (Records of a Grand Historian), between 604 and 547 bc, the State of Chu invaded the Zheng State nine times, and the State of Jin attacked Zheng twelve times.

Toward unification The documentation of many wars and battles are recorded in the classic texts. Visual depictions of warfare are found on contemporary objects. As well oral accounts of great conflicts were passed down through the generations. Over the years many historians have made considerable efforts to list all known battles. Gu and Zhu (2001: 529–564) list major events of the Spring and Autumn period including military actions between 770 bc and 453 bc. Yang (1998: 696–722) made a similar table of events of the Warring States period between 481 bc and 221 bc. Counting their lists, both coincidentally came to similar numbers: about 300 military engagements among vassal states have been extracted from historic texts. However, the scale of military conflicts and efforts in participation in these wars dramatically increased over time. The aforementioned conversations regarding the city sizes and population (p. 133) between General Zhao She and Primary Minister Tian Shan of the State of Zhao took place in 266 bc. The primary minister disagreed with his general, who wanted to dispatch 100,000 to 200,000 military personnel to a battle, as Tian was worried about his people being taken away from farming. Tian further stressed to General Zhao She why 30,000 troops would no longer be sufficient to win a battle as it did in the past. In refutation, the general simply pointed out that times had changed, and that the Zhao state was now one of the Seven States that survived out of the hundreds of vassal states that came before. Each of the Seven States could easily control an army of more than a few hundred thousand soldiers, making the battle last as long as it needed to. The general further demonstrated to the primary minister the fact that the State of Qi deployed an army of 200,000 soldiers to attack the State of Chu, but the battle lasted five years, and that when the State of Zhao sent 200,000 soldiers to attack the State of Zhongshan, the soldiers did not return after five years. With this bold declaration, Primary Minister Tian had a very high opinion of his generals after all. This historical account clearly reflects the changes of intellectual opinion toward the political and military situations at the dawn of unification. After 259 bc, when the last King of Zhou 139

Chen Shen

was dead, the lands of China encompassing the valleys of the Yellow River and the Yangtze River were now controlled separately by the Seven States and their small allied states. Changes brought forth during the last 500 years had come to the point of no return – each of these superpowers could no longer sustain co-existence with the others: they needed to move unilaterally towards unification. Lords of the Seven States, now the kings, felt that the only way of survival was to conquer neighbouring territories and utilize key resources from their rivals. This allowed them to satisfy the rapidly growing needs of the populations and logistical demands required for expanding military forces. Mencius, a great thinker and philosopher at the time, actively condemned the practice, saying that most wars among vassal states were unjustified, as they tended to serve one purpose only – eliminating their opponents. The battles listed here suggest how several wars launched during the third century bc were unjustified. In 293 bc, the State of Qin army defeated the united army of Han and Wei states at Yijue, killing 240,000 Han and Wei soldiers. In 273 bc, the State of Qin army defeated that of Wei state, killing 150,000 Wei soldiers. In 260 bc, the State of Qin army defeated that of Zhao state, burying alive 400,000 Zhao soldiers. In 251 bc, the State of Zhao deployed 200,000 men, with the army defeating 600,000 soldiers of the State of Yan at Haodai. Historians continue to debate the legitimacy of these records, questioning if they were reliable or not. One thing is clear: defeating the enemy was the constant objective on the minds of these territorial lords, regardless of the number of battles fought and the enormous causalities of war. While there has been no verification between archaeological discoveries and known historic events, archaeological data can still provide some insights into the brutal situations presented. In the south outside the Yanxiadu city site, archaeologists identified fourteen mass graves. One of the graves, which is in fact a very shallow pit 23 meters long, 0.6 meters wide, and 0.7 meters deep, contained over 2000 skulls and no other skeletal remains. Although the excavators believed that the mass graves are linked to the massacre of civil rebellions and social unrest in the early fourth century bc recorded in Shiji (Records of a Grand Historian), archaeologically there is no proof of the status of the dead. They may be war captives of the State of the Yan army. However, the following archaeological case suggests clear evidence of an unknown battle between states of Qin and Han. In 2006, a battlefield was discovered just outside the ancient capital of Han State, Xinzheng, at Longhu township of Henan province. It is a defensive stronghold with a nearly square trench, about 2 meters deep and 10 meters wide at the surface. Forty bodies were identified within a limited excavated area, but all without heads (Figure 5.11). Examining the remains of cervical vertebra shows a clear metal cut in all skeletons. There are in particular three overlying bodies buried in a small earthen pit; their postures suggest being killed while conscious. One of the bodies was completely burned, whereas skeletons from twenty-five bodies exhibit various wounds and breakages. Cultural remains were nothing but a few fragments of armour and metal arrowheads. More importantly, four bronze round coins identified as banliang – the State of Qin coinage – were recovered in the field, indicating the presence of the army from Qin State, coming from about 1000 kilometers away in the west. This is the first battlefield to be identified in China that has yielded so much information by way of identifying armies between two major warring states.

140

Figure 5.11 A newly discovered battlefield site in Longhu, near Xinzheng, the late ancient royal city of the State of Han, whose soldiers defended their capital from an aggressive invading army from the Qin State in the third century bc. All unearthed headless skeletons were possibly Han soldiers who lost their heads to Qin military personnel.

Chen Shen

The missing heads of war victims are no mystery once it is known that the opponents were from the State of Qin. The objective of Qin’s soldiers was to retrieve as many of the enemy’s heads as possible. Returning victoriously with them to obtain promotions allowed Qin soldiers to rise through the military ranks. This was one of the institutions established in reforms set forth by Shang Yang (see earlier) in the middle of the fourth century bc. This macabre military incentive gave soldiers the opportunity to alter their futures and improve their lives by elevating themselves from farmers and commoners to high-ranking military officials.Therefore, in the text records, whenever an army of Qin was winning a battle, their victims had to lose their heads. The discovery of the Longhu battle trench is the first vivid archaeological proof of this record. This testament also provided us with an understanding that reforms which took place during the Warring States period triggered electrifying acts by both territorial lords from the top and commoners at the bottom.The rest, which lie in the middle, had been able to seek out new opportunities in the age of transformation and had an opportunity to move up within the ranks, taking advantage of the unrest in the world. To live at a time when the future is unpredictable and filled with disorder allows one to long for a final resolution toward unification. Now was the time for determination and a need for unification, all of which were driven by the state’s political strategies, economic incentives, centralized government, and standardized systems. Each state competed for the best governing strategy advocated by various doctrines, choosing between Confucianist, Daoist, Mohist, Legalist, or others. Regardless of which governing doctrines were being adopted by the rulers, changes in government management were inevitable. As early as the fifth century bc, states like Qin, Jin, and Chu established a county system whose governors were appointed by the state lords instead of nobility enfeoffment. The county system was later expanded to a three-level hierarchy management, with a Xiang unit equivalent to villages and townships today at the bottom and a Jun unit like prefectures or provinces at the top. The network of governing systems provided substantial and sustainable supplies for manpower and resources to central governments rather than to local noblemen.The new system of centralized government also provided new opportunities for trade and market networks, which in turn gave way to economic incentives in the states. These states had further built economic networks and markets to allow trading of local resources (including some strategic resources like iron and salt) and precious products. But the trading system further stimulated the desire for standardization of currency, measures, weight, trading routes, and transportation across regions and even states. Under such circumstances, the blooming trade markets across commercial cities gave rise to affluent businessmen and entrepreneurs who were now heavily involved in political unrest and took advantage by supporting their next lordship or kingship to maximize profits or political returns. In a case like Lü Buwei 呂不韋 of the Zhao state, this wealthy businessman purposefully befriended the Qin prince, Zi Chu 子楚, who resided in Handan, the capital of Zhao State, as the state hostage. Lü invested his fortune on this hostage prince from the State of Qin. Later, Lü successfully returned Zi Chu to his mother in Xianyang, the capital of the prince’s home state, in 250 bc, which led to him ascending to the throne in the State of Qin. Zi Chu, later the King of Zhuang Xiang, was the father of Ying Zheng 贏政 (born 259 bc).Ying Zheng became the King of Qin state at the age of 13 and spent his first twenty-five years as king directing many battles, small or big, against Qin’s six rivals on the east side of the Yellow River. After defeating the State of Qi in 221 bc, he assumed the title huangdi, a title never before used in China, to represent his ruling over all heaven and earth. He had consolidated the territories that his ancestors, lords of the State of Qin, had worked to annex for several hundred years, and he intended his family line to rule the Qin Empire for 10,000 generations. Thus he is known to be the First Emperor of Qin, Qin Shihuangdi. 142

The age of territorial lords

Works cited Anhui Sheng Wenwu Kaogu Yanjiusuo (2013), ‘Chunqiu Zhongli jun mu fajue baogao 春秋钟离君墓发 掘报告’ [The report of the excavation at the tomb of Bai the lord of the Zhongli State in the Spring and Autumn period], Kaogu Xuebao Vol. 189. No. 2, pp. 239–282. Bai, Shouyi (1994), Zhongguo tongshi 中国通史 第三卷 [History of China Vol.3]， Shanghai: Shanghai People’s Press. Barbieri-Low, Anthony J. and Robin D.S.Yates (2015), Law, State, and Society in Early Imperial China, A Study with Critical Edition and Translation of the Legal Texts from Zhangjiashan Tomb no. 247, Leiden: Brill. Chen, Jian (2013), Zhanguo zhushu lunji 战国竹书论集 [Essays on Bamboo books of the Warring State period], Shanghai: Shanghai Guji Chubanshe. Chen, Pan (1997), Chunqiu dashibiao lieguo juexing ji chunmie yiyi 春秋大事表列国爵姓及存灭譔异 [Essays and Annotation on States during the Spring and Autumn period mentioned in the Chunqiu Dashi Biao], Taipei: Sanming Shujue. Chen, Peifeng (2004), Xia Shang Zhou qingtongqi yanjiu 夏商周青铜器研究 [Studies of Xia-Shang-Zhou Bronzes], Shanghai: Shanghai Guji Chubanshe. Chen, Ping (2005), Guanlong wenhua yu yingqin wenming 关陇文化与赢秦文明 [Quanlong Culture and Qin Civilization]， Nanjing: Jiangshu Education Press. Cook, Scott (2012), The Bamboo Texts of Guodian: A Study & Complete Translation. Ithaca, NY: Cornell University East Asia Program. Fan, Wenlan (1949), Zhongguo tong shi jian bian 中国通史简编 [Brief history of China], Beijing: People’s Press. Fang, Hui and Chen Shen (1999), ‘Ji Huangjia Andalüe bowuguan shoucang de yijian huaxiang qingtongqi 记皇家安大略博物馆收藏的一件画像青铜器 [on a pictorial bronze vessel from the collection of Royal Ontario Museum]’ Gugong Wenwu Yuekan Vol. 194, pp. 68–77. Fang, Qing and Wu Hongtang (2015), mumu zenghou: zaoyang guojiammiao zengguo mudi 穆穆曾侯:枣阳 郭家庙曾国墓地 [Cemeteries of Zeng State at Guojiamiao, Caoyang City]. Beijing: Cultural Relics Press. Gao, Guangren and Shao Wangping (2005), Haidai wenhua yu qilu wenming 海岱文化与齐鲁文明 [Haidai Cultures and Qi-Lu Civilizations], Nanjing: Jiangshu Education Press. Gu, Derong and Zhu Shunlong (2001), Chunqiu Shi 春秋史 [History of the Spring and Autumn], Shanghai: Shanghai’s People Press. Guo, Dewei (1995), Chuxi muzang yanjiu 楚系墓葬研究 [A Study of Chu Cemeteries], Wuhan: Hubei Education Press. Guojia Wenwuju (2016), Zhongguo zhongyao kaogu faxian 中国重要考古发现 [Major Archaeological Discoveries in China in 2015], Beijing: Cultural Relics Press. Han, Guanghui (1996), ‘Qidu Linzi hukou kaobian 齐都临淄户口考辨 [Analyses of populations at Linzi, Qi State capital]’ Guanzi Xuekan No.4, pp. 25–30. Hebei Wenwu Kaogu Yanjiusuo (1996), Yanxiadu 燕下都 [the lower capital of Yan State]. Beijing: Cultural Relics Press. Hubei Provincial Museum (2007), Jiuliandun: Changjiang zhongyou de Chuguo guizu damu 九连墩:长江中 游的楚国贵族大墓 [Jiuliandun: large tomb of a Chu Noble in the Middle Reaches of the Yangtze], Beijing: Cultural Relics Press. Hsu, Cho-yun (1965), Ancient China in Transition, Stanford: Stanford University Press. Hsu, Cho-yun (1999), ‘The Spring and Autumn Period,’ in Michael Loewe and Edward L. Shaughnessy (eds.) The Cambridge History of Ancient China: from the Origins of Civilizations to 221 B.C, Cambridge:The University of Cambridge Press, pp. 545–586. Irwin, Sara and Chen Shen (2016), ‘A Question of Provenance: The Bishop William White “Jincun” Collection in the Royal Ontario Museum,’ Orientations Vol. 47, No. 3, pp. 34–41. Jiang, Gang (2002), ‘Dongzhou shiqi zhuyao lieguo ducheng renkou wenti yanjiu 东周时期主要列国都 城人口问题研究 [A study of populations at capital cities during the Eastern Zhou period]’ Wenwu Chunqiu.Vol. 68. No. 6 (2006), pp. 6–14. Lewis, Mark Edward (1999), ‘Warring States: Political History,’ in Michael Loewe and Edward L. Shaughnessy (eds.) The Cambridge History of Ancient China: from the Origins of Civilizations to 221 B.C., Cambridge: The University of Cambridge Press, pp. 587–650. Li, Feng (2003), Landscape and Power in Early China, Cambridge: Cambridge University Press. Li, Feng (2014), Early China: A Social and Cultural History, Cambridge: Cambridge University Press.

143

Chen Shen Li, Ling (2008), Jianbo gushu yu xueshu yuanliu 简帛古书和学术源流 [Archaic bamboo and silk books and the origins of intellectual studies], Beijing: SDX Joint Publishing Company. Li, Xueqing (1985), Eastern Zhou and Qin Civilization, New Haven:Yale University Press. Liu, Dunyuan (1990), ‘Zhongguo qingtongqi shang de caisang tuxiang 中国青铜器上的采桑图像 [on the image of mulberry picking from pictorial bronze vessels of China]’ Wenwu Tiandi No. 5, pp. 4–6. Loewe, Michael and Edward L. Shaughnessy (eds.) (1999), The Cambridge History of Ancient China: From the Origins of Civilizations to 221 B.C, Cambridge: Cambridge University Press. Luoyang Wenwu Gongzuodui (2009), Luoyang Wangcheng guangchang Dongzhoumu 洛阳王城广场东周墓 [Eastern Zhou tombs from the Wangcheng Square in Luoyang], Beijing: Cultural Relics Press. Ma, Chengyuan (2002), Zhongguo qingtongqi yanjiu 中国青铜器研究 [A Study of Chinese Bronzes], Shanghai: Shanghai Guji Chubashe. Ma, Chengyuan (ed.) (2012), Shanghai bowuguan cang Zhanguo Chu zhushu 上海博物馆藏战国楚竹书 [Warring State bamboo book collections of Chu State from Shanghai Museum], Vol 9, Shanghai: Shanghai Guji Chubanshe National Museum of History and Henan Museum (2001), Bronzes from the Prince Zheng Tomb, Xin Zheng, Taipei: National Museum of History Pu, Maozuo (2007), ‘Jing gong nue,’ in Ma Chengyuan (ed.) Shanghai bowuguan cang zhanguo chuzhushu 上海博物馆藏战国楚竹书 [Warring State bamboo book collections of Chu State from Shanghai Museum], Vol 6, Shanghai: Shanghai Guji Chubanshe, pp. 157–191. Shandong Wenwu Kaogu Yanjiusuo (1975), ‘Linyi Yinqueshan sizuo Xihan muzang 临沂银雀山四座西汉 墓葬’ [Four tombs of the Western Han from Yinqueshan in Linyi County], Kaogu No. 6, pp. 363–379. Shandong Wenwu Kaogu Yanjiusuo (2013), Linzi Qi gucheng 临淄齐故城 [Linzi City Site of Qi State], Beijing: Cultural Relics Press. Shanxi Sheng Kaogu Yanjiusuo (1996a), Jindu xintian 晋都新田 [Xintian: the Capital of Jin State], Taiyuan: Shanxi People’s Press. Shanxi Sheng Kaogu Yanjiusuo (1996b), Taiyuan Jinguo zhaoqing mu 太原晋国赵卿墓 [Tomb of Jin State Minister Zhao at Taiyuan], Beijing: Cultural Relics Press. Shao, Wangping (2013), Shao Wangping shixue kaoguxue wenxun 邵望平史学考古学文选 [Selected Essays of Shao Wangping on History and Archaeology], Jinan: Shandong University Press. Shen, Chen (1994), ‘Early Urbanization in the Eastern Zhou in China (770–221 bc): An archaeological view’, Antiquity Vol. 68, pp. 724–744. Shen, Chen (2004),‘Compromises and Conflicts: Production and Commerce at the Royal Cities of Eastern Zhou, China’, in Monic L. Smith (ed.) The Social Construction of Ancient Cities, Washington, DC: Smithsonian Books, pp. 290–310. Shen, Chen (2010), The Warrior Emperor and China’s Terracotta Army, Toronto: Royal Ontario Museum Press. Shaughnessy, Edward D. (1991), Sources of Western Zhou History: Inscribed Bronze Vessels Berkeley: University of California Press. Shi, Quan (1979), ‘Gudai Zengguo – Suiguo diwang chutan 古代曾国-随国地望初探’ [Geopolitical assessments of Zeng State or Sui State] Wuhan Daxue Xuebao No. 1. So, Jenny (1995), Eastern Zhou Ritual Bronzes from the Arthur M. Sackler Collections, Washington, DC: Arthur M. Sackler Gallery, Smithsonian Institution. Tong, Shuye (1946), Chunqiu shi 春秋史 [History of the Spring and Autumn], Shanghai: Kaiming Books. Teng, Mingyu (2002), Qin diguo: Cong fengguo dao diguo de kaoguxue guancha 秦帝国:从封国到帝国的考古 学观察 [From Vassal state to empire: an archaeological examination of Qin culture], Beijing: Xueyuan Chubanshe. von Falkenhausen, Lothar (1999), ‘The Waning of the Bronze Age: Material Culture and Social Development, 770–481 B.C.’ in Michael Loewe and Edward L. Shaughnessy (eds.) The Cambridge History of Ancient China: from the origins of civilizations to 221 B.C., Cambridge: The University of Cambridge Press, pp. 450–544. von Falkenhausen, Lothar (2006), Chinese Society in the Age of Confucius (1000–250 bc): The Archaeological Evidence, Los Angeles: University of California Los Angeles, Costen Institute of Archaeology. White, William C. (1934), Tombs of Old Lo-Yang, Shanghai: Kelly & Walsh. Limited. Xian Shi Wenwu Baohu Kaogusuo (2006), Xi’an nanjiao Qinmu 西安南郊秦墓 [Cemeteries of the Qin State near southern suburban of Xian], Xi’an: Shaanxi People’s Press. Xu, Hong (2000), Xianqin chengshi kaoguxue 先秦城市考古学 [An Archaeological Research on Cities in Pre-Qin period], Beijing:Yanshan Publishing House.

144

The age of territorial lords Xu, Shaohua (1994), Zhoudai nantu lishi dili yu wenhua 周代南土历史地理与文化 [Historic-Geographic and Cultures of Southern Lands in Zhou Dynasty], Wuhan: Wuhan University Press. Yang, Jianhua (2004), Chunqiu Zhanguo shiqi Zhongguo beifang wenhuadai de xingcheng 春秋战国时期中国北 方文化带的形成 [the Formation of Northern Chinese Frontier Belt during the Spring and Autumn and Warring States Periods], Beijing: Cultural Relics Publishing House. Yang, Kuan (1998), Zhanguo shi 战国史 [History of the Warring States], Shanghai: Shanghai’s People Press. Yuan, Zhongyi (1980), ‘Qindai de shiting taowen 秦代的市亭陶文 [Marks on Shi and Ting from pottery of the Qin State]’, Kaogu Yu Wenwu Vol. 1, No. 1, pp. 93–98 + 72. Zhang, Changping (1998), ‘Anju Zhoudai chengzhi de faxian jiqi yiyi 安居周代城址的发现及其意义’ [Discovery of Zhou period city-ruin at Anju and its implications], zhongguo wenwu bao August 26. Zhang, Changping (ed.) (2007), Zengguo qingtongqi 曾国青铜器[Bronzes from State of Zeng], Beijing: Cultural Relics Press. Zhao, Congcang and Guo Yanli (2004), Liangzhou kaogu 两周考古 [Archaeology of Western Zhou and Eastern Zhou], Beijing: Cultural Relics Press. Zhongguo Shehuikexueyuan Kaogu Yanjiusuo (2004), Zhongguo kaogu xue: Liangzhou juan 中国考古学: 两周卷 [Chinese Archaeology: Western Zhou and Eastern Zhou], Beijing: China Social Sciences Press. Zhu, Fenghan (1995), Gudai Zhongguo qingtongqi 古代中国青铜器 [Bronzes of Ancient China], Tianjing: Nankai Daxue Chubanshe.

145

6 THE QIN DYNASTY

CHARLES SANFTTHE QIN DYNASTY

Charles Sanft

Song dynasty scholar Sima Guang (1019–1086) – the greatest historian in China between the Han dynasty and the modern period – began his comprehensive account of China at the end of the fifth century bce.1 The Zhou dynasty was fading, and it was then, Sima Guang explains, that the states of Qi, Chu, Jin, and Qin began to expand. Sima Guang ended the annals of the Zhou dynasty and began those of the Qin in the middle of the third century bce because in 256 bce the state of Qin defeated the Zhou and began its final ascent to supremacy. Looking back over more than a millennium of imperial history, Sima Guang saw the path of empire leading from Zhou to Qin and on to Han. Writing over a thousand years earlier, Sima Guang’s ancestor, the Han-era historian Sima Qian (ca. 145–ca. 86 bce) had conveyed a different image in his Historian’s Records. Sima Qian presented as context a panoply of states in conflict with each other and gave individual attention to them and their rulers. No few have devoted chapters among what Sima Qian called the ‘Hereditary Houses’. This was part of Sima Qian’s concentration on the Qin emergence as the product of competition and struggle and above all violence. He did something similar by giving a chapter of ‘Annals’ to the primary rival of the Han dynasty founder, Xiang Yu, putting him alongside the rulers of the Qin and Han dynasties. Sima Qian’s narrative of Qin origins in the Historian’s Records is the usual starting place for discussions of the Qin. It links the origins of the Qin ruling lineage to myth, tracing its roots to Nüyou, fabled descendant of the mythological Thearch Zhuanxu. According to the Historian’s Records, when this Nüyou was once weaving, a mysterious bird laid an egg. She swallowed it and subsequently gave birth to a child, Daye. It is to this Daye that the Historian’s Records traces the descent of the Qin ruling house. This account connects Qin to the rulers of the Shang by means of Zhuanxu, who is supposed to be the Shang ancestor. It would make the Qin a collateral branch of the same lineage as the Shang dynasty rulers. The rest of the story of the Qin clan’s origins is only a bit less nebulous. So deep is the confusion that early modern scholars held two precisely opposite viewpoints about where the Qin arose originally: some argued that the Qin came from the west, while others held that they came from the east (Ma 1982, 3–4). In the absence of additional information in the form of archaeological evidence, trying to establish an origin seems futile. And finding reliable material evidence of Qin origins is, in the words of one archaeologist who has sought it, ‘extremely difficult’ (Zhao 2014, 67). All this validates Sima Guang’s approach, which is to start 146

The Qin dynasty

the story at the end of the fifth century bce and focus on developments in the fourth and third centuries. Not only is that firmly within the time of written records, it is also when the rise of the Qin to power takes place.

Conquest The Qin enter Sima Guang’s narrative with a military action in 401 bce. That year, from their state in the west, in what is now Shaanxi, the Qin attacked the state of Wei. Qin appearances in the following decades of history are primarily military actions: success against Hann in 391 bce; a battle with Jin in 390 bce followed by an invasion the next year; the capture of territory from Shu in 387 bce; and a series of Qin victories in the 360s bce. It is at around this time that two men emerge into the Qin historical narrative and give shape to all that follows. In 362 bce, Lord Xiao of Qin succeeded his father as ruler of Qin and announced his intention to resume what he called his forefathers’ project to expand Qin power. To that end, he offered land and high position to anyone who could propose plans that would strengthen Qin. In response, Shang Yang (ca. 385–338 bce) went to Qin from the state of Weih. Shang Yang had ideas about how to strengthen the army and enrich the Qin state that pleased Lord Xiao, and he quickly became involved in its government. Traditionally scholars associate Lord Shang with proposing and carrying out alterations to the Qin legal system known as ‘changing the laws’. These reforms actually went much further than this might seem to imply, for they brought about deep shifts in Qin society. The first set of changes came in 359 bce.The government instituted a series of twenty ranks that spanned all of Qin society, with privileges such as age of retirement and reduction in or release from corvée labour for those at certain ranks. Every male Qin subject held one of these ranks and could win advancement through success in battle and other contributions to the general good as the state defined it. The reforms also strengthened sumptuary rules governing clothing and lands, with the goal of clarifying social status without respect to wealth. Qin law at this point arranged commoner households into groups of five, with statutory requirements for members to report each other’s illegal acts. This reporting would bring special reward equal to that for beheading an enemy in battle, while failure to report a crime was punishable with death.The laws were changed to encourage men to farm and women to weave, and those whose productivity was outstanding enjoyed reduced corvée requirements. One of the most thoroughgoing and enduring changes to Chinese society also dates to around this same time. The Qin established for the first time a population registry that was intended to include every single person. Archaeologists have recovered registration documents from the third century bce that reflect a system of just this configuration, recording men and women, children and adults, and including slaves (Sanft 2015). According to an account that Sima Qian included and Sima Guang repeated, Shang Yang doubted that the people of Qin would put stock in the new statutes and came up with a plan to ensure they would. He commanded that a tree trunk be placed at the southern gate of the capital and posted a sign offering reward for moving it. When no one dared move the log, he raised the reward, and paid it when someone finally shifted it. Whether true in its details or not, this reflects the importance that the Qin placed on their text-based legal system and its acceptance by the general population. In 350 bce, Shang Yang proposed and carried out a second round of changes to Qin state governance, focused around the reorganization of land practices. They moved the capital to Xianyang, near modern Xi’an (Shaanxi). Around the same time, the Qin state divided up the small communities in which most of its population lived. They created thirty-one prefectures, 147

Charles Sanft

each with a prefect and an assistant to run it as part of a state-wide bureaucracy. Even the fields, upon which this agrarian society depended, were reorganized and assigned to households – a reorganization quickly followed by the establishment of a land tax. The goal of these changes was to give Qin the strength necessary to fight with its competitor states to the east. For all along, even as these internal changes were under way, Qin was engaged in a protracted military struggle for supremacy in the realm. Shang Yang’s reforms brought quick results, putting Qin into a position of strength that none of its rivals could match. Within a few years, through a combination of military force and deception, Qin repeatedly defeated the forces of Wei, which had once been the most powerful state, and took a great deal of its territory. When Lord Xiao died in 338 bce, his son succeeded him as ruler of Qin. There was another shift, as well, which was indicative of changing politics among the states. For Lord Xiao’s son, like many of his contemporaries, would rule as king and not merely as lord. This signalled a claim to power and rejection of nominal Zhou sovereignty. As King Hui, he got rid of Shang Yang, executing him due to reports that the Shang Yang sought to overthrow him. The systems put into place at the advice of Shang Yang, however, remained and indeed expanded. Historians have often depicted Qin dynasty rule as harsh and totalitarian, something that owes much to the actions of Qin at the end of the Warring States period. That picture is changing as we learn more about the details of Qin governance from archaeologically recovered sources and the reconsideration of received historical accounts. Rather than totalitarian, the emergence of Qin and its dominance over the other states has come to seem like the ascendance of highly competent technocrats operating on a field without anyone else quite up to their level. With new information and objective analysis, even the changes associated with Shang Yang, which writers long criticized, emerge as sophisticated and highly cooperative enterprises (Sanft 2014b). The slow rise of the Qin bureaucratic state to dominance began no later than the fourth century bce and accelerated over time. The background of Qin’s gradual expansion in the late fourth and into the third centuries bce was a mix of military conflict and diplomacy. As the Qin gradually expanded their territory through force, they became involved in much but not all of the conflict. Even as the Qin did so, they evoked a response from the other states, which slowly began to realize their peril from the highly effective governance of Qin. That did not stop the states from fighting each other and attempting to manoeuvre for advantage, in doing so even sometimes joining with Qin against others. King Hui died in 311 bce, and his son King Wu succeeded him, followed in turn by several others. The other states slowly realized that Qin was a real threat, yet they still did not effectively join together. Even after the state of Qin briefly declared its king the Western Emperor in 288 bce and suggested that the king of Qi, the other most powerful state, call himself Eastern Emperor, there was intrigue and small-scale conflict among the Qin’s competitors. Qin remained measured in its movements, which perhaps prevented panic among the other states. As the first stage of Qin’s emergence began with Lord Xiao and Shang Yang, so did its thirdcentury culmination similarly rely upon two men: King Zheng (r. 246–210 bce) – who would become the First Emperor of the united realm – and Lü Buwei (291–235 bce). Lü Buwei’s part in the narrative begins already in the time when King Zheng’s father was still a youth. The future King Zhuangxiang was the middle son of the Qin king and then a hostage in Zhao, where Lü Buwei was. Lü Buwei recognized an opportunity in the prince and engineered the future King Zhuangxiang’s return to Qin and establishment as heir designate. In 250 bce, Zhuangxiang ascended the throne and made Lü Buwei his chancellor.When Zhuangxiang died a few years later, his son King Zheng came to rule, and Lü Buwei remained chancellor. It was 246 bce. 148

The Qin dynasty

Lü Buwei was a man of great and varied talents. In addition to success in politics and trade, he was also a philosopher. He led the compilation of a text that bears his name, Mr. Lü’s Annals, a compendium that gives some impression of the breadth of intellectual life in the late third century bce. Sima Qian writes that while King Zhuangxiang was young and a hostage in Zhao, he took a fancy to Lü Buwei’s concubine. Lü gave her to him, and she not too long later gave birth to the future King Zheng. According to the story, she was pregnant already before going to Zhuangxiang. Thus, the story goes, Lü Buwei was the father of the eventual First Emperor. Such a tale is salacious, enjoyable, and deserving of no credence. Yet it reflects the undeniable truth that there was something very special about King Zheng, a man who combined great ability with greater ambition in a manner reminiscent of Lü Buwei, and about the relationship between the two. When he became king in 246 bce, Zheng was too young to rule.The queen mother and the high officials, particularly Lü Buwei, made decisions in his place. The Qin rulers immediately broadcast a summons across the realm for talented persons who wished to join them.Their goal, according to Sima Qian, was by that time explicitly to unify the realm under Qin. One of the men who went to Qin around this time was an engineer. The state of Hann, one of Qin’s rivals, had sent him to persuade the Qin to persuade them to carry out extensive works that would sap Qin’s strength.The engineer convinced the Qin to build a tremendous canal.When the Qin realized his duplicity, the engineer protested that the project would in fact serve Qin. The Qin let him finish it, and their farmlands benefitted from the irrigation the canal provided. This tells us something about the Qin focus on practical considerations over appearances. The first years of King Zheng’s nominal rule carried forth the pattern of previous decades. Lü Buwei’s influence and wealth continued to expand, and there was alternating conflict and peace between Qin and the other states. Meng Ao, originally from the state of Qi, had served as high official and general under earlier Qin kings and became an important part of Qin’s military expansion. As general, he helped subdue a revolt in the first year of King Zheng’s rule, then went on to win important victories by taking thirteen cities in the state of Hann in 244 bce and laying extended siege to two cities in Wei beginning that same year, taking them the following. In 242 bce Meng Ao invaded Wei again, capturing a great number of cities – twenty or thirty, depending on the record you read. Qin turned the conquered territory into a commandery, an administrative division that the Qin government ruled through a bureaucracy. The other states had been fighting each other all along, but this annexation of territory scared them. In 241 bce, five of the states (Chu, Zhao,Wei, Hann, and Weih) banded together to strike Qin. Their campaign turned into a rout when the combined force fled rather than face Qin troops at Hangu Pass – an embarrassment that passed into legend. Qin, for its part, continued to conquer territories in Wei and Weih that year, with still more in Wei the next. Meng Ao died in 240 bce, but Qin encroachment into Wei territory did not stop. They defeated and absorbed further cities in 239 and 238 bce. In the ninth year of his reign, 238 bce, King Zheng underwent the ritual ‘capping’ that formally made him an adult. Since those who wielded power did not have official regency, there was no orderly process by which King Zheng could come to rule in fact as well as name. The transfer of power was instead chaotic, and it is hard to tell truth from tale in the records. In one version, the story goes back to the start of King Zheng’s reign in 246 bce. Supposedly Lü Buwei was carrying on an affair with the queen mother and, as the king grew older, became fearful of discovery. Lü Buwei disguised an accomplice, a certain Lao Ai, as a eunuch and presented him to the queen mother in the expectation that Lao Ai would act as his agent. The queen mother liked Lao Ai so much she became pregnant by him, twice. She granted him territory and gave him rein to decide matters of government. Later, shortly after the king 149

Charles Sanft

had reached adulthood, Lao Ai argued with someone in the king’s entourage, and that person denounced him to the king as a false eunuch. When the king sought Lao Ai’s prosecution, Lao Ai falsified the king’s seal and led Qin troops in rebellion. Two high officials went out with soldiers, killed hundreds of Lao Ai’s men, and eventually captured Lao Ai himself. Lao Ai was executed and his clan wiped out, his associates executed or banished. King Zheng killed the queen mother’s two children by Lao Ai and locked her in a palace, although he released her when reminded of how it would look for the king to imprison his own mother. Lü Buwei was implicated in this matter and dismissed in 237 bce. The king initially did not want to execute someone who had done as much for the state of Qin as Lü Buwei had.Yet he came to fear Lü’s power. Two years later Lü Buwei committed suicide after receiving a letter from King Zheng that indicated his contributions to the Qin state were no longer valued. In 237 bce, prominent members of the Qin royal lineage recommended expelling from the state the many men from elsewhere who had come to Qin, on the suspicion that they were acting on behalf of other rulers. One of those caught up in the ensuing purge was Li Si, a native of the southern state of Chu, who was serving as an official in Qin. As he made his way out of Qin, Li Si submitted a letter to the throne in which he pointed to historical precedents of men who aided another state’s king and the king benefitted. Li Si also noted the irony of Qin’s king surrounding himself with musicians and beautiful women from other states while sending away useful men who would then go help his rivals. His conclusion compares this course of action to arming and provisioning bandits. The letter persuaded the king, who summoned Li Si, restored him to his official position, and cancelled the expulsion of non-Qin persons. Despite having demonstrated some inkling of the threat that Qin posed to their existence, the other states continued mostly to fight among themselves. Qin took advantage of the opportunities this presented. When Zhao attacked Yan – a smaller state to the northeast – in 236 bce, for instance, three Qin generals and their forces invaded Zhao and captured a number of its walled cities.Two years later, in 234 bce, one of those same generals led yet another force against Zhao but after initial victories was defeated.Yet then in 233 bce he again struck Zhao and captured three more towns. That same year, the state of Hann turned over lands and its king’s seal to Qin, asking to become a vassal of Qin, which was by then rapidly emerging as the dominant power in the realm. Hann would cede still more lands to Qin in 231 bce, as would Wei. But Hann efforts at self-preservation were unsuccessful. Qin destroyed Hann, captured its king, and turned its territory into a commandery in 230 bce. Already in 232 bce, King Zheng had sent several of his armies to invade the state of Zhao. Zhao was able at first to repel the Qin incursion, but famine struck in 230 bce. Still, the best Zhao general repelled a Qin offensive in 229 bce, which led the Qin to change tactics. They bribed the Zhao king’s favourite to slander the successful general, leading the Zhao king to order his replacement. When the general refused to accept removal, he was arrested and killed. The Qin revisited Zhao the following year, destroying its army and capturing its king. A final member of the Zhao ruling lineage took a few hundred soldiers and fled to the state of Dai, where they joined forces with the state of Yan. A prince of Yan worried about Qin vengeance and decided on a stratagem to strike at the heart of the enemy. In 228 bce he dispatched an assassin to kill the king of Qin. The attempted stabbing of King Zheng happened during an audience in 227 bce, but the king managed to evade death. In retaliation King Zheng sent more troops to Zhao and ordered one of his generals to attack Yan, where Qin defeated the combined forces of Yan and Dai. The following year the king of Yan killed the prince and sent his head to King Zheng in hopes of conciliation. The Qin king only redoubled his efforts. 150

The Qin dynasty

In 225 bce the Qin general leading the invasion of Wei flooded its capital, destroying it. The king of Wei committed suicide, marking the end of the state as an independent polity. Around the same time, the Qin struck Chu to the south. But the force for the invasion of Chu was of insufficient size, and the defenders repulsed it. King Qin dispatched a more senior general and a larger body of troops in 224 bce. In 223 bce, Qin took over the territory of Chu and turned it, too, into a commandery. As part of the same campaign they pushed the next year still further south, into the lands of non-Chinese peoples, creating there another commandery. That same year, 222 bce, another Qin general captured the kings of Yan and Dai. At the end of 222 bce, only the state of Qi remained to challenge Qin. Qi had preserved itself through a combination of subservient and harmonious relations with Qin and maintaining good faith in its dealings with the other states.This approach had preserved the king of Qi safely on his throne for more than forty years. Yet once the other states were gone, Qin’s attention inevitably turned to Qi. The fall of Qi, the end of the Warring States period, and the military unification of the realm in the twenty-sixth year of King Zheng’s reign was quick and prosaic in comparison to the sanguinary elimination of other states. A Qin general marched southward into Qi from territory to its north that formerly belonged to the state of Yan. The sudden arrival of those troops in the Qi capital went unopposed. The Qin then offered the Qi king a grant of territory in exchange for surrender, and the king accepted. The king was, however, instead immediately imprisoned and starved to death shortly thereafter.The Qin turned the Qi lands into commanderies and the takeover was complete. It was 221 bce.

Empire The Qin dynasty was by all accounts a self-conscious dynasty. Its founder conceived of himself as beginning a line of rulers that he saw stretching out into the future and known by number rather than name. He used a new title, emperor, and called himself the First Emperor, with his successor to be the Second Emperor and the Third to come after that. This choice of title reflects the Qin willingness to do things in a new way. Indeed, the whole Qin dynasty was a period of innovation, in that they used existing systems and practices in ways or to extents that previous rulers had not. On the other hand, the process of unification was in many respects one of bringing to the realm systems that had already existed in the state of Qin. The following inscription embodies both sides of Qin rule: In the twenty-sixth year of his reign, the emperor unified the lords of the realm, the common people had great peace, and he established the title of emperor. Now he commands chancellors Wei Zhuang and Wang Wan: ‘As for the laws and units of measure that are disparate or doubtful, in all cases clarify and unify them’. (Sanft 2014a, 59) Archaeologists have found examples of this inscription in many places, reflecting that the Qin dispersed the announcement of the realm’s new condition across the whole of their state in a variety of forms. The system of weights and measures that the inscription refers to was also that of the state of Qin. These became the new standard for the realm, replacing the variety of systems previously in place – at least ideally. The inscription’s dating to the twenty-sixth year links it to the reign of the erstwhile king and does not declare it the first year of the new emperor. Yet it refers to him by the title emperor. This widely broadcast announcement’s simplicity made it suitable for 151

Charles Sanft

easy comprehension, while the change it commanded made it likely to attract notice from the peasants whose taxes were paid in grain and weighed in the new units. The unification of the former states under the emperor raised the question of governance. Most of the high officials, including the chancellor, proposed appointing the emperor’s sons to rule the distant parts of the new realm as lords. Li Si offered the sole recorded dissent, blaming the violence of the Warring States on the Zhou pursuit of a similar policy. Li Si proposed instead that the system of commanderies and prefectures should encompass the whole of the realm, with appointed officials to administer them on behalf of the imperial government.The emperor followed Li Si’s advice and established a bureaucratic form of governance centred on administrative divisions controlled by officials.2 The systems of social ranks and grouped households likewise expanded across the realm. The First Emperor in 220 bce set out on the first of a series of travels through his territories. The first trip was short and took him through lands that were originally part of the Qin state. The remainder of his years were spent alternately in the capital and on the road on one of the four subsequent journeys that took him quite far from the capital. The emperor’s travels come to structure the historical narrative of his rule over the empire. And while scholars sometimes refer to these trips as inspection tours, the indications are that the emperor travelled to be seen at least as much as to see himself. The First Emperor’s travels did not interrupt his habitual close engagement with the details of bureaucratic governance or lead to any slowdown in the pace of Qin changes and developments. In this same year as the first journey, there were construction projects that included the Supreme Temple (Jimiao) and the extension and improvement of roads into a highway system that linked together existing roads to reach the ends of the realm. The emperor spent much of 219 bce travelling through the eastern portion of the realm. Early in this trip, the emperor set up for the first time a stone marker with an extensive and elaborate inscription at Mount Zouyi describing and praising the First Emperor’s deeds. After this, the First Emperor summoned some seventy ritual experts and travelled to the famous Mount Tai (Shandong), where the experts suggested he carry out a ritual offering. But the dozens of experts together produced a variety of bothersome and contradictory recommendations as to method that irked the emperor. He ignored them all and instead simply drove up the mountain, put up another stele, and drove back down. He borrowed the rituals for his sacrifices from those for other offerings. Then the First Emperor headed still further east. His travels took him all the way to the ocean, and he followed the coast, performing offerings to mountains and rivers, even building a terrace at a favourite spot, Langye (Shandong), and erecting still another stone marker. Then he made his way back to the capital, travelling on land and by river. One of the most famous stories told about the First Emperor is part of this trip, and it reflects simultaneously the hubris, violence, and willingness to waste his subject’s labour that later accounts attribute to him. While making his way back to Xianyang, his capital, the First Emperor was going to make a sacrifice to Mount Xiang (Hunan), on an island. While crossing the Yangtze River, the emperor encountered strong winds that nearly sunk his boats. The emperor was incensed and, upon learning the identities of the spirits said to dwell in the mountain, he decided to punish them. He sent some three thousand convict labourers to denude the mountain, cutting down its trees and exposing its soil as punishment for the spirits. The next year, in 218 bce, the First Emperor travelled east once more. While on the road he escaped assassination only because the would-be assassin struck the wrong chariot. After commanding a search for the assassin, the emperor continued his travels. He ascended Mount Zhifu (Shandong) and commanded the inscribing of a text onto stone there. He again visited Langye and then travelled once more to the eastern coast, where he visited Jieshi (Liaoning) and left 152

The Qin dynasty

yet another inscription in the stone there. As he travelled, he commanded the destruction of dikes and defensive walls before heading north to visit the border area. Then he returned to the capital. An account from this year gives a further impression of the types of stories that circulate about the First Emperor. He sent a man off to explore the sea, and the man came back with a mysterious document bearing the cryptic message, ‘The destroyer of Qin will be Hu’. The First Emperor misunderstood this warning due to an ambiguity in its phrasing. The word Hu often denoted non-Chinese people, meaning something like ‘barbarian(s)’. But it was also part of the name of one of the emperor’s sons, Huhai, who would in fact preside over the main part of the Qin empire’s collapse a few years later. Thinking this was Hu in the first sense, the emperor sent an army under Meng Tian – grandson of Meng Ao – against the ‘barbarians’ living north of the empire. A sceptic might point out that the First Emperor had just visited that very area and perhaps noticed a need for military action and that the pun is a bit too convenient to be more than a storyteller’s device.Yet this story often appears. The northern campaign of 215 bce continued into 214 bce as the Qin pushed their seminomadic northern neighbours back to the Yinshan Mountains, a sort of cultural demarcation between Chinese and non-Chinese peoples (Di Cosmo 2002, 88). Meng Tian took still other lands. The Qin built walls and military fortifications and turned some of the territory they captured into prefectures. Expansion along the southern border of the reign accompanied this. The government dispatched persons who had fled their places of registration and had been recaptured and other undesirables to territory south of the empire, which became commanderies. The Qin banished people to populate their new territories in the north and the south, and sent still more to both places in 213 bce. An event that counts among the most infamous occurrences in Qin history – indeed one of the most decried events of all Chinese history – also dates to 213 bce: the notorious Qin burning of the books. According to the received account, Li Si wrote to the emperor, asserting a connection between political contention and a lack of productive work with the presence of scholars who studied and taught ancient texts. He proposed destroying historical records other than Qin’s own annals and burning the texts of the classics and many other works, except for the copies that the official erudites held. Li Si furthermore proposed banning the use of ancient texts or examples to criticize the Qin. In laconic, legally correct fashion, the First Emperor authorized Li Si’s proposal with one word: ‘Approved’. The reputation of Qin would never recover. Many writers, down to the present day, treat this as a description of an attempt to destroy the classical corpus. That overlooks a few things. First, Li Si never proposed the eradication of the classics. The text of his proposal as we have it would limit the copies to those held by official specialists at the capital. That would be a significant restriction, but that is something quite different from total destruction. There is also the obvious point that the enforcement of such a decree would have been very difficult. Sima Guang included in his narrative assertions attributed to an eighth generation descendant of Confucius alive when the edict took effect, who avers that he was unlikely to face an official search but would hide his copies of the classics anyway, just in case. If this story deserves credence, it is not as an account of the persistent bibliophilia of Confucius’ lineage. Rather it is as a reflection of the fact that such a subterfuge would have been within the powers of nearly anyone wishing to retain copies of those texts. That scholars and students would have committed to memory much of what was in fact a relatively limited corpus would also act as a barrier to the destruction of the classics. One may well criticize the Qin for seeking by means of coercion to limit access to particular texts and assert control over particular kinds of discourse 153

Charles Sanft

and writing. But this is something quite different from what many later historical accounts allege. Such is the case even without considering the more sophisticated critiques of the event’s portrayal levelled by revisionist historians (Petersen 1995). Accounts of the year 212 bce mix the two manias – violence and massive construction – that historians have attributed to the First Emperor ever since the Han dynasty. It was in this year that he constructed the most famous segment of the Qin highway system, the famous Direct Road (Zhidao), which ran from an area near the capital (and itself linked to the capital) to the northern border. The Qin built it running atop mountains and across valleys along the area of the western border to end in the region of the Yinshan Mountains that marked a cultural boundary between China and the steppes peoples. While according to standard accounts the emperor merely wished to travel the realm, it seems that such a construction would also signal Qin power and capability to observers inside and outside the realm. Archaeological work on reliably identified Direct Road remains indicates that the road in many places followed existing ways. Yet there are sections of it that run across the desert in the north that seem to have been built by the Qin. The histories tell us the road was not complete. Histories of the Qin furthermore record that the First Emperor commanded the construction of a new capital on the other side of the Wei River next to Xianyang.That project began in 212 bce with work on the Epang Palace.3 Planned as the fore-palace of an even larger structure, it was to be rectangular and roughly 690 by 116 meters. Although by all accounts the palace was never completed, most people assumed it lacked only finishing touches. Writers exercised their imaginations in describing it, and the Tang dynasty poet Du Mu (803–852 ce) wrote describing the extent of its vain and opulent extravagance. Modern archaeologists destroyed these figments when their excavations revealed that the palace’s construction had not progressed much beyond its very large packed earth foundation (Sanft 2008). Around the same time the emperor commanded the construction of Mount Li, an artificial hill to mark his tomb. Mount Li, at least, is there, but archaeologists have not yet excavated it. Another aspect of the construction burst of 212 bce was a set of means to conceal the emperor as he travelled in the area of the capital. A combination of walled roads and elevated walkways connected palaces and other buildings, permitting him to go among them without being seen. It was also forbidden to speak of the emperor’s location. According to Sima Qian’s account in the Historian’s Records, the interest in concealment resulted from the advice of a practitioner of mystical arts, who told the emperor that it would help him avoid evil spirits. That in turn was supposed to bring about the arrival of the True Man.The story makes the First Emperor look gullible and ridiculous, as no doubt Sima Qian intended it to do. But there is, of course, nothing intrinsically absurd about a leader like the First Emperor wishing to remain concealed or to keep his whereabouts a secret. The repeated assassination attempts show that, if common sense would leave any question. This same practitioner and another together bear putative responsibility for an event in 212 bce that, together with the assertion of control over classic texts, forms in large part the Qin reputation for brutal anti-intellectualism. After criticizing the emperor as someone whose desire for power left him unable to achieve the longevity they had on offer, the two men departed the court suddenly. The emperor had treated the two generously, and their criticism angered him. He deputed officials to go question remaining scholars about disparaging the emperor, and all of the scholars informed on each other. The emperor is recorded to have executed 460 of them by burying them alive and to have banished still more. The burial found its place in history alongside the alleged destruction of texts, and it became a truism that the Qin ‘burned books and buried scholars’. The only person who comes out of the story looking good is the eldest 154

The Qin dynasty

son of the emperor, Fusu, who remonstrated with his father and was in consequence sent off to oversee the army in the northern border region. In 211 bce there were disturbances among the stars and a meteorite that fell to earth in the area of Dong Commandery. Someone – supposedly a member of the common population – carved it with a prediction of the First Emperor’s death and the disintegration of his empire. When news of this reached the emperor, he sent officials to find the culprits. The Historian’s Records tells us that no one confessed, so they executed everyone living in the area and destroyed the stone. In the final year of the First Emperor’s life, 210 bce, he set out once again to travel through his empire, heading east and south. The two chancellors – left and right – accompanied him, as did Huhai, his favourite son.The entourage visited the Jiuyi mountains (Hunan) and there made a sacrifice to mythical sage kings. Then they travelled in boats along the Yangtze and eventually ascended Mount Kuaiji (Zhejiang) to make an offering to another mythical ancient ruler and a sacrifice to the Southern Sea. They placed an inscribed stone praising the emperor at Kuaiji, then began their long and circuitous route back to the capital. While they were still following the seacoast, the emperor became sick. Sima Qian records that the First Emperor had an aversion to speaking of his death, so that none among the courtiers dared to speak of the succession as his illness progressed. Of course, the emperor had in fact spoken of this in the abstract, at least, when he chose to call himself the First Emperor and imagined generation after generation of successors. This suggests his reasons for not naming a successor were more complicated than a simple unwillingness to speak of the inevitable.The intrigue that surrounded his death indicates that concern about what would happen after he had named a successor, for example, would have been well founded. Whatever the reason for his previous delay, as the emperor grew sicker he commanded an official named Zhao Gao to create a document bearing the emperor’s seal for Fusu, still with the army in the north. In it, the First Emperor ordered Fusu to bury his corpse at the capital, Xianyang, and to entrust the army to Meng Tian, the general. Either by implication or some piece of the text now missing from the record, this document designated Fusu the emperor’s successor. But instead of sending the document north to Fusu, Zhao Gao retained it. The First Emperor died of his illness in the summer of 210 bce.4 He had ruled a total of thirty-six years as king and as emperor. Since the entourage was still far from the capital, Li Si feared that immediately announcing the emperor’s death would lead to disorder. He decided to keep it a secret, and working together with Huhai, Zhao Gao, and a number of favoured eunuchs, concocted a ruse. They placed the coffin in a wagon, pretending the emperor reclined there. When officials presented memorials, one of the eunuchs in on the scheme would approve them in his place. In the version of the events that has come down to us, Zhao Gao is the villain among villains. For he comes up with the idea of installing Huhai as emperor in place of Fusu and proposed the conspiracy. Huhai was initially resigned to the throne going to his brother and resisted Zhao Gao’s plan as a violation of duty and filiality, and because of his own insufficient ability. Zhao Gao persuaded him by appealing in turn to self-interest, to classical precedent, and to vainglory. Once he had convinced Huhai, Zhao Gao turned to Li Si, who at first rejected the proposal even more strenuously than Huhai had.To convince Li Si, Zhao Gao employed his considerable eloquence in laying out the degradation and danger that Li Si could expect if Fusu ascended the throne. In place of the document the First Emperor had dictated to Zhao Gao, which ought to have made Fusu the next emperor, the conspirators supposedly falsified a letter that accused Fusu of failure and waste and criticizing the throne. It also blamed Meng Tian, general of the border 155

Charles Sanft

region and member of a lineage that had helped Qin rulers for at least three generations, of failing to correct these errors. Both were to commit suicide. Meng Tian tried to convince Fusu that the letter could be fake, but Fusu refused even to consider the possibility of resisting his father’s command and killed himself. Meng Tian did not immediately accept the order and went to prison, where he eventually committed suicide by poison. Huhai’s succession was thus secured. They buried the First Emperor in the ninth month of the year at Mount Li, in a tomb said to have been filled with treasures and wonders and guarded by traps. Some of the Second Emperor’s first acts upon coming to his position were in direct emulation of his father. Aware of his youth and troubled by the lack of loyalty to him among the populace, he thought of his father’s travels and the demonstration of strength they had constituted. He travelled east with Li Si in attendance, visiting many of the same places the First Emperor had and adding to the stone inscriptions an addendum proclaiming his filial devotion. A few months later he would restart work on Epang Palace, though never making much progress. The high officials and imperial relatives presented a difficult problem for the young ruler, as they would not accept his authority. It seems likely that they knew of Huhai’s insufficient abilities and probably even had some idea about the contravention of the First Emperor’s wishes in the matter of succession. But when the Second Emperor questioned Zhao Gao about the situation, Zhao pointed instead to the fact that many of the top officials came from families that had been illustrious for generations and looked down upon him. He argued that the moment called for violence. What followed was a purge of high officials and imperial clan members through new laws and new punishments designed to achieve just that goal. The massive loss of expertise, experience, and prestige surely weakened the bureaucracy. In all likelihood, this is what led directly to the brevity of Qin imperial rule. Given these sudden and acute changes in the upper echelons of administration, it is not surprising that disquiet in the realm began already in the autumn of the Second Emperor’s first year. The story of Qin’s fall begins here. Chen She and Wu Guang, lowly and otherwise unexceptional men, were leading some nine hundred troops in the area that is now Hubei, on their way to a new station. Heavy rains delayed them en route, and Chen She and Wu Guang reckoned they had missed the appointed time for their arrival. According to received account, this delay was a capital crime. Whatever the proximate cause was, Wu Guang and Chen She decided to rebel. After Wu Guang had killed a superior – for reasons variously given – the two called together the troops traveling with them.They reminded their comrades of the punishment that, in the traditional narrative, awaited the entire force – and strengthened their point with the dismal odds for making it through their military service alive even if they escaped execution. This persuaded the troops, and together they began their rebellion under the name of the former state of Chu. Chen She declared himself general, with Wu Guang under him. Their small force attacked and captured a series of fortified towns, gathering soldiers and making its way north. They ended up in the area of Chen (in Hunan and Anhui), where Chen She became king. Word of the rebellion spread through the empire, and other men followed their example, even as the forces of the new Chu expanded their numbers and reach. Wu Guang became ‘temporary king’ without a kingdom and pushed west. The Second Emperor, for his part, is said to have punished messengers who reported on the rebellions, with the result that he later heard only that difficulties were minor. Chen She continued to expand his territory and army. He sent forces north to former territories of Zhao and Wei and elsewhere. He also dispatched westward as general a man he selected on the basis of reputation alone, Zhou Wen, in what would turn out to be an unsuccessful attack on the Qin. The forces in Zhao initially took more than ten fortified towns, killing the Qin officials in charge of them. When the general leading the campaign promised to spare the lives of officials 156

The Qin dynasty

manning other towns in Zhao as well as in former Yan, more than thirty surrendered without resistance. At that point, the general declared himself king of Zhao.The news enraged Chen She, but he feigned acceptance. The Zhao king’s advisors knew that Chen’s acquiescence could only be part of a long-term strategy. But all those involved played along. Around this same time, other significant local rebellions against the Qin emerged. These included that of Liu Bang, eventual Han founder, and Xiang Liang, whose nephew Xiang Yu would end up being Liu Bang’s main rival. A member of the former Qi royal lineage,Tian Dan, declared himself king of Qi; a one-time Qin official named Han Guang did the same in Yan at the behest of elites there; and with Chen She’s assistance the former nobleman Wei Jiu claimed the kingship of Wei. The Second Emperor’s first year ended with the empire lurching toward dissolution. Liu Bang began the following year with a series of victories.Yet the Qin government was still powerful. Qin forces defeated Zhou Wen’s army, and while Wu Guang had encircled a fortified Qin town, he was unable to break its defences. The other generals involved in the siege grew fearful that Qin reinforcements would arrive and turn the stalemate into a victory for Qin.They complained of Wu Guang’s arrogance and thought him unworthy of joining their plans – in other words, he failed to accede to their wishes. The generals faked a command, executed Wu Guang, and sent his head to Chen She. All this was to no avail, and the Qin arrived to defeat their forces, killing the general in charge. Still another general, unsuccessful against the Qin, was executed by Chen She. Nor was that general alone. For around the same time, Chen She was busy executing, offending, and alienating all his associates, friends, officials, and commanders. In the Historian’s Records account of Chen She’s life, Sima Qian blames this series of actions, which left Chen without supporters, for his fall. Whatever the path by which he reached it, Chen She came to a quick and ignominious end when his driver, wanting to surrender to the imperial government, killed him. Chen She had been king for some six months. The driver did not live to enjoy the fruits of his perfidy. In the capital, the Second Emperor was becoming increasingly isolated from the bureaucracy that ran the empire. Here, again, the various accounts do not match up in their details, though the general shape is clear. It seems this isolation resulted from the influence and advice of two men. One of these two men was Li Si, who faced his ruler’s unreasonable expectations and a realm that was falling apart and consequently feared loss of his position. Li Si advised the Second Emperor to use surveillance and punishment of officials combined with sole discretionary authority, which he described as the methods of the sage ruler. The other man involved was Zhao Gao. According to one account, Zhao Gao had, from selfinterested motives, advised the emperor to cease holding audiences with his officials.When Li Si subsequently sought an opportunity to address the emperor about the realm’s exigencies, Zhao Gao pretended to help him. But in fact he repeatedly sent Li Si to Huhai while the emperor was taking his ease, angering him. Zhao Gao then calumniated Li Si, alleging he sought a kingship as reward for his earlier support and casting doubts upon his family’s loyalty. Li Si’s response was to criticize Zhao Gao in a letter. The emperor shared the content of the letter with Zhao Gao, and at the latter’s suggestion imprisoned Li Si for trial. According to another account, Li Si and two other high officials remonstrated with the Second Emperor about his failure to deal effectively with the rebellion. The emperor grew angry and commanded all three be tried. Only Li Si, convinced that his years of service and loyal intentions would give him a chance, chose not to commit suicide. He wrote a letter reminding the emperor of his many services to the Qin, but it never reached its intended recipient, and he was executed in 208 bce in the market at Xianyang. The Second Emperor made Zhao Gao chancellor. 157

Charles Sanft

At the same time this was happening in the capital, the situation of the Qin Empire was worsening. Heavy rains began in the autumn, and Xiang Liang was winning victories on the battlefield, repeatedly defeating Qin troops. The Second Emperor finally sent a reinforced body of men, which defeated Xiang Liang’s army and killed him at Dingtao (Shandong). Xiang Liang had dispatched Xiang Yu and Liu Bang to capture Qin towns in a successful joint campaign. When news of Xiang Liang’s death reached them, they turned back east and separately encamped their armies.The following months saw increasing disorder that lasted until the death of the Second Emperor the following year. The armies of Chu, especially those of Liu Bang and Xiang Yu, won a series of victories. At the same time, forces associated with other states clashed with the Qin and with each other. The Qin military became increasingly unimportant as Xiang Yu and Liu Bang conquered town after town. Soon the conflict resolved into a rivalry between Xiang Yu and Liu Bang, in which the Qin military, its defeats and disloyalties, played what seems only a supporting role. Zhao Gao engineered the death of the Second Emperor, just as he had the emperor’s accession to power. The narrative tells of a violent dream that troubled the emperor, who decided to fast and make a sacrifice to avert the misfortune he feared. While preparing himself, Huhai also sent a message blaming Zhao Gao for the rebellion that threatened his rule. This frightened Zhao Gao, who entered into a conspiracy with his son-in-law Yan Le and his younger brother Zhao Cheng. Zhao Cheng and Yan Le faked the arrival of rebellious troops at the palace where the emperor was purifying himself, then forced their way in to where the emperor was. When Yan Le confronted the emperor and refused his request to see Zhao Gao, Huhai made a series of decreasing requests in exchange for submission. As Yan Le finally approached to kill him, Huhai committed suicide. They buried him as a commoner, away from the capital. Zhao Gao was at this point resigned to the breakup of the realm and sought to preserve himself as best he could. He called together the high officials and members of the imperial clan and told them what he had done. Zhao Gao then acknowledged that the realm had reverted to a situation in which Qin was but one kingdom among many, with a territory that was growing smaller. As such, he proposed that their next ruler should hold the title of king. The man they chose as king of Qin was Ziying, whom accounts variously describe as the nephew of the Second Emperor or the younger brother of the First (Ma 1982, 101). Ziying had a reputation for humaneness and frugality that Zhao Gao thought would make him acceptable to the common population of Qin. Ziying himself, for obvious reasons, distrusted Zhao Gao. After five days of ritual purification, Ziying was supposed to go to the Qin ancestral temple to receive the jade seal of rule. He did not go to the temple at the appointed time and ignored repeated summonses. When Zhao Gao came to fetch him, Ziying had him killed and his family exterminated to three generations. After Ziying had ruled as king of Qin for forty-six days, Liu Bang entered the Qin homeland through a mountain pass and defeated its army. The officials refused to resist, and Liu Bang sent an emissary to make an agreement with Ziying in exchange for his surrender. Ziying, with a rope around his neck and riding an undecorated chariot, met Liu Bang outside the capital.There he presented the seal of rule to the future Han emperor. Liu Bang spared Ziying, but after more than a month, Xiang Yu killed him and the rest of the former ruling clan, ending the Qin line. The historiography of the Qin dynasty derives from transmitted texts and palaeographic material recovered by archaeologists. The transmitted record bases itself on Han sources. The Han dynasty, for all its rhetoric to the contrary, and after a short hiatus at its start, took over and developed most Qin governmental systems. This presented an obvious legitimation problem: loyal subjects should have restored the previous dynasty, not founded a new one. Han writers and government officials responded by blaming the Qin for every sort of failure, even as 158

The Qin dynasty

the Han in fact built upon the many Qin successes. Historians have too often repeated these criticisms. The continuities between Qin and Han are such that it would be no misrepresentation to characterize the changeover mainly as a change in ruling houses rather than a system or method of rule. So while scholars once linked the Qin dynasty with the Warring States, historians now more frequently conceive of a Qin-Han period marking the beginning of the Chinese empire. It is that imperial system that, continually developing and repeatedly interrupted, ended – or at least paused – with the fall of the Qing dynasty in 1911.

Notes 1 The historical narrative in this chapter derives mainly from Sima Guang (1956; 1–298 and his preface); and Sima Qian (1959). I draw extensively from Ma Feibai (1982), who gathers together and organizes the received sources on the Qin. I have also referred to Jian (1999), Lao (1980), and Yang (1980). I synthesize many sources and argue points of interpretation in Sanft (2014a), which I draw from throughout this chapter. I have inserted specific notes when drawing from other scholars’ work, from Ma Feibai’s analyses, and my work outside my book. 2 Although received accounts say there were thirty-six commanderies at this time, modern research and evidence from excavated sources has complicated the picture greatly by proving that previously unknown names existed; see, e.g., Zhou Zhenhe (2005). 3 The name Epang is variously transliterated. The characters that make it up look like they should be pronounced Afang. 4 The date given in Sima Qian (1959: 264) is the bingyin 丙寅 day of the seventh month that year. According to the calendar conversion in Wang Shuanghuai (2006: 1192), however, there was no bingyin day that month, although one did come in the preceding month. Such discrepancies are not uncommon. There is no way to resolve this conflict based on the information available to us, but no doubt exists about the general season.

Works cited Di Cosmo, Nicola. (2002) Ancient China and Its Enemies, Cambridge: Cambridge University Press. Jian, Bozan. (1999) Qin-Han shi, second edition, Beijing: Beijing daxue chubanshe. Lao, Gan. (1980) Qin Han shi, Taipei: Wenhua daxue chubanbu. Ma, Feibai. (1982) Qin jishi, Beijing: Zhonghua shuju. Petersen, Jens Østergård. (1995) ‘Which Books Did the First Emperor of Ch’in Burn? On the Meaning of Pai Chia in Early Chinese Sources’, Monumenta Serica, 43: 1–52. Sanft, Charles. (2008) ‘The Construction and Deconstruction of Epanggong: Notes from the Crossroads of History and Poetry’, Oriens Extremus 47: 160–76. Sanft, Charles. (2014a) Communication and Cooperation in Early Imperial China: Publicizing the Qin Dynasty, Albany: State University of New York Press. Sanft, Charles. (2014b) ‘Shang Yang Was a Cooperator’, Philosophy East and West, 64.1: 174–91. Sanft, Charles. (2015) ‘Population Records from Liye: Ideology in Practice’, in Ideology of Power and Power of Ideology in Early China, ed.Yuri Pines, Paul R. Goldin, and Martin Kern, 249–69, Leiden: Brill. Sima, Guang. (1956) Zizhi tongjian, Beijing: Zhonghua shuju. Sima, Qian. (1959) Shiji, Beijing: Zhonghua shuju. Wang, Shuanghuai. (2006) Zhonghua rili tongdian, Jilin: Jilin wenshi chubanshe. Yang, Kuan. (1980) Zhanguo shi, Shanghai: Shanghai renmin chubanshe. Zhao, Huacheng. (2014) ‘New Explorations of Early Qin Culture’, transl. Andrew H. Miller, in Birth of an Empire:The State of Qin Revisited, ed.Yuri Pines et al., 53–70, Berkeley: University of California Press. Zhou, Zhenhe. (2005) ‘Qindai Dongting, Cangwu liangjun xuanxiang’, Fudan xuebao (shehui kexue ban), 5: 63–67.

159

7 THE FORMER HAN EMPIRE

VINCENT S. LEUNGTHE FORMER HAN EMPIRE

Vincent S. Leung

The Former Han (206 bce – 9 ce) was in many ways an improbable empire.The humble origin of the founding emperor and the improvisational nature of its early administration, as we shall discuss in this chapter, hardly augured well for its political viability and sustainability. But surprisingly, the empire did manage to survive and persist for more than two centuries. Moreover, the complex body of ideas and institutions that the empire had cultivated would come to serve as the foundation for the state-building efforts of countless regimes in and around the North China Plain in the next two millennia until the early twentieth century. In fact, this successful afterlife of the Former Han empire has provided a common context for its historiography, in the sense that the historical significance of the Former Han is largely ascribed to the fact that it marks the beginning of the imperial period in Chinese history. Against the swift collapse and apparent failure of its predecessor the Qin empire (221–207 bce), the Former Han is fondly remembered as a rousing political success that inaugurated the long imperial period in Chinese history. The dynastic institutions that it had created and the political culture that they embodied had proved to be so rich and robust that they would be endlessly adapted by various regimes not only in China proper but throughout East Asia in the next two millennia. In this historiographical context, which implicitly equates temporal origin of a political entity with its ideological foundation, the Former Han empire is meaningful and relevant largely for the fact that it constitutes the historical, and therefore ideological, fountainhead of the imperial period of Chinese history. To study the Former Han this way, however, is to relate it to a future that it did not anticipate. We do see it clearly as the historical origin for the many institutions and ideas that would come to define the imperial period, but of course, this perspective is only possible with the benefit of hindsight. Standing in the here and now more than two millennia after the Former Han, we are in a chronologically privileged position to appreciate its great legacy for the rest of the history of China. This retrospective appreciation of the Former Han, a reading of its history with an apprehension of its consequential afterlife, naturally lavishes attention on those moments and episodes of the Former Han that came to have a lasting impact while obscuring much of the debates and failed voices that were also an integral, essential part of its history. In this well-established framework for the history of the Former Han, the improbability of its success is often downplayed as it shoulders the burden of being the originary site of imperial China.1 160

The Former Han empire

In this chapter, I would like to give a survey of the political history of the Former Han from a different perspective. In fact, I would like to approach this from a diametrically opposite perspective, one that purposefully forgets about the great legacy of the Former Han.This is justified by the simple fact that the legacy of the Former Han had nothing to do with the making of the Former Han. Its legacy is true but irrelevant to those who actually lived through it. Those who lived through the Former Han had no idea that their era would become the historical origin of the imperial China in the future.They could not have known that what they had created would come to have such an eventful afterlife in the next two millennia. If anything, from their own vantage point, the Former Han was hardly a grand beginning but more obviously the tail end of a long series of failed political experiments. The end of the Western Zhou (1046–771 bce) half a millennium ago ushered in a long period of interstate warfare, and the one state that managed to put a definitive end to this era of violence and bloodshed, namely the Qin empire, collapsed in just about a decade and a half. This defunct political inheritance, rather than the fantasy of a grand beginning, must have been what weighed most heavily on the minds of those who toiled so arduously for the success of the Former Han. It was through their anxious reflection on this troubling political inheritance and the many institutional innovations that came from it that they ultimately created something that would come to have such a lasting legacy. With this spirit of reorienting our view of the Former Han by way of its past rather than its future, let us now turn to this survey of its very eventful political history. It will be divided into three sections. The first section will discuss the unlikely founding of the empire and the administrative schizophrenia of the first decades of its rule. The second section will focus on the watershed reign of Emperor Wu (Han Wudi 漢武帝, r. 141–87 bce), which saw the consolidation and expansion of the empire.The final section will discuss the negotiation of Emperor Wu’s legacy in the last decades of the empire and the end of the Former Han, with the usurpation of the throne by Wang Mang 王莽 (45 bce–23 ce) in the year 9 ce.

Early Former Han (206–142 bce) Administrative schizophrenia and political ambivalence In the first year of his reign, the First Emperor of Qin (Qin shi huangdi 秦始皇帝, r. 221–210 bce) confidently declared that this new empire that he had founded would last forever. After centuries of failed political experiments by one state after another, the Qin alone had arrived at the one viable form of government, a centralized bureaucratic empire that would liberate all from their own ignorant customs and ushered in a period of perpetual peace. In time, he would be succeeded by the “Second Emperor, then the Third Emperor, continuing forever and ever for ten thousand generations,” while the basic structure of the Qin government will remain essentially unchanged (Sima 1959: 6.236). However, like most regimes in world history that declared themselves to be the end of history, the Qin empire did not last very long. What was supposed to last for ten thousand generations barely lasted for more than a single decade (Pines 2013; Pines 2014; Kern 2000). The empire started to unravel immediately after the First Emperor’s death in 210 bce. A power struggle between various factions broke out at court, most famously between the Chancellor Li Si 李斯 (d. 208 bce) and the eunuch Zhao Gao 趙高 (d. 207 bce), as they vied to exercise influence over the incompetent successor of the First Emperor.2 Old aristocratic kingdoms which were vanquished just a few decades ago by the fearsome Qin infantry army worked quickly to reconstitute themselves upon the news of the emperor’s death (Sima 1959: 6.273). And it was not just the political elite, both at the capital and the various old kingdoms, 161

Vincent S. Leung

that became restless upon the news of the demise of the emperor. The great majority of the population, those belonging to the lower echelons of the Qin bureaucracy as well as the peasantry throughout the country, also saw this as an opportunity to break away from the harsh dictatorial rule of the Qin. The First Emperor, as it turned out, was much more a lynchpin to the Qin empire than he had ever imagined. He piously thought that this world that he had fashioned, animated by “great principles and utmost brilliance,” would remain intact for generations (Sima 1959: 6.243). He could not have been more wrong. He was never just a figurehead presiding over a self-perpetuating bureaucratic machine but the first political domino whose fall precipitated the collapse of the entire empire in just the span of a few years.The Qin empire did not manage to transcend the vicissitudes of history after all. Grievances against the Qin empire were deep and widespread, despite the fact it was founded just a decade or so ago. Resistance did not just come from a single group or a single region but materialized independently among different constituents throughout the country. While the chorus of discontent was unified in its unhappiness with the Qin regime, the motivation and agenda of the different clusters of insurgency were vastly different.Their shared desire to see the Qin topple did not entail a common agenda in what exactly should replace it afterwards. This became especially clear after the Qin empire officially came to an end in the year 207 bce, when the Third Emperor was assassinated by the rebel forces (Sima 1959: 6.275).The remaining rebels, including the eventual founder of the Former Han empire Liu Bang 劉邦 (256–195 bce), were no longer fighting the defunct regime of the Qin but each other over how the country should be governed in the wake of its destruction. This civil war, therefore, was not just a military affair but was also a war of ideas, a fierce contest between different political ideals. In this context, the eventual victory of the Former Han was also the victory of a very specific political vision of how the world should be governed after the fall of the Qin. Let us now turn to this period of civil war and study these competing political visions that emerged among the ruins of the Qin in more details. The beginning of the end of the Qin empire – or conversely, the rise of the Han empire – is usually ascribed to the peasant uprisings that took place throughout the country in the months after the news of the First Emperor’s death. In particular, historians from as early as the Former Han have long ascribed particular significance to the revolts initiated by Chen She 陳涉 and Wu Guang 吳廣 in the year 209 bce. Despite their humble status as no more than conscript laborers of the Qin empire, they quickly inspired a following of thousands of men, and at one point, Chen She even declared himself to be the king (wang 王) of a newly restored Chu 楚 kingdom. We would never know what might have been their endgame, however, for they were both assassinated by their own men just months after the start of the revolt (Sima 1959: 48.1949–1965). Notwithstanding the abortive end of their revolt, it was emblematic of what was happening throughout the country. Rebels arose in quick succession, and many of the old kingdoms of the Eastern Zhou period were hurriedly reconstituted in short order, including Qi and Yan in the northeast and Hann, Wei, and Zhao in the north, in addition to the aforementioned Chu in the south (Sima 1959: 6.273). The rebels behind these newly restored states might have had different agendas, but at the outset they were all united in their defiance against the Qin empire. The Qin was able to quell much of these revolts in the beginning, but eventually it began to falter. Among the different rebel forces in the country, the ones behind the newly restored Chu kingdom quickly emerged as the most consequential of them all. They would be the ones who eventually sacked the Qin capital Xianyang 咸陽, assassinated the Third Emperor, and put a definitive end to the Qin empire in the year 207 bce. Despite the success of the leadership of the Chu rebels, they actually did not share a unified front. By the time that the Qin empire was laid to waste, they were 162

The Former Han empire

divided into two main contingents. One was led by Xiang Yu 項羽 (232–202 bce), a descendant of a noble military family of the Chu kingdom. The other was led by the aforementioned Liu Bang, the eventual founder of the Former Han. In contrast to the aristocratic pedigree of Xiang Yu, Liu Bang hailed from a very modest background. Under the Qin empire, he was merely the head of a postal station in his home commandery, Sishui 泗水. Both men began their rebel career with an act of assassination; Xiang Yu murdered the governor of Kuaiji 會稽 and Liu Bang killed the magistrate of his hometown, Pei 沛. Separately, they each assembled military forces that ultimately overwhelmed the Qin army. After they demolished their common enemy, the Qin empire, they did not join hands to celebrate, but instead, they turned against one another (Sima 1959: 7.295–317; Sima 1959: 8.341–379). The dramatic battles that ensued between Liu Bang and Xiang Yu have long been part of the legendary lore of the founding story of the Former Han empire. But what were they fighting over? What disagreement prompted them to prolong the bloodshed even after the Qin had fallen? It was more than a simple matter of military predominance or future political leadership of the Central Plain. As it turns out, Xiang Yu and Liu Bang had two very different visions for the sort of polity that should succeed the Qin. The former preferred a reversion to a decentralized political arrangement, while the latter was committed to recreating a more centralized state pioneered by the Qin. In the early months of 206 bce, after the assassination of the Third Emperor of Qin, Xiang Yu declared himself to be the “Overlord of the Western Chu” (“Xi-Chu bawang” 西楚霸王). Based at Pengcheng 彭城 (in modern-day Xuzhou, Jiangsu Province), he would provide leadership for the rest of the country, divided into eighteen kingdoms. Liu Bang was named the king of Han 漢, one of these eighteen kingdoms (in present-day Sichuan and southern Shaanxi Provinces) (Sima 1959: 7.317; Sima 1959: 8.365). In many ways, including this particular titular nomenclature (e.g. “bawang” 霸王 and “wang” 王), this confederacy of kingdoms was reminiscent of the decentralized multistate system of the Spring and Autumn period (771–476 bce).We would never know how this confederacy might have actually worked, because it never had a chance to fully materialize. Few of the rebels, including Liu Bang, had the intention to submit to this confederacy proposed by Xiang Yu. Should Liu Bang have accepted the position granted to him by Xiang Yu, that might very well have led to the end of the civil war. Xiang Yu would have been the new overlord of the Central Plain, and there would never have been a Former Han empire. But Liu Bang was defiant from the beginning. He never entertained the legitimacy of this confederacy. To the contrary, Xiang Yu’s presumptive claim to the overlordship had prompted him to escalate his military operations even more. In the histories written under the Former Han, compiled from the perspective of the eventual victor, they would consider the year 206 to be the first year of Liu Bang’s reign, but in truth, he did not yet have control over the entire country.3 In the next few years, Xiang Yu persisted in quelling opposition to his claim of overlordship, while Liu Bang continued to expand his military campaign. By the year 203 bce, after a series of significant setbacks, Xiang Yu was obviously no longer committed to his original vision of a confederacy. Instead, he proposed to Liu Bang that as the two leading military overlords, they should cease fire and coexist peacefully, with each claiming half of the country as his own. Specifically, there would be the Han kingdom in the western half of the country, while the rest in the east would fall under his sovereignty. Liu Bang accepted the title but only nominally, while he continued to pursue Xiang Yu and his followers. The next year, 202 bce, after a fateful battle that ended at the Huai River, Xiang Yu committed suicide by slitting his own throat.The death of Xiang Yu is the event that, unofficially and effectively, marks the birth of the Han empire. Later historiographies may say that the Han empire started in the year 206 bce, immediately after the last Qin emperor was assassinated, but it was really this moment, when blood gushed out of Xiang Yu’s throat, that Liu 163

Vincent S. Leung

Bang was in a position to proclaim absolute sovereignty over the entire country and proceeded to the work of political reconstruction among the physical, institutional, and ideological ruins left behind by the fall of the Qin empire and the subsequent civil war (Sima 1959: 7.315–340; Sima 1959: 8.365–379). What might have been if Xiang Yu had prevailed over Liu Bang? We will never know, of course, but given the two different proposals that he had put forth, it is safe to assume that he would not have been keen to recreate the type of centralized legalist bureaucracy that the Qin empire had attempted and failed at just a decade ago. Under either one of his two different proposals, the country would not have been governed by a single bureaucratic administration, like the Qin empire, but would have been divided into a number of kingdoms with strong feudal ties. It seemed eminently sensible not to replicate the Qin-style administration at the time, since it had failed so miserably just a few years ago. In contrast, the eventual victor, Liu Bang, the founding emperor of the Former Han empire, did not seem quite as eager to toss out this idea of a legalist administration. While this was not explicitly articulated during the civil war with Xiang Yu, it will be borne out by what he actually created in the early years of the Han. His penchant for the idea of a centralized state, the type that was pioneered by Shang Yang 商鞅 in the fourth century bce and perfected by the First Emperor of the Qin, was intimated early on by reports of his deference to the cult of Yellow Emperor (huangdi 黃帝) which had, by this time, become a familiar deity associated with centralized statecraft and sanctioned violence (Sima 1959: 8.350; Lewis 1990; Puett 2001). His defiance against Xiang Yu’s proposals, not once but twice, was less evidence of his distaste for Xiang Yu than his disagreement with the idea of a decentralized country. The protracted battle between Liu Bang and Xiang Yu was not only a competition for military supremacy but was also a war of political ideals. What Xiang Yu had proposed was effectively an erasure of the Qin legacy, while Liu Bang was keen to give its political ideal a second hearing. Liu Bang prevailed, and so the Qin legacy lived. Liu Bang’s partiality to the imperial artifice of the Qin was everywhere evident in his first sovereign acts. Perhaps most tellingly, he opted to style himself not just a wang (“king”), the common sovereign title in past dynasties, but a huangdi 皇帝 (“august god”), the neologistic title that the First Emperor of the Qin fashioned for himself upon the founding of the Qin empire (Sima 1959: 6.236; Sima 1959: 8.378; Puett 2004). It is merely a title, for sure, but it is nevertheless indicative of Liu Bang’s desire to identify himself with the role assumed by the First Emperor. He wished to be the type of ruler that the First Emperor was, whatever that may have meant to him specifically. By adopting the exact same title that the First Emperor had coined for himself, Liu Bang also forged a sense of historical continuity between himself and the First Emperor, and by extension, the Qin and the Former Han empires. A more subtle example of his commitment to the Qin imperial artifice is his promulgation of the new Han code when he first overtook the Qin capital in the year 207 bce. He declared that the legal code under his rule will only consist of three articles: murder is punishable by death, while thefts and personal injuries will be punished according to the severity of the crimes. The rest of the Qin code, according to Liu Bang, will be cast aside in order to undo the empire’s draconian legalistic regime (Sima 1959: 8.362–364).This may be interpreted as a repudiation of the Qin legalist regime, and to a certain extent, it was; after all, it was a declaration of the necessity of pruning the Qin code. But the repudiation was far from total. Liu Bang was calling for only a downscaling of the Qin legal code (“yue fa” 約法) but did not renounce the very idea of a legal code (“fa” 法) itself. It signaled a tacit appreciation for the idea of a legalist regime, the core political principle that defined the Qin empire, and criticized only its excesses (Bodde 1938). It was a measured, calculated inheritance of the Qin legacy without a repudiation of its core.4 164

The Former Han empire

This strategic preservation of disparate pieces of the Qin imperial edifice was paradigmatic of the state-building efforts at the court of Liu Bang. For instance, after his promulgation of the aforementioned “three articles,” he ordered the Chancellor of State (“Chengxiang” 丞相) Xiao He 蕭何 (d. 193 bce) to work through the laws of the Qin and preserve statutes appropriate for the conditions of the new empire to arrive at a legal code, namely the “Nine Sections” (“jiuzhang” 九章) (Ban 1962: 23.1096). Once again, it was a move to rework the legal legacy of the Qin empire rather than a repudiation of it. In some other areas, the appropriation of the Qin legacy was even more forthright. When Liu Bang charged Shusun Tong 叔孫通 (fl. 190s bce), the first Superintendent of Ceremonial (“Fengchenag” 奉常) of the empire, to devise state bureaucratic titles and ritual protocols, he largely borrowed those of the Qin empire (Sima 1959: 23.1159–1160). Similarly, when the time came for the new empire to adopt an official calendar, it simply opted for the Qin calendar wholesale. It was the so-called Zhuanxu 顓頊 calendar adopted on the advice of Zhang Cang 張蒼 (d. 152 bce), a former Qin official who had become a close advisor of Liu Bang (Ban 1962: 21.974). It was also Zhang Cang who proposed, successfully, that the Han empire should follow the Qin in adopting water as its patron element (Sima 1959: 96.2675). Among these scattered first initiatives at constructing the state apparatuses of the Former Han, we can see a basic pragmatic tendency not to undo the Qin legacy but to build on it. The Qin legacy was an object of critical salvage for Liu Bang and his associates. But of course, in politics, there is always a gap between the ideal and the real. Liu Bang’s ideological commitment to the Qin had prompted him to repurpose institutions of the recently fallen empire, but at the same time, he also made many compromises that were necessary at the time for the survival of the new empire. The most significant departure of all, perhaps, was in the political administration of the country. The Qin empire, famously, divided the country into thirty-six commanderies (jun 郡) under the governance of a centralized bureaucracy (Sima 1959: 6.220). It was a deliberate, radical departure from the feudal aristocratic governance of the Western Zhou dynasty, where the country’s territories were enfeoffed to various regional lords in exchange for their pledge of services. Liu Bang, despite his express commitment to so many aspects of the imperial institutions of the Qin, took to recreating the centralized bureaucracy only for a limited part of the country, roughly one-third of its territory in the west. This area was divided into thirteen commanderies in addition to the capital city, Chang’an 長安, and its metropolitan area. The rest of the country, east of these centrally governed commanderies, was divided into ten kingdoms, each enfeoffed to a key ally of Liu Bang who was instrumental to his rise to power. It was a decidedly hybrid political administration, established in the year 202 bce, and it would persist for many decades until the reign of his great-grandson, Emperor Wu, who would take to a radical expansion of the centralized bureaucracy at the expense of these regional kingdoms in the eastern part of the empire (Loewe 2006). The decision to enfeoff the eastern two-thirds of the empire, breaking it up into a number of regional kingdoms, was a politically necessity at the time. They allowed for the young empire to delegate much of the governance of the eastern territory to the various regional lords, while it worked to reconstruct the centralized bureaucracy in the west. Moreover, the enfeoffment also secured, for the time being, the continual allegiance of some of the most powerful supporters of the empire, many of whom had already established strongholds in the eastern territory with expectation of hereditary proprietorship over their home territories. It was an amalgamation of two distinct political systems – centralized bureaucracy and decentralized kingdoms – that gave the Former Han a semblance of political unity in its early decades. But it was a unity that was decidedly schizophrenic. The two systems were never meant to work together. In fact, the Qin devised the idea of a centralized bureaucracy precisely to displace the old aristocratic enfeoffed kingdoms of the Western Zhou.What Liu Bang had created was an 165

Vincent S. Leung

uneasy bricolage of administrative systems, informed by two fundamentally antagonistic governing principles. It was an expedient solution at the time that quickly became a thorny problem for the early Former Han court. As it worked assiduously to rebuild a version of the Qin centralized bureaucracy in the west, it had little faith that the regional kingdoms in the east had any reason to remain subservient to the central court.The supposed loyalty of the regional lords was suspect from the beginning. Liu Bang, in just about a decade, replaced all except one of the regional lords with his sons and brothers. The idea was that kinship ties promised and promoted greater allegiance to the emperor and his court. Meanwhile, by the end of his reign in the year 195 bce, the number of commanderies had also increased from thirteen to sixteen, subdivided into smaller counties that allowed for greater control by the central court (Loewe 1986: 123–27). The management of this volatile hybrid administration, to a large extent, defined the political history of the first decades of the Former Han empire. Liu Bang passed away in the year 195 and was posthumously granted the title Emperor Gaozu 高祖 (“Venerable Ancestor”) (Sima 1959: 8.392). He was succeeded by one of his sons, Liu Ying 劉盈; he was only 15 years old at the time and would only reign for seven years. Effective control of the central court lay not with this teenage emperor but with his mother, the empress dowager Lü 呂 (241–180 bce). Under her regency, which lasted for a decade and a half until her death in the year 180 bce, an inordinate number of her family members came to fill key official positions, and at times, it seemed as if the Lü family was in a position to overthrow the imperial family (Sima 1959: 9.395–412). But the dynasty did survive this near usurpation. In the next four decades, the Former Han would be ruled by two effective leaders in succession, first Emperor Wen (Han Wendi 漢文帝, r. 180–157 bce), a son of Emperor Gaozu, and then his own son Emperor Jing (Han Jingdi 漢景帝, r. 157–141 bce). Neither of them radically remade the hybrid administration first struck by their forebear Liu Bang, but rather, they aggressively managed it in order to preserve the sovereignty of the Former Han. Emperor Gaozu’s installation of his sons and brothers as regional lords, as it turned out, did not ensure the subservience of the kingdoms and peace in the realm after all. Both Emperor Wen and Emperor Jing expended considerable resources on military campaigns and political machination to contain and undercut the power of the regional kingdoms. Take, for example, the tumultuous history of the kingdom of Huainan 淮南. It was one of the ten kingdoms originally created by Emperor Gaozu upon the founding of the empire. He nominated one of his key generals Ying Bu 英布 (d. 196 bc) to be its lord. His lordship was short-lived, however. He and his family were summarily executed less than a decade later in 196 bce for his supposed seditious plot. Then Emperor Gaozu replaced him with his seventh son, Liu Chang 劉長, who was still an infant at the time, as the new lord of Huainan. About two decades later, under the reign of Emperor Wen, Liu Chang launched an unsuccessful revolt against the central court in the year 174 bce, and subsequently the kingdom was abolished and governed as a number of commanderies. A decade later, in the year 164 bce, for reasons that are not entirely clear, Emperor Wen decided to restore the Huainan kingdom, but with only a fraction of its original territory. The rest of the original territory was divided into two smaller kingdoms. All three were enfeoffed to sons of Liu Chang, namely Emperor Wen’s nephews.The Huainan kingdom, in this diminished form, would survive for another few decades until 122 bce, when it was abolished, once again, when its lord staged yet another a revolt. This time, however, it was gone for good and was permanently remade as commanderies (Sima 1959: 118.3075;Vankeerberghen 2001; Major et al 2010: 1–40). Huainan is but one regional kingdoms of the early Former Han, but nonetheless, its history is typical of the others. Regional lords, despite the fact that they were kinsmen of the imperial family, often harbored seditious intent, or at least allegedly so, and staged armed rebellions against the empire. In addition to the necessary military response, the central court often took to 166

The Former Han empire

breaking up a kingdom into smaller fiefdoms to undermine its stronghold in a particular region. In more drastic cases, the central court would abolish a kingdom and convert it into centrally governed commanderies. What was supposed to be a peaceful cooperative existence between the central court and the regional kingdoms quickly turned into a zero-sum game of power and survival. The empire wished to rule over as much of the country as possible with the central bureaucracy, but at the same time, it still had to rely heavily on the regional kingdoms for the governance of the eastern territory far away from the capital. The regional lords, on the other hand, realized quickly that they were in a perpetually precarious position; their titles and all the attendant economic and political privileges were granted to them by a central court that had developed great suspicion of their allegiance. And as the history of the Huainan kingdom had shown, the regional lords were often at the mercy of the much more resourceful central court. On several occasions, they did forge alliances to protest against the empire. Most notably, in the year 154 bce, seven kingdoms joined together to form one of the largest military revolts yet. But in the end, the Former Han empire prevailed; all the regional lords who rebelled were executed and replaced by young sons of the thenreigning Emperor Jing.The central court did not always enjoy victory in this complex matrix of conflicts with the various regional kingdoms, but over the few decades under Emperor Wen and Emperor Jing, the long arc was clearly bending towards the central court. By the end of their reigns around the mid-140s bce, there were as many as forty commanderies and twenty-five regional kingdoms, indicative of a greatly expanded central bureaucracy since the founding of the empire (Loewe 1986: 139–44). In the next few decades, the regional kingdoms would have all but perished under the even more aggressive reign of Emperor Wu, as we shall discuss next. In this very uneasy assemblage of administrative systems, which required constant diplomatic negotiation and military intervention by the central court, we see a great deal of ambivalence among the ruling elite towards the political inheritance from previous regimes. On one hand, there is the bureaucratic legalist system pioneered by the Qin empire. Despite the central court’s evident commitment to it, the central court shrewdly realized that it would be logistically infeasible to extend it throughout the whole country. In order to retain a measure of control over areas that lie beyond the extent of this now diminished central bureaucracy, the only recourse it had was the enfeoffment system of the Western Zhou, whose collapse led to the centuries of chaos for which the Qin legalist empire was supposed to be the solution. In these first decades of the early Former Han empire, the central court decided to pay the high price of this administrative schizophrenia for a semblance of political unity. It forged together two historically failed systems that were not supposed to work together in the first place. This was hardly the articulation of a coherent political vision but a scrambling effort at survival. The early Former Han was less an inauguration of a new order, or even a thoughtful recreation of an old one, but a cobbling together of pieces of what had worked once before for a barely functional whole. This tentativeness of the political condition of the early Former Han, marked by an ambivalence towards the past and an uncertainty about the future, provoked an imaginative, contentious debate over a wide-ranging set of political concerns among the ruling elite (He 1988; Xu 2014; Goldin 2007). In both the received and excavated corpora, we see competing diagnoses of the political predicament of the new empire and divergent reflections on the proper course forward. Among them, the new empire’s commitment to recreating a form of the Qin’s legalist bureaucracy was a particular source of consternation. For those who were sympathetic to the imperial cause, we see their attempts to work out a viable legacy of the Qin experiment. For instance, Lu Jia 陸賈 (d. 170 bce), a close advisor of the founding emperor Liu Bang, argued in his essay “Dao ji” 道基 (“Foundation of the Way”) that the introduction of state violence, as the Qin empire had done, was a necessary stage in our civilizing progress. The next step, therefore, 167

Vincent S. Leung

for the new empire, was not to demolish the Qin institutions but to build on their foundation by moralizing them (Puett 2001: 152–57; Ku 1988). A generation later, there was Jia Yi 賈誼 (d. ca. 169 bce), who argued similarly that the Qin legalist state was not intrinsically immoral, but it was the ruling elite who perverted it (Sima 1959: 6.276–284). In the excavated text “Jing fa” 經法 (“Canonical Standards”) from the Mawangdui 馬王堆 tombs, we see an attempt to recast the idea of law (“fa” 法), the very core of the political philosophy of the Qin empire, as a natural entity rather than an arbitrary creation by men (Peerenboom 1993; Yates 1997). At the same time, we also see elaborate arguments for a more decentralized government in texts such as the Huainanzi 淮南子. Compiled by Liu An 劉安 (179–122 bce), a regional lord of the eponymous kingdom in the last years of the reign of Emperor Jing, this monumental text proffers a cosmology so vast and multitudinous that no single empire should or could ever hope to achieve complete dominion over it. In this text, there is no truly viable legacy from the Qin political experiment. The only lesson is a negative one. It was a foolish attempt to introduce order into an naturally orderly world, and the precipitous demise of the Qin was an entirely expected and a fitting conclusion to the hubris of the Qin rulers (Major et al 2010). These were the anxious voices of a political elite uncertain of the future. They had no foresight that the Former Han will have actually persisted, grandly in fact, in the generations immediately after their time. The only world that they knew was this schizophrenic administration, with persistent unrest among the regional kingdoms, constantly threatening to undo the unity of the empire. What must have further eroded confidence in empire’s strength was the fraught relations that the empire had with its various neighbors beyond the frontier, most importantly the Xiongnu 匈奴 confederation in the steppe region in the north. From the start of the empire, they posed a clear and present danger all along the northern frontier. After a particularly significant defeat in the year 200 bce, the founding emperor Liu Bang wisely aimed for a diplomatic accord rather than a military submission with the Xiongnu. Under this first “harmonious kinship” (“heqin” 和親) treaty, the Former Han agreed to offer various gifts, including Han princesses and various precious goods, in exchange for a voluntary ceasefire and withdrawal to north of the border, as roughly defined by the Great Wall. While it was initially a success, the Xiongnu nevertheless continued to expand their sphere of influence in the steppe region. Terms of the “harmonious kinship” treaties continued to escalate in favor of the Xiongnu in the next decades under the reigns of Emperor Wen and Emperor Jing, while small-scale attacks on territories south of the Great Wall never entirely ceased. Within, the Former Han was beset by a schizophrenic administration, and without, the menace of the steppe nomads was everimminent (Sima 1959: 110.2879–2920; di Cosmo 2004; Chin 2014). But against all odds, the Former Han did survive these tumultuous early decades. Not only did it survive, but in fact, somewhat fantastically, it will become a most expansive empire in the next few decades. This pivotal development came largely under the reign of the next emperor, namely Emperor Wu. Under his charismatic leadership and the support of his many notable supporters, the central court pursued a series of radical policies that would ultimately transform the very nature of the Former Han empire. Let us now turn to a survey of this watershed reign and the fundamental transformation that it had wrought.

Mid-Former Han empire (141–87 bce) A grand beginning under the reign of Emperor Wu In the year 104 bce, Emperor Wu of the Former Han, successor to Emperor Wen, who had been in power for more than three decades by this time, issued a new calendar for the empire and 168

The Former Han empire

along with it an imposing new reign title, the “Grand Beginning” (“taichu” 太初).The very idea of using reign titles (nianhao 年號) to enumerate years was an invention under Emperor Wu’s reign. It started in the year 113 bce, when an ancient tripod (ding 鼎) was discovered at Fengyin 豐盈, and to commemorate this auspicious occasion, Emperor Wu declared that this will be year one of the “First Tripod” (“yuan ding” 元鼎) era (Sima 1959: 12.476; Ban 1962: 21.974–976; Loewe 1974: 17–36). Since then, it had become customary for an emperor to choose an expressive “reign title” to designate a particular era. With this bold reign title, the “Grand Beginning,” Emperor Wu was possibly referring to more than just the new calendar but also the brave new world that he had assiduously fashioned over the last few decades. In sharp contrast to the early decades of the Former Han, which were marked by a profound ambivalence between different governing models materializing itself in a strained schizophrenic administration, the fivedecades-long reign of Emperor Wu saw the confident – and by and large successful – pursuit of a distinct, clear political vision for what sort of empire the Former Han should be. An act of hubris, perhaps, to declare that the empire had started anew under his reign, but it was not an entirely unwarranted claim. So what sort of empire did Emperor Wu envision for the Former Han? What new world did he in the end create? One of Emperor Jing’s fourteen sons, Liu Che 劉徹, the future Emperor Wu, was born in the year 156 bce. When he was still a toddler, he was already enmeshed in the politics of the empire’s schizophrenic administration when he was named the regional lord of the Jiaodong 膠東 kingdom in the northeast, whose insurgency was recently suppressed by the central court. He relinquished the title three years later when he was named the heir apparent of the empire. In the year 141 bce, upon the passing of his father, he succeeded to the throne at the tender age of 15 (Ban 1962: 6.155). By this point, after more than four decades of military campaign and political machination, the central court had largely contained the growth and influence of the various regional kingdoms, but nevertheless the empire remained essentially a schizophrenic whole. The young Emperor Wu, however, was obviously not content with the status quo. It is difficult to say with any precision what motivated him and his advisors, but what happened under his reign was clear. Through complex military and political measures, he labored for many decades to greatly expand the central government at the expense of the regional kingdoms. He saw to it that the country would neither devolve into a loose confederation of regional kingdoms nor remain as the volatile administrative hybrid that it had been since its founding. Instead, it would be governed by an expansive legalist bureaucracy presided over by the emperor and the central court at the capital. From the schizophrenic administration that it had been, it would become a strong bureaucratic whole. This, of course, did not happen overnight. It was a protracted transformation that spanned decades, the result of a complex series of initiatives that the emperor and his trusted associates patiently pursued. The rest of this section, devoted to this eventful reign of Emperor Wu, will focus on key aspects of this historical remaking of the Former Han. Let us begin with this most consequential and dramatic transformation of the political landscape of the Former Han, namely the willful destruction of the regional kingdoms by the central court. This was not a new development, by any means, as we have discussed in the previous section; from the beginning of the empire, the loyalty of the regional lords was always suspect, and the central court had always worked to curb the growth of their power. And indeed, by the end of Emperor Jing’s reign, the central court had made great strides towards that end in having secured the prerogative to appoint most of the high officials of all the regional kingdoms. What was new under the young Emperor Wu, however, was the intensity of the campaign against the regional lords and their kingdoms. While his predecessors would have been content with unseating a regional lord and replacing him with one of his own docile sons, for instance, Emperor Wu often did 169

Vincent S. Leung

not bother with such a circuitous strategy to nominally preserve the regional kingdoms and the schizophrenic administration at large. Instead, Emperor Wu would simply annex the territory of the regional kingdoms and convert them into a number of commanderies, entirely governed by officials appointed by the central court. More than ever before, under Emperor Wu’s reign, it was a zero-sum game between the survival of old regional kingdoms and the creation of new commanderies. In the first three decades of Emperor Wu’s reign, no fewer than fourteen kingdoms were annexed by the central court, diplomatically or militarily, and subsequently converted into centrally governed commanderies. Particularly pivotal was the aforementioned dissolution of Huainan kingdom in the year 122 bce; the largest of the remaining kingdoms at that point, its failed revolt and subsequent disintegration into five commanderies in many ways signaled the end of the era when regional kingdoms could represent meaningful opposition to the rulership of the central court. By the year 108 bce, there were eighty-six commanderies and only eighteen kingdoms, with the latter occupying scattered territories mostly in the eastern half of the country, a miniscule fraction of the two-thirds of the empire that they once held collectively about a century before. Parallel to this steady destruction of the regional kingdoms was Emperor Wu’s subtle manipulation of the nobility status, specifically the “hou” 侯 nobility. Often translated as “marquisates,” the hou nobility was not a Former Han invention but an institution of the Qin empire. Highest among the twenty ranks of social status, a hou is a hereditary nobility that often comes with a small fief and various other material privileges.The founding emperor Liu Bang, over the course of his reign, bestowed the hou nobility on 147 individuals, often as rewards for their services or sometime to secure the loyalty of resourceful figures in various parts of the country. While in theory the hou nobility is hereditary in perpetuity, the emperor may also confiscate it from its holder arbitrarily in extraordinary circumstances. Emperor Wu, ever wary of this old elite which enjoyed a great degree of autonomy from the central court, campaigned to purge as many of its holders as possible. Most notably, in the year 112 bce, he confiscated the title of over a hundred marquisates, and afterwards only seven of the original 147 marquisates remained (Loewe 1986: 157–60; Ban 1962: 13.363–390, 14.391–426, 15A.427–483, 15B.483–526). Behind the destruction of the regional kingdoms and the confiscation of the marquisates, the motivation was essentially the same; these old hereditary elites who were seen as an integral part of the old power structure of the empire must be broken down for a proper expansion of the central bureaucratic government throughout the empire. Emperor Wu did indeed confiscate many of the old marquisates, but at the same time, interestingly, he also took to creating many new ones. While the confiscation of existing, inherited marquisates undercut the old nobility, the conferral of new ones created a new elite who would be personally loyal to the emperor. Over the course of his reign, he created seventy-five new marquisates, more than half of which were for individuals from territories that were newly incorporated into the Former Han empire. This relates to another key event in Emperor Wu’s transformative reign, namely its imperialistic expansion of the empire’s territory. Not only did he remake the domestic administrative structure in favor of the central bureaucratic government, he also pursued a very aggressive foreign policy aimed at expanding the empire’s frontier in virtually all directions. By the end of Emperor Wu’s reign, the Former Han empire would proclaim its imperial sovereignty over a vast territory that was significantly larger than its predecessor, the Qin empire. In fact, it was the largest single political entity that the East Asian region had ever seen by this point in its history. In the first two decades of his reign, regarding frontier relations, Emperor Wu was largely focused on diminishing the Xiongnu confederation as a significant military threat in the northwest. While his predecessors relied on the “harmonious kinship” policy to contain the threat of 170

The Former Han empire

the Xiongnu by way of material appeasement, Emperor Wu pursued a decidedly different course. Instead of diplomatic accords with an exchange of gifts aiming for a peaceful co-existence, Emperor Wu opted for hostile engagement with the Xiongnu as a more effective approach in achieving a peaceful northwestern frontier. His argument, in one of the few surviving speeches that we have of him, was simply that after years of appeasement policy, the northern border simply had not yet attained peace (Sima 1959: 30.1422). Between the years 135 and 119 bce, he launched no fewer than seven major military campaigns against the Xiongnu. There were both victories and losses, but on the whole, these were largely effective campaigns that achieved their overt goals not only in defusing the threat of the Xiongnu but also in appropriating new territories for the Former Han. In the year 127 bce, two new commanderies, called Wuyuan 五原 and Shuofang 朔方, were founded in the region beyond the northwestern loop of the Yellow River, representing a significant expansion of the Former Han empire into Central Asia. After a major punitive campaign in the year 119 bce, there would be no more Xiongnu forces infiltrating Former Han territory for almost two decades until the year 103 bce (Sima 1959: 110.2904–2920). Though continuing to be a looming presence just north of the border, largely defined by the physical barrier of the Great Wall, for the rest of Emperor Wu’s reign and the rest of the Former Han empire, it never managed to regroup itself into the grand nomadic empire that it once was in the first decades of the Former Han (di Cosmo 2004: 161–254). This westward expansion into Central Asia was not solely a military operation for keeping the Xiongnu at bay, though it did accomplish that, but it was also informed by the evolving economic interest of the empire in establishing trade routes with various groups in Central Asia. The celebrated journey of Zhang Qian 張騫 (d. ca. 113 bce) from the capital city Chang’an to various Central Asian countries contributed significantly to this development. A military official (lang 郎), Zhang Qian volunteered to lead a diplomatic mission to Yuezhi 月氏to strike an alliance with them against their common foe, the Xiongnu, who were physically located between the two. Unbeknownst to him at the time, it would be almost twenty-five years before he would return to capital city of the Former Han. He did eventually make his way to Yuezhi after a decade of captivity by the Xiongnu, and even though he was unsuccessful in securing the alliance with the Yuezhi, he had nevertheless gathered important information about many regions in Central Asia, as far as Bactria (modern-day northern Afghanistan). Upon his return to the Former Han capital in 126 bce, his report was by far the most detailed and richest account of the various Central Asian countries that the Former Han had ever secured. His observation that precious goods of the empire, in particular silk, were coveted in these faraway westward countries fortified Emperor Wu’s conviction to extend the reaches of the empire westward. Under his directive, there would be additional expeditions into Central Asia in the next couple decades; he also established new commanderies further westward, with Dunhuang 敦煌 as a major outpost, a key portal to the various Central Asian regions (Ban 1962: 61.2687–2698; Yü 1967: 135–137; Hulsewé 1979). Emperor Wu’s colonial ambition, not surprisingly, was not limited to the northwest frontier and Central Asia. Soon after his successful campaigns against the Xiongnu, a major preoccupation in the first two decades of his reign, he turned his attention to the rest of the complex frontier of the empire. He sent forth imperial troops, almost yearly for more than a decade, to colonize ever more territories, or when that failed, to establish diplomatic or tributary relations with its neighboring countries. In the northeast, it went as far as the northern Korean peninsula, and at one point, it had established there as many as four commanderies. In the southwest, it managed to push into Yunnan 雲南, adjacent to modern-day northeastern Myanmar, and in the south, it ventured as far as the Hainan island and what is the present-day country of Vietnam. By the year 104 bce, Emperor Wu had added twenty new commanderies to the Former Han 171

Vincent S. Leung

as a result of this colonial enterprise in virtually all directions along the empire’s frontier. It was by far the largest empire, geographically speaking, that the world had ever seen (Sima 1959: 30.1417–1444). It was in this same year, namely 104 bce, that Emperor Wu declared the adoption of a new calendar and the start of a new era the “Grand Beginning,” as we discussed earlier. Argumentative and hyperbolic though it may have been, this new reign title was not entirely unwarranted. In the few decades since he assumed the throne, the Former Han had been remade in fundamental ways. The old nobility and their regional kingdoms had largely fallen and reintegrated into the bureaucratic order of the empire, which had expanded well into Central Asia and the far reaches of East Asia. This monumental remaking of an empire, of course, required not only a distinct political vision but also great material resources. The many military campaigns, in particular, were tremendously expensive. How did Emperor Wu finance this great expansion of the empire? The confiscation of old nobility titles, the annexation of regional kingdoms, and the creation of new commanderies from territories both old and new added to the wealth of the empire immediately by increasing the landholding and taxable households of the empire. But that was not nearly enough. Emperor Wu, in fact, pursued a very ambitious fiscal agenda that would ultimately transformed the very nature of the Former Han empire. In the economic realm, the world of commodities and their exchanges, Emperor Wu once again envisioned a very activist role for the central government. This is, of course, consistent with his expansion of the central bureaucracy at the expense of the regional kingdoms which had enjoyed certain independence in the management of their local economies.Towards the end of the 120s bce, when most of the major regional kingdoms had fallen and the state treasury was heavily exhausted by the long campaign against the Xiongnu, Emperor Wu began to explore the possibility of establishing state monopolies of key industries such as salt, iron, and liquor. Oversight of salt and iron mines began to grow steadily, and then, in the year 119 bce, he declared that the Former Han would impose a state monopoly on the production and sales of salt and iron. Salt, of course, was essential to the diet of the entire population, and iron tools were indispensable to the millions of farmers whose livelihood depended on them. State monopolies would ensure affordable access to these two essential goods for every household within the country, and at the same time the empire stood to profit greatly from their sales. For the same economic rationale, the empire imposed a state monopoly on the liquor industry about two decades later in 98 bce, when it was once again pressed to cultivate new sources of revenue after another two decades of imperial expansion. The empire never attempted to monopolize the production and sales of grain, another key commodity in the Former Han economy. It did, however, implement the so-called balanced standard (“pingzhun” 平準) system to control price levels of grains; the government participated as a direct buyer and seller of grains in different regional markets in order to control and stabilize prices throughout the country. Ordinary households would be protected from inflated grain prices or supply shortage, while the central government stood to profit greatly from being the largest grain trader (Wagner 2002; Swann 1950; Sima 1959: 3.1441).5 At the same time that the empire was inserting itself into the commercial sectors, it also began to pursue monetary measures to further its grip on the economy. In the same year that the salt and iron monopolies were imposed, the central court started to circulate a new type of copper coins, called “wuzhu” 五銖, to be a legal tender for the entire empire. Six years later, the state proceeded to ban private coinage of any kind. Private coinage was fairly common up until this time, especially among the regional kingdoms, which often found it necessary to mint their coins to sustain their local economies. With this abolition of private coinage, the empire was now the sole supplier of the very medium of economic exchanges.The Former Han empire was 172

The Former Han empire

now the political entity that facilitated and sanctioned all commodity exchanges, and as such, it had also assumed the monetary mechanism to greatly influence commodity prices throughout the country (Sima 1959: 30.1429; Ban 1962: 24.1165–1185). All in all, these economic policies that Emperor Wu pursued were all heavily interventionist, with a strong preference for state participation in virtually all sectors of the economy. Similar to its governing philosophy that the world is only well-ordered when it is integrated into the bureaucratic order of the central government, the economy was also thought to be disorderly on its own until the strong hand of the empire deliberately regulated it. Before we take leave of Emperor Wu and discuss the last decades of the Former Han, let us consider one more event during his reign which, though a modest initiative at the time, would turn out to be immensely consequential in the decades to come. It was the creation of the Imperial Academy (taixue 太學) around 124 bce, and with it the introduction of the Confucian curriculum of the “Five Classics” (“wu jing” 五經). It is fairly easy to see that much of the work that Emperor Wu had done in remaking the Former Han was inspired by its predecessor the Qin empire; the very idea of a centralized bureaucracy engineered for imperialistic conquest had a clear genealogy from the rise of the Qin empire to Emperor Wu’s reign. But yet it would be a mistake to understand the Former Han of Emperor Wu as simply a successful reiteration of the Qin empire with some variations. The case of the creation of the Imperial Academy and the way that Emperor Wu tried to reshape the world of ideas provide a particularly strong case for his political innovativeness. The Qin empire, despite the brevity of its rule, was notorious for its intolerance for ideological dissent, a popularly held historical verdict enshrined in notorious episodes such as the bibliocaust ordered by the chancellor of the Qin empire (Sima 1959: 6.254–255).The Former Han empire under Emperor Wu, despite its strong institutional indebtedness to the Qin, was remarkably open to the articulation of competing ideologies and the very idea of learning at large. Critical voices against initiatives by Emperor Wu were often heard, even though they were rarely heeded by the emperor himself. The Imperial Academy, established to train officials to serve in the empire’s quickly expanding bureaucracy, was structured around the “Five Classics” associated with the editorship and teachings of the historical Confucius. These five corpora, namely the Spring and Autumn Annals (Chunqiu春秋), Documents (Shu 書), Odes (Shi 詩), Rites (Li 禮), and Changes (Yi 易), were precisely the ones targeted for their subversive potential by the bibliocaust of the Qin. The Confucian scholars, namely those who identified themselves as followers of the teachings of Confucius and whose ideological capital was vested in their erudition of these classical texts, found themselves in a curious position. Long subjected to the scorn and derision by the political elite under the rise of the Qin, they now found themselves in the center of the intellectual world of Emperor Wu’s Former Han.With the Five Classics elevated as the standard curriculum at the Imperial Academy, they now enjoyed an ideological currency that they never had before. Enrollments at the Imperial Academy were relatively modest at the beginning, with its graduates comprising only a small fraction of the bureaucracy; in time, however, especially in the decades after the reign of Emperor Wu, the Confucians would steadily grow to be a most significant, powerful faction among the empire’s political elite (Cai 2014; Nylan 2001; Loewe 1986: 464–465). In contrast to the heavy-handed suppression of dissenting voices under the Qin empire, Emperor Wu’s relative willingness to grant audience to ideological diverse voices was striking. To be sure, there are still many episodes where Emperor Wu showed his displeasure at dissent, like when he ordered the grand scribe Sima Qian 司馬遷 (d. 86 bce) to be castrated when he dared to defend a military general defeated by the Xiongnu (Sima 1959: 130.3300). But at the same time, we do see figures such as Bu Shi 卜式 (fl. 110s bce), the Imperial Counsellor (Yushi dafu 御史大夫) in the late 110s, who was openly critical of the many interventionist economic 173

Vincent S. Leung

initiatives pursued by Emperor Wu. In one of the most memorable vitriols from the Former Han, he said that the Control of Grain Commandant (Zhi su wei 治粟尉) Sang Hongyang 桑弘 羊, a key architect of the economic policies under Emperor Wu, should be “boiled alive” (Sima 1959: 30.1442). Bu Shi’s official career did suffer because of his vocal opposition, but it was not unusual for eminent officials to engage in such ardent debates at the Emperor Wu’s court. If one diagnoses the ideological repressiveness of the Qin empire as what had prompted, in part, the eventual revolt against its rule, then the apparent ideological openness of Emperor Wu’s court can be interpreted as a subtle strategy aimed at vesting his court with a certain moral authority and therefore greater political legitimacy. It was a deliberately nimble intellectual environment where oppositions were given a voice, if only to be subverted and suppressed in the end. Emperor Wu reigned until his death in the year 87 bce. It had been a long, eventful reign. Over more than five decades, he and his coterie had fundamentally transformed the Former Han, in the many ways that we have discussed. The schizophrenic administration which he had inherited was now a centralized bureaucratic whole governing a greatly expanded territory with a moral authority that had eluded its predecessor, the Qin empire.The founding emperor of the Former Han, Liu Bang, could have hardly imagined that the rickety empire that he had hurriedly put together could have grown into this colossal empire in just over a century. But where do they go from here? What happened to this imperial project after its lead architect, namely Emperor Wu, passed away? Let us now turn to the last section of this chapter and discuss the last decades of the empire.

Late Former Han (86 bce – 9 ce) A return to antiquity and Wang Mang’s usurpation In 87 bce, after Emperor Wu’s death, he was succeeded by his youngest son, Liu Fuling 劉弗 陵, posthumously known as Emperor Zhao of the Former Han empire (Han Zhaodi 漢昭帝, r. 87–74 bce). There will be six more emperors in the next decades before the Former Han came to an end in the year 9 ce, when the throne was usurped by a certain Wang Mang 王莽 (45 bce–23 ce), a relative of the imperial family, who declared the founding of a new dynasty, the Xin 新 (9–23). In many ways, the political history of this last century of the Former Han can be understood as an extended consideration of the immense legacy of Emperor Wu. Critical voices against Emperor Wu had always been there, even at his own court, but in the decades after his death, they only grew louder. Living in the world that was intimately shaped by Emperor Wu, their diagnosis was that the empire had overextended itself in almost every way and that his expansionist policies were neither sustainable nor desirable over the long term. They argued for a return to the ways of antiquity, variously defined but typically understood to have been best exemplified by the Western Zhou dynasty, over the Qin imperialist ideals embraced by Emperor Wu. This debate over Emperor Wu’s legacy, a contentious assessment of his reign as a way to argue for the empire’s proper way forward, started almost immediately after his death. A major court debate was convened in the year 81 bce, just half a decade after the end of Emperor Wu’s reign, and a record of this contentious conference was compiled into what we know today as the Debate on Salt and Iron (Yan tie lun 鹽鐵論). Compiled some years after the debate, this is a key document on the political culture among the ruling elite in the post-Emperor Wu era. Supporters and critics of Emperor Wu’s imperialistic government argued over a wide range of policy matters, beyond just the state monopolies of the salt and iron industries noted in the title of the text. The critics, often drawing on the supposed authority of the Five Classics and the idea of 174

The Former Han empire

antiquity, argued for a retrenchment of the state in virtually all areas. Excessive state intrusion should be scaled back in order to allow for a greater degree of local autonomy. Their interlocutors, those who supported Emperor Wu’s policies, argued that the empire was not only there to create the condition for a good order but is the good order itself. There is simply no order in the absence of a state. In this way, the reign of Emperor Wu cast a very long shadow over the rest of the Former Han; almost all policy discussions were imagined as a response to his reign. In retrospect, the world that Emperor Wu had created did remain for the most part in the end. Out of this debate in the last century of Former Han rule, what eventually happened was less an unmaking of Emperor Wu’s creation than a retrenchment of it.There was no fundamental repudiation of his legacy but only a recalibration of its governing priorities. Take, for example, the political administration of the late Former Han. Regional kingdoms, and the old nobility, came tumbling down in the half-century under Emperor Wu’s reign. Their destruction cleared the ground for a vast expansion of the centralized bureaucratic government. In the late Former Han, this basic structure of the political administration remained. Old regional kingdoms continued to be demolished, remade into either smaller kingdoms or commanderies. Till the end of the empire, the great majority of the empire’s territory remained under the administration of the centralized bureaucracy. At the same time, there was also a notable surge in the number of marquisates being conferred by the central court. Between the reigns of Emperor Zhao and his successor, Emperor Xuan (Han Xuandi 漢宣帝), from 87 to 49 bce, seventy-four marquisates were conferred to sons of regional lords, with hereditary landholding and various privileges. In the late 60s bce, the central court even made efforts to seek out the descendants of the marquisates who were enfeoffed at the start of the empire but were stripped of their nobility by Emperor Wu in the late 110s. While the central court was committed to maintaining this vastly expanded central government, they also found the idea of hereditary nobility less of an anathema to the integrity of the empire. Toward the end of the empire, in the first census we have dating back to the year 2 bce, the empire had 12,366,470 registered household, or roughly 57.5 million individuals, with a great majority of them under the rule of the central government in the eighty-three commanderies and the rest living in one of the twenty small regional kingdoms (Ban 1962: 28B.1639–1640; Loewe 1986: 157–160). The retrenchment was more dramatic in the area of foreign policies. After the very aggressive imperial expansion under Emperor Wu, the late Former Han largely focused on just maintaining the expanded frontier that it had inherited. Xiongnu, after its protracted battle with the troops of Emperor Wu, was no longer the existential threat that it once was for the Former Han in earlier times. Divided between the rival tribes within its confederation, it lacked the coherent organization to coordinate major military strikes against the Former Han.While the late Former Han largely succeeded in maintaining its presence and control throughout the new colonial territories that Emperor Wu had secured, there were cases where retreat was necessary. On the Korean peninsula, for example, the four commanderies that Emperor Wu established in 108 bce were reduced to just one, the Lelang 樂浪 commandery, in the late 80s bce (Ban 1962: 7.223; Byington 2013). Similarly, the commandery on the Hainan Island in the far south, established in 110 bce by Emperor Wu, was abandoned in the mid-40s bce due to financial strain and the court’s inability to maintain local order (Ban 1962: 70.2835). To the west, the Former Han did maintain its military presence in Central Asia, but the diplomatic alliances with the various groups had considerably weakened. All in all, the war machine that was running at full speed under Emperor Wu had largely decelerated in these last decades of the empire. Emperor Wu’s economic initiatives had also occasioned much debate among the ruling elite in the late Former Han. The aforementioned Debate on Salt and Iron was ostensibly about the state monopolies on these two key industries of the country, as the title suggests. Between the 175

Vincent S. Leung

wide range of positions between laissez-faire and state interventionism held by the political elite, there was the larger question on the proper role of the central government in the economic life of the people of the empire. While Emperor Wu seemed incapable of imagining an orderly, flourishing economy in the absence of state participation, his critics argued for a much more circumscribed role of the empire in favor of greater local autonomy. Some of the most vehement objections were articulated under the reign of Emperor Yuan (Han Yuandi 漢元帝, r. 48–33 bce), who appeared to have been sympathetic to these complaints. One of his eminent counsellors, Gong Yu 貢禹, for examples, argued for dismantling almost all the fiscal institutions that Emperor Wu had created, including the state monopolies of key industries, state coinage, and the “balanced standard” price level control. Appealing to the supposed ways of antiquity, he argued for a return to a pre-monetary economy where wealth was denominated in agricultural goods rather than metallic coins minted by the state. The many deviations and ruptures from the ways of the ancients, first under the Qin empire and then the reign of Emperor Wu, had done nothing but introduced immoral profiteering into the world, according to Gong Yu (Ban 1962: 72.3075). He did manage to persuade Emperor Yuan to abolish the state monopolies on salt and iron in 44 bce, but it was a short-lived victory; the monopolies were restored a few years later, as the loss of state revenue was simply too great. While Gong Yu was alarmed by the state becoming the largest, most resourceful competitor in the economy, to the detriment of its people, there were others who were alarmed by the rapidly rising economic inequality. Private landownership, which had become legal since Emperor Wu, had led to the steady growth of family estates with a large, dependent, landless peasantry. Shi Dan 師丹, an eminent official at the capital towards the end of the empire, once proposed redistributing land equally among the populace, just as sage kings in antiquity had done, in order to alleviate gross economic inequality (Ban 1962: 24.1142–43). These and other radical proposals, voiced at the late Former Han court, were largely unheeded, however. The economic world that Emperor Wu’s policies had wrought largely remained. This rhetoric of a return to the ways of antiquity, prevalent among critics of the legacy of Emperor Wu, was perhaps the most pronounced in the debate over state cults. There were calls, among eminent officials time and again, for the emperors to abandon, once and for all, the sacrificial system and pantheon of deities that the Former Han, specifically Emperor Wu, had inherited from the Qin empire in favor of the venerable cult of Heaven (tian 天) practiced long ago by the Western Zhou dynasty. One of the most eloquent articulation of this call for a cultic reform can be found in the joint memorial by Kuang Heng 匡衡 and Huan Tan 桓譚 for the recently crowned Emperor Cheng (Han Chengdi 漢成帝, r. 33–7 bce). In it, they argued that under the existing state cult, the emperor had to regularly travel to sacred sites to worship the various deities throughout the country, as the First Emperor of the Qin and Emperor Wu had done before; instead, the emperor’s virtue had already secured the blessing of Heaven, which followed the emperor wherever he may be. The cult of Heaven, they argued, obviated the exorbitant, wasteful expenses of these lavish imperial pilgrimages, and in the end, it simply accorded more with the proper ways of antiquity. Emperor Cheng did assent to this proposal and restored the state cult of Heaven (Ban 1962: 25B.1253–54; Puett 2004: 307–316; Tian 2015). Identifying retrenchment of the state as the way to return to antiquity is a trope commonly deployed among critics of the legacy of Emperor Wu in the late Former Han. No doubt, these critics would have agreed with Emperor Wu’s proclamation that the world had begun anew under his reign. For them, however, this “grand beginning” was not the start of a new, better world order but a deviation from the ways of antiquity with the state transgressing its proper boundary. This antiquarian rhetoric was an ideological front that provided support for a wide range of political interests among the ruling elite during the late Former Han. In fact, the person who eventually brought the Former Han empire to its end, namely the aforementioned 176

The Former Han empire

Wang Mang, was one such politician. A nephew of the empress dowager Wang Zhengjun 王 政君 (71 bce-13 ce), he had a long and intimate relationship with the imperial family. After the young Emperor Ping (Han Pingdi 漢平帝, r. 1–6 ce) acceded to the throne, Wang Mang not only became the regent but also managed to enthrone one of his daughters as the empress. When Emperor Ping passed away a few years later, Wang Mang declared himself to be the “acting emperor” (jia huangdi 假皇帝) (Ban 1962: 99.4094). A few years later, in 9 ce, he declared himself to be the new emperor, not of the Former Han, but a new dynasty called the Xin 新 (“New”). And with that, the Former Han came to an end after more than two centuries of rule (Ban 1962: 99.4039–4096). Wang Mang’s Xin dynasty is mostly remembered as a disastrous interregnum. It lasted only for a short time, just about a decade and a half until Wang Mang’s death in 23 ce, and in this short time, he attempted to entirely undo the Former Han empire and replace it with a new society and economy that would have accorded with the ways of antiquity as he understood them based on his understanding of the classical canon including the Five Classics. His reforms, based on the scattered sources that we still have today, were greatly unpopular. The landed elite detested his attempts at land redistribution, while his attempts at currency reform appeared to have quickly eroded economic confidence among the populace at large. Disruptive reforms led to social unrest, especially among the peasantry, and soon revolts broke out throughout the country. Rebels eventually stormed the capital city of Chang’an, and the dynasty came to an end in the year 23 ce when they assassinated Wang Mang. Two years later the Liu family reclaimed the throne and declared a restoration of the Han dynasty. What ensued would be another two centuries of imperial rule by the Liu family, namely the Later Han empire (25–220). All empires end, sooner or later, in one way or another. The story of how the Former Han came to an end through the usurpation by a relative of the imperial family was rather unremarkable in the annals of hereditary monarchy in world history. In retrospect, one truly remarkable thing about the Former Han was that it managed to become a viable empire at all that lasted for as long as it had. Its beginning was humble; the man who founded it, Liu Bang, was but a lowly postal official of a failing empire. Charismatic, perhaps, but admittedly uncouth, he left behind a hurriedly assembled empire with a very volatile administration that was a hybrid of two historically failed models, as we have discussed at length. But yet, the empire survived. It persisted despite the ceaseless rebellions among the regional kingdoms. And then, the empire came under the custodianship of a young teenager, namely Emperor Wu, who quickly grew up to be one of the most vigorous imperial leaders in all of world history. Over more than five decades, he transformed the political house of cards that he inherited into a colossal empire, thoroughly bureaucratic within and imperialist without, with an apparent moral authority that had eluded its predecessor, the Qin empire. The political ambivalence, and its material expression in the schizophrenic administration that defined the early decades of the Former Han was expelled at last. Then, in the wake of Emperor Wu’s pivotal reign, the political elite turned to a fierce debate over the cost of this newfound imperial confidence. The empire did remain entirely viable for another century, and then, somewhat unceremoniously, the throne was usurped by someone who wished for a return to antiquity by undoing all the historical accomplishments of the Former Han. But as he found out in the end, painfully so, that there was no going back. The unlikely empire that was the Former Han had struck a fundamentally new historical course from which there will be no return.

Notes 1 See Qian (2004) and Jian (1983) for examples of this typical narrative of Han history. 2 Translations of all official titles follow Loewe (2000).

177

Vincent S. Leung 3 This is the case in the two major histories written under the Han empire, namely Sima Qian’s Shiji (1959) and Ban Gu’s Hanshu (1962), the two key sources for the Former Han. 4 In recent decades, our understanding of the laws of the early Han empire and their relationship to the Qin legal code have been considerably enriched by the discovery of excavated legal codes. See, in particular, the excellent discussion in Barbieri-Low, et al (2015) of the spectacular find in the Zhangjiashan tomb from the Hubei Province from the first few decades of the Han empire. 5 The translation for “pingzhun” as “balanced standard” here follows Watson (1993).

Works cited Ban, Gu (1962) Hanshu, 12 vols, Beijing: Zhonghua shuju. Barbieri-Low, Anthony J., and Yates, Robin Y.S. (2015) Law, State, and Society in Early Imperial China: A Study with Critical Edition and Translation of the Legal Texts from Zhangjiashan Tomb no. 247, 2 vols., Leiden and Boston: Brill. Byington, Mark E. (ed.) (2013) The Han Commanderies in Early Korean History, Cambridge: Harvard University Press. Cai, Liang (2014) Witchcraft and the Rise of the First Confucian Empire, Albany: State University of New York Press. Chin, Tamara (2014) Savages Exchange: Han Imperialism, Chinese Literary Style, and the Economic Imagination, Cambridge: Harvard University Press. Derk, Bodde (1938) China’s First Unifier: A Study of the Ch’in Dynasty as Seen in the Life of Li Ssû (208?–208 bc), Leiden: Brill. di Cosmo, Nicola (2004) Ancient China and its Enemies: The Rise of Nomadic Power in East Asian History, Cambridge: Cambridge University Press. Goldin, Paul (2007) “Xunzi and Early Han Philosophy”, Harvard Journal of Asiatic Studies, vol. 67, no. 1, pp. 135–166. He, Lingxu (1988) Xi-Han zhengzhi sixiang shi, Taipei: Wunan tushu chuban gongsi. Hulsewé, A.F.P. (1979) China in Central Asia:The Early Stage 125 B.C.-A.D. 23, Leiden: Brill. Jian, Bozan (1983) Qin-Han shi, 2nd edn, Beijing: Beijing daxue chubanshe. Kern, Martin (2000) The Stele Inscriptions of Ch’in Shih-huang:Text and Ritual in Early Chinese Imperial Representation, New Haven, CT: American Oriental Society, 2000. Ku, Mei-kao (trans.) (1988) A Chinese Mirror for Magistrates: The Hsin-yü of Lu Chia, Canberra: Australian National University. Lewis, Mark Edward (1990) Sanctioned Violence in Early China, Albany: State University of New York. Loewe, Michael (1974) Crisis and Conflict in Han China, 104 B.C. to A.D. 9, London: Allen & Unwin. Loewe, Michael (1986) Cambridge History of China, Volume 1, The Ch’in and Han Empires, 221 bc–ad 220, Cambridge: Cambridge University Press. Loewe, Michael (2000) A Biographical Dictionary of the Qin, Former Han and Xin Periods (221 bc–ad 24), Leiden: Brill, 198. Loewe, Michael (2006) The Government of the Qin and Han Empires: 221 bce–220 ce, Indianapolis: Hackett Publishing Company. Major, John, Queen, Sarah, Meyer, Andrew, and Roth, Harold D. (2010) The Huainanzi: A Guide to the Theory and Practice of Government in Early Han China, New York: Columbia University Press. Nylan, Michael (2001) The Five “Confucian” Classics, New Haven, CT:Yale University Press. Peerenboom, R. P. (1993) Law and Morality in Ancient China, Albany: State University of New York Press. Pines,Yuri (2013) “From Historical Evolution to the End of History: Past, Present and Future from Shang Yang to the First Emperor”, in P. Goldin (ed.) Dao Companion to the Philosophy of Han Fei, Berlin: Springer. Pines, Yuri (2014) “The Messianic Emperor: A New Look at Qin’s Place in China’s History”, in Y. Pines, G. Shelach, L. von Falkenhausen, and R. D.Yates (eds.) Birth of an Empire:The State of Qin Revisited, Berkeley, Los Angeles and London: University of California Press. Puett, Michael (2001) The Ambivalence of Creation: Debates Concerning Innovation and Artifice in Early China, Stanford: Stanford University Press. Puett, Michael (2004) To Become a God: Cosmology, Sacrifice, and Self-Divinization in Early China, Cambridge: Harvard University Press. Qian, Mu (2004) Qin-Han shi, Beijing: Sanlian shudian.

178

The Former Han empire Sima Qian (1959) Shiji, 10 vols., Beijing: Zhonghua shuju. Swann, Nancy Lee (1950) Food and Money in Ancient China: The Earliest Economic History of China to A.D. 25 (Hanshu 24), Princeton: Princeton University Press. Tian, Tian (2015) “The Suburban Sacrifice Reforms and the Evolution of the Imperial Sacrifices”, in Nylan, Michael and Vankeerberghen, Griet (eds.) Chang’an 26 bce: An Augustan Age in China, Seattle: University of Washington Press. Vankeerberghen, Griet (2001) Huainanzi and Liu An’s Claim to Moral Authority, Albany: State University of New York Press. Wagner, Donald (2002) The State and the Iron Industry in Han China, Copenhagen: Nordic Institute of Asian Studies. Watson, Burton (trans.) (1993) Records of the Grand Historian: Han Dynasty, Volume 2, New York: Columbia University Press. Xu, Fuguan (2014) Liang-Han sixiang shi, 3 vols., Beijing: Jiuzhou chubanshe. Yates, Robin (1997) Five Lost Classics: Dao, Huanglao, andYinyang in Han China, New York: Ballantine Books. Yü, Ying-shih (1967) Trade and Expansion in Han China: A Study in the Sino-Barbarian Economic Relations, Berkeley: University of California Press.

179

8 THE LATTER HAN EMPIRE AND THE END OF ANTIQUITY WICKY W.K. TSETHE LATTER HAN EMPIRE

Wicky W.K. Tse

An empire in the name of Han In the spring of 190 ce, capitalizing on the chaos which ensued from a palace coup, a general named Dong Zhuo 董卓 (d. 192 ce) led the garrison troops on the northwestern frontier of the Latter Han empire (25–220 ce) to storm the capital city of Luoyang 洛陽. Upon taking control of the imperial court, Dong deposed the reigning young emperor and enthroned another, who would merely be a captive of different warlords and the last monarch of the dynasty. Dong’s actions immediately aroused fierce opposition from regional governors of the area east of the capital. As a response to the threat, Dong forcibly moved all residents of Luoyang, from the emperor to commoners, westwards to Chang’an 長安, an erstwhile imperial capital in his sphere of power. His troops looted and burned Luoyang to the ground during the relocation. What Dong and his men did to the monarchs and the capital was indeed sacrilege against the imperial authority. And it marked a de facto end of the ruling Latter Han dynasty – although its formal and somewhat dramatic end did not come until 220 ce, when the last emperor announced abdication, responding to the transfer of heaven’s mandate from his dynasty (Leban 1978: 321– 24; Goodman 1998: 122–25; de Crespigny 2016: 465–73). The Latter Han empire had lasted for nearly two hundred years when meeting its formal end, but when the early fifth-century historian Fan Ye 范曄 (398–445) noted the abdication in his History of the Latter Han Dynasty (Hou-Han shu 後漢書) – a work later acknowledged as the standard history of the period – he lamented the demise of the dynasty by using the phrase ‘to end our four-hundred [years]’ (zhong wo sibai 終我四百) (Fan 2003: 392), counting both the Former and Latter Han dynasties (202 bce–220 ce) together.1 Latter Han claimed itself, and is most commonly interpreted, as a continuation of the Former Han dynasty. When Liu Xiu劉秀 (6 bce–57 ce, r. 25–57 ce), posthumously known as Emperor Guangwu 光武, and his followers founded a new dynasty in 25 ce, they cloaked it in the rhetoric of the revival of the Han empire after Wang Mang’s interregnum (9–23 ce) and hence adopted the same dynastic name. The designations of ‘Former’ and ‘Latter’, of course, were coined only after both dynasties had passed into history. For the Latter Han people, they referred to Former Han as the era of the Western Capital (xijing 西京) – the period when the imperial capital was located at Chang’an, so as to distinguish it from their own time. Based on the geographical location of the imperial capitals – Chang’an in the west and Luoyang in the 180

The Latter Han empire

east – the two empires are also conventionally named as Western Han and Eastern Han, after the model of the Western and Eastern Zhou dynasties (ca. 1045–256 bce). For the founders of the Latter Han, the legitimacy of the dynasty rested upon a self-fashioned image of restoring the Han empire usurped by Wang Mang. And the key person to link the two Han dynasties was Liu Xiu. As a distant kinsman of the Han imperial house, Liu Xiu was at first no more than a common member of a local strong clan in the Nanyang 南陽 commandery, and he would never expect nor be able to ascend the throne in peacetime – in fact, his initial ambition was probably no more than to be the commandant of the capital police (zhijinwu 執 金吾, literally the bearer of the gilded mace) (Fan 2003: 405). In addition, there is no evidence whatsoever to suggest that Liu Xiu and his family ever considered restoring the Han dynasty from Wang Mang’s usurpation. But Wang Mang’s governance failure and the ensuing civil war accidentally brought Liu Xiu to centre stage. When the waves of rebellion reached the Lius in Nanyang, it was his elder brother Liu Yan 劉縯 (? – 23 ce; also known as Liu Bosheng 劉伯升) who first capitalized on the situation, gathering forces and collaborating with the insurgents; in contrast, Liu Xiu was overshadowed by his mighty elder brother and could play a supporting role only. Riding the waves of anti-Wang Mang sentiment and of nostalgia for the Han dynasty, different power contenders made use of the name of Han – either by installing a Han imperial kinsman as emperor or by proclaiming Han officialdom – to gather support; the Liu brothers made no exception and exploited their inborn advantage. Liu Yan led his clansmen and allies to join a conglomerate of rebels and served the Gengshi 更始 emperor, who was a cousin of his and was enthroned by some rebel leaders. Conflicts, however, grew tense among these upstarts. The Gengshi faction perceived Liu Yan as a threat and had him killed. Liu Xiu barely survived and fled to the region north of the Yellow River, where he went through considerable hardship and finally established a foothold with the support of prominent local families.2 Liu Xiu and his supporters then translated his imperial kinship into valuable political capital, promoting the idea of Liu Xiu as the legitimate person to restore the Han dynasty in their propaganda campaign; the establishment of a new dynasty was interpreted as a revival and renewal of the glorious Han. This self-fashioned image was very successful and was generally accepted, not only by the contemporaries but also the successive generations. Liu Xiu’s achievement was dubbed in traditional historiography as the ‘Guangwu Restoration’ (Guangwu zhongxing 光武中興). And the two Han dynasties are always counted together, as in Fan Ye’s work noted earlier. To justify such a ‘restoration’, merely adopting the name of Han would not suffice; the Latter Han founders had to take certain measures. Among them, the most significant symbolic meanings were to restore the worship rites of the Liu imperial ancestors – the deceased Former Han emperors – and to re-establish the fiefs of imperial blood relatives, maternal relatives, and meritorious officials that had ceased to exist (Brashier 2011: 148). These measures conveyed an explicit political statement – the Han empire had made a comeback and the imperial house and its associates had resumed their seats. In addition, the restoration of ancestral orders and fiefs of the Lius and meritorious officials resonated with the cardinal principle of filial piety promoted by the Former Han dynasty. It performed the utmost duties of filial piety, i.e., the perpetuity and unity of lineage blood and the assurance of making offerings to the ancestors in an unceasing manner. Latter Han followed its predecessor’s policies of promoting filial piety in various ways. It modeled on the Former Han practice of entitling the emperors – except for the founder, who was himself an object of filial piety – with the posthumous title of filial (xiao 孝); the imperial state also granted filial subjects exemption of levies and corvée duties, entitlement of meritorious ranks and even minor government posts. Latter Han therefore not only forged its image of upholding the Former Han orthodox ideology of filial piety but also deepened the influence of 181

Wicky W.K. Tse

such a principle on its people. A major consequence was a widespread approval of filial behaviors in Latter Han society, which increasingly made the performance of filial piety a crucial feature of contemporaneous political culture (Brown 2007: 32–9). To further demonstrate itself as a continuation of the Han empire in more explicit and mundane ways, the Latter Han founders adopted, at least nominally, an array of Former Han official titles and institutions – although recent studies have shown that quite a number of those institutions in fact had been modified or even invented by Wang Mang (Pu 2002: 96–103; Puett 2010: 148–53). Furthermore, the Latter Han state did not keep its predecessor’s institutions intact but adjusted and even abolished them, if necessary, when the dynasty proceeded; the abolition of universal military service was an oft-discussed example (Lewis 2000: 33–9). The Latter Han empire, in fact, inherited political and military institutions and certain intellectual and cultural trends from its predecessors, whose origins can be traced back to the Warring States period, if not earlier; the empire therefore in many aspects embodied the development of classical cultures of early China. At the same time, Latter Han also witnessed the end, or the beginning of the end, of those institutions and intellectual and cultural trends; the seeds of changes that sprouted in the third and fourth centuries were sown during this period. In retrospect, Latter Han was a transitional period that marked the end of antiquity and the birth of medieval China in various aspects, about which I will elaborate in the last section of this chapter. Despite its cloak of rhetoric as a revival of the Former Han empire, Latter Han was in fact a different regime, no more than a dynasty adopting the same name of Han.The most fundamental difference between the two Han empires was the composition of their ruling elites. Such a difference was significantly determined by their different geographical bases of power. Former Han was a western-centric empire, and its ruling elite had an uneasy relationship with the east since the early days of its establishment. Although many Former Han founding members were of eastern origins, the dynasty was born in the west during the civil war and eventually chose to situate the imperial centre in the strategic Guanzhong region – ‘the land within the passes’. Former Han thus inherited the homeland of Qin and its model of conquering and governing the east from a western base. The divergence between east and west was sharpened by the fact that in the early years of the dynasty territories east of the Guanzhong region, well known as Guandong (east of the passes), were composed of various kingdoms, principalities and marquisates which enjoyed a considerable degree of autonomy and consequently posed serious threats to the imperial court. That the Former Han state strictly delineated the boundary between the Guanzhong and Guandong regions and cautiously examined the people and goods passing through also clearly manifested its entrenched suspicion of the easterners (Oba 2001: 122–37). The Former Han state finally dissolved the eastern rival kingdoms by military and political means. It also enforced sporadic relocation of local strong clans, convicted criminals and refugees of the east to the Guanzhong region and the adjacent northwestern frontier so as to remove the potential troublemakers from the east. A mindset of east-west confrontation had ingrained upon the Guanzhong-based ruling elite, resulting in prejudices and suspicions about the easterners and attempts to restrict their participation in central politics. Coming hand in hand with the political predominance of the western-based elite was the military ascendancy of the west over the east, which is vividly reflected in the civil-military division between the eastern and western halves of the empire as well. Guanzhong and its environs, the homeland of the former Qin state, had long been renowned for the martial valor and fighting spirit of its people, who consequently constituted a vital component of the Qin and Former Han military forces. A group of martial elite from the northwestern frontier became increasingly influential in the military and consequently enjoyed political privileges as the Former Han dynasty progressed. They were the “sons of impeccable families from the six commanderies” 182

The Latter Han empire

(liujun liangjiazi 六郡良家子).3 With their outstanding skills of archery and horsemanship, they played a crucial part in military operations, especially those of northern and northwestern expeditions since the last quarter of the second century bce. It was common for them to assume the commandership in the military; some of them even reached the upper echelons of the imperial bureaucracy (Tse 2012: 116–43). The rise of these military elite came along with the increasing importance of the northwestern frontier in the Former Han imperial strategy. The northwestern frontier, covering the flank of the Guanzhong area, first and foremost served as the first line of defense for the Former Han state against the rival Xiongnu 匈奴 empire. During the reign of Emperor Wu (140–87 bce), when Former Han’s territorial expansionism reached its climax, the region functioned as the bridgehead for Han’s westward expansion and projection of power into Central Asia. Military operations provided the northwestern military elite with ample opportunities for upward mobility.The predominance of the northwestern military elite, in contrast with the emphasis on classical education in the east, further strengthened the stereotypical geographical civil-military division. As a contemporaneous proverb goes, ‘The land west of the passes produces military officers, whereas the land east of the passes produces civil officials’ (Guanxi chu jiang, Guandong chu xiang 關西出將，關東出相). As long as the imperial centre was in the west and the momentum of northwestward territorial expansion still functioned, the composition and vision of the Former Han ruling elite would largely remain western-oriented and militaristic. In contrast, Latter Han was an eastern-centric empire dominated by the elite of eastern origins, who had neither direct connection nor continuity with their Former Han counterparts. Liu Xiu first joined the anti-Wang Mang rebels with his kinsmen around Nanyang commandery and later consolidated his power in the northeast after gaining support from the local strong clans. Most, if not all, of the Latter Han founding members were based in the Guandong region. The selection of Luoyang, instead of Chang’an, as the imperial capital was also a reflection of their interests. For these nouveau ruling elite, the lands west of the passes were not only geographically distant but also politically peripheral. The roles of the conqueror and the conquered reversed between the west and the east, with the former now becoming newly conquered territories. Under the new regime, the eastern-based imperial relatives, local strong clans and officials trained by classical education rose to prominence and occupied positions of influence. They were primarily concerned with their deep-rooted interests in the east and were more prone to defensive actions with respect to the affairs of northwestern frontier. The frontier was different during the two Han dynasties: expanding in Former Han, whereas contracting in Latter Han (Tse 2012: 76–112). As the Latter Han state adopted a policy of retrenchment, the momentum of territorial expansion that once prevailed during the Former Han times disappeared. As a result, the northwestern frontier lost the strategic importance it once enjoyed, and so did its military elite. The northwestern military men were now too rustic in the eyes of the nouveau eastern-based ruling elite with classical training, as mastery of ancient classics became central to education and the principal criterion of distinction between the educated and the uneducated; their path of upward mobility was blocked. Some of them had to transform themselves into civil officials and tried to be as well-educated as their eastern peers in order to achieve career advancement (Tse 2012: 187–213). But for most of the northwesterners, the deprivation and discrimination against them were increasing as the dynasty progressed; it consequently fomented the centrifugal force and triggered uproar, on which Dong Zhuo and his troops capitalized. In short, from the early days to the last years, the mainstays of the two empires’ ruling elites were geographically different. Such a difference further resulted in divergent imperial visions and strategies of the two regimes. From this perspective, no continuity is seen between the two Han dynasties. 183

Wicky W.K. Tse

Even the perpetuity of the imperial lineage between the two Han dynasties, which was the core of the Latter Han’s claim of legitimacy, was not as strong as the traditional records would like to propagate. Comparing with other cases of dynastic restoration or revival in Chinese history, the founding sovereigns of Eastern Zhou, Eastern Jin and Southern Song all belonged to the main line of the imperial families of Western Zhou, Western Jin and Northern Song, respectively. Liu Xiu, however, was merely a member of the imperial family’s collateral branch and was far distant from being a legitimate heir of Former Han.While continuity is more or less evident in the composition of ruling elites of the other three cases, it is not clear in the Former Han–Latter Han transition. Even in the case of reinstating the Former Han fiefs, as Hans Bielenstein points out, Liu Xiu was not wholeheartedly willing to revive the old order fully. He did not renew the majority of the old marquisates, whereas a batch of new ones were created for his own relatives (Bielenstein 1986: 256). Based on this analysis, I would hereby suggest considering the Latter Han as another dynasty rather than a continuation of the Former Han. Only by treating the Latter Han’s self-fashioned image of being a restoration of the Former Han as no more than political rhetoric shall we be able to understand the rationales behind certain policies of the Latter Han which were in conflict with the preceding practices and to situate the dynasty as a transitional phase between antiquity and early medieval China.

An embattled empire Conventional Chinese histories always depict the Latter Han as a feeble and lackluster empire. Comparing with the Former Han, the Latter Han was weaker, as reflected in its shorter reign years, smaller territories and smaller registered population (Bielenstein 1947: 125–63); being less aggressive and expansionistic in foreign affairs was also a sign of its inferiority. Partly because of its inferior status, Latter Han has long been overshadowed by the mighty Former Han, and thus understudied. Although not as splendid as its predecessor, Latter Han in fact went through various challenges and adapted itself to new situations quite successfully – for it lasted for nearly two hundred years. The image of a weaker empire was significantly contributed to by the more troubled and complicated circumstances it faced, both externally and internally. In this section, I shall highlight the complexity and diversity of interactions between external and internal conflicts in shaping the military, political and social circumstances that the Latter Han needed to deal with and their impacts on the decline of the empire.

External challenges The Latter Han ushered in a new era of inter-polity relations in eastern Eurasia, in that the predominantly bilateral confrontation between the Former Han and the Xiongnu transformed into multilateral conflicts – partly due to the decline of the Xiongnu hegemony over the steppes, involving not only the Latter Han and the Xiongnu but also the Qiang 羌, the Wuhuan 烏桓 and the Xianbei 鮮卑, not to mention the recurrence of conflicts with the indigenous peoples along the southern frontier. The relationship between the Latter Han and its neighbors was one of uneasy peace interrupted at frequent intervals by the outbreak of armed conflicts in certain border regions. A major hot spot was the northwestern frontier, where the Latter Han fought bitterly with the Qiang tribes in a series of intermittent wars from the mid-first to the late second centuries. Among all the rivals, the Qiang people brought the most devastating disasters to the Latter Han, since the wars with them were costly and lengthy – the prodigal expenditure led to the financial collapse of the empire. Dong Zhuo’s armies also grew out of that battlefield.The protracted conflicts, however, demonstrated the Latter Han’s undeniable capability for survival. 184

The Latter Han empire

It was able to cope with external enemies from various directions and lasted for two centuries until being disintegrated by the warlords from within. Generally speaking, different from the Former Han’s aggressiveness, the Latter Han adopted an attitude of self-restraint toward foreign affairs since its early days and throughout most of its course of history, which in turn shaped its embattled situation. In the process of empire building, Liu Xiu put priority on domestic rehabilitation to give the people who suffered from the civil war a respite and therefore employed a policy of retrenchment in many aspects. Wang Mang’s overambitious foreign policies and the subsequent fiasco cast a shadow on Liu Xiu’s planning and dissuaded him from any foreign adventure. As a result, while eliminating his power contenders one after another, Liu Xiu carefully dealt with those who were backed by the Xiongnu to avoid drawing the steppe power into the conflicts. The Xiongnu, however, had somewhat rejuvenated during the last decades of the Former Han and militarily challenged Wang Mang’s claim as the overlord of all under heaven. With the Xiongnu once again becoming a military threat to the northern frontier of the Chinese empire, the situation in the Western Regions – over which the Former Han and the Xiongnu had long struggled for domination – also fluctuated and was complicated by a mixture of the rise of ambitious local powers, the influence of the Xiongnu and the retreat of Chinese forces ensuing the downfall of Wang Mang. When some of the Central Asian polities sent delegates to Luoyang in 45 ce and requested the reinstatement of the Protector General of the Western Regions (Xiyu Duhu 西域都護) to maintain the region’s balance of power, Liu Xiu took a restrictive stand and declined the request. The refusal implied Latter Han giving up the Western Regions; mutual official diplomatic connections were hence suspended. Obviously, Liu Xiu was not willing to let his nascent empire get into the Central Asian conflicts. Another reflection of Liu Xiu’s unwillingness to involve his empire in foreign affairs was his adoption of the time-honored policy of ‘playing off one barbarian against another’ (yiyi zhiyi 以 夷制夷) when facing the division of the Xiongnu; the aim was to prevent a unified power in the steppe. Serious struggles over succession broke out among the Xiongnu leadership in the late 40s ce and tore the steppe power into two halves. Liu Xiu chose to recognize the southern Chanyu – the weaker one – and provided his tribes with a shelter within his empire’s northern borders. The ideal situation would be that the southern Xiongnu buffered the Latter Han from the northern Xiongnu, and that the two Xiongnu powers would check each other from becoming too strong to threaten the Chinese empire. The Latter Han would thus pay the least cost to defend its northern frontiers if everything worked out as Liu Xiu had expected.The decision of embracing the southern Xiongnu, however, has long been criticized by commentators of later generations as the underlying cause of the Western Jin’s tragic end in 310 ce – the dynasty was toppled by the descendants of the southern Xiongnu who resided in the imperial domain. For example, modern sinologist Hans Bielenstein harshly criticized Liu Xiu, who had ‘committed the greatest error of his reign, a blunder which belongs among the worst in Chinese history’. Instead, he believed Liu Xiu should, in coalition with the southern Xiongnu, have attacked the federation of the northern Xiongnu, [for] such a campaign . . . almost certainly would have been successful. The southern Chanyu would have returned to the lands north of the Gobi as the sole ruler of the Xiongnu, and China would have regained the northwestern border commanderies. For Bielenstein, Liu Xiu’s failing to destroy the northern Xiongnu once and for all and send their southern cousins back to the steppes was a ‘short-sighted policy’ (Bielenstein 1986: 267–8). Such an argument would be echoed in the conventional explanation of how the barbarian 185

Wicky W.K. Tse

migrants and mercenaries toppled the Roman empire – in fact, an analogy is commonly made between the fall of the Roman empire and its Chinese counterpart. Bielenstein’s comments, however, are based on a series of unproven assumptions and not unproblematic. First of all, there was no evidence to support his theory that the northern Xiongnu would be defeated by a joint military campaign and the southern Xiongnu would return to their homeland. Even if so, there was no guarantee that the Latter Han and the southern Xiongnu would maintain a peaceful relationship thereafter. Most important, those comments oversimplify and underplay the situations Liu Xiu faced and the options – resources and tools – that were available to him. For Liu Xiu, an economical way of dealing with the Xiongnu and protecting the empire was preferable. Such a self-restrictive attitude was clearly shown in a statement made by Liu Xiu in 51 ce, when he rejected the northern Xiongnu’s request for resuming the marriage-alliance with the Latter Han, as their ancestors did with the Former Han. Two courtiers submitted a petition, saying, Now, the caitiffs [northern Xiongnu] are suffering from the plague that has killed their people and herd and from the drought and locusts that have devastated their land; they are exhausted and feeble, and their strength is not able to match that of a commandery of the Han. The life and death of these people living ten-thousand li away now hinge on the decision of Your Majesty. Is it still appropriate to hold steadfastly the principle of civil virtue and therefore disrupt the military affairs? We propose dispatching armies to the frontiers, issuing warrants that promise to pay handsomely for killing and capturing the Xiongnu, ordering the foreign allies like Gaojuli 高句麗, Wuhuan, and Xianbei to attack the Xiongnu from the left, and launching the troops of the four commanderies of Hexi 河西 and the commanderies of Tianshui 天水 and Longxi 隴西, as well as the forces of the Qiang and the Hu 胡 to assault the Xiongnu from the right. If so, only in a few years will the northern caitiffs be erased. (Fan 2003: 695) Liu Xiu, however, elaborated his thought in an edict, saying, The Book of Huang Shigong 黃石公 reads, ‘The soft can overcome the tough, and the weak can overcome the strong. The soft is the virtue, whereas the tough is the vice. The weak earns the aid of the benevolence, whereas the strong incurs discontent.’That said, ‘the virtuous monarch employs what pleases him to please the people, while the monarch without virtue employs what pleases him to please himself. He who pleases others will enjoy pleasure for a long time, whereas he who only pleases himself will not last long. He who ignores the nearby and plots for the distant will trouble himself without gaining any merit; whereas he who ignores the distant and focuses on the nearby will enjoy one’s ease and have a good ending. An administration enjoying its ease is filled with loyal ministers, whereas an administration troubling its labour is filled with fussy men.’ That said, ‘he who focuses on expanding his territory will become destitute, whereas he who focuses on expanding his virtue will become strong. He who satisfies with what he has is content, whereas he who covets what others have is ruthless. Though a ruthless administration could attain initial success, it will fail eventually.’ Now, the empire does not enjoy good administration, and there are unceasing natural disasters.The people are frightened and not able to save themselves. How could We turn our labour to the land beyond the frontier? Confucius says, ‘I am afraid that the worries of the Jishi季氏will not be caused by the Zhuanyu 顓臾.’ Besides, the 186

The Latter Han empire

northern barbarian is still strong, and there are always unfounded rumors that circulate around the farm colonies and sentry posts along the frontiers. If it is possible to destroy the arch enemy by mobilizing half of the empire, how could it not be Our utmost wish? But if it is not the time, We would rather rest the people. (Fan 2003: 695–96) Henceforth, according to the historical record, no general dared to talk about military affairs again. The case cited here tells, first, that there were voices in the court who proposed launching large-scale expeditions as a once-and-for-all solution to the northern Xiongnu problem; second, that Liu Xiu was not an absolute pacifist but was only reluctant to launch any military campaign that would incur huge costs for his empire and subjects – in other words, maintaining domestic stability was Liu Xiu’s prime consideration. By rejecting the aggressive approach, Liu Xiu insisted on coping with the Xiongnu in his self-restrained way. He consistently and staunchly sided with the southern Xiongnu when the northern Chanyu tried to make good terms with him, lest any change break the delicate balance between the two Xiongnu powers and cause trouble for his empire. Facing the need for rehabilitation and the limitation of resources he could expend, Liu Xiu was pragmatic to employ a policy of self-restraint. Quite contrary to what Bielenstein criticized as a short-sighted policy, Liu Xiu was aiming to lay a solid foundation for the long-term development of his dynasty. His policy was in fact admired by the intellectuals of the Latter Han; the renowned scholar Ban Gu 班固 (32–92) was an example. In the History of the Former Han Dynasty compiled by the Ban family, Ban Gu lamented that the Qin and Former Han empires had made futile attempts to deal with the northern barbarians without working out any good method. The ideal way that Ban Gu praised, however, was close to what Liu Xiu carried out – to leave the barbarians in their own way and not to take initiative in engaging them (Ban 1962: 3833–4). Even though Ban Gu himself participated in a military expedition against the northern Xiongnu, he still honored Liu Xiu’s policy of conducting foreign relations; so did other literati trained in classics. Other measures consistent with Liu Xiu’s self-restraint and domestic-oriented policy included the reduction of the numbers of counties and official posts, demobilization of soldiers, abolition of regional military officers and universal military service (Lewis 2000: 33–75), propagation of defensive measures rather than offensive actions along the frontiers and inward migration of frontier residents. All these measures coherently served the clear purposes of downsizing the state apparatus, cutting government spending and minimizing the chances of border conflicts. Defense-prone as his foreign policy was, however, Liu Xiu’s empire was never free from external military confrontations. Border conflicts, in varying intensities, intermittently broke out between the Latter Han and its neighbors. In the early years of the dynasty the Xiongnu problem was acute, and later the Qiang replaced them as the most serious threat and cost the Latter Han expensive and exhaustive military campaigns; not to mention the Man and Yue peoples scattered along the southern frontiers and the Wuhuan and Xianbei in the northeast, which all caused chronic headaches to the empire. Facing the multi-frontal challenges and the changing circumstances, Liu Xiu’s successors found the strategy of self-restraint might not always be the best option.4 Salient changes happened during the reigns of the three succeeding emperors after Liu Xiu, in which the Latter Han took the initiative to attack its rivals – as a means of effective defense. Two military expeditions against the northern Xiongnu were launched in 73 ce and 89 ce, under the command of General Dou Gu 竇固 (d. 88 ce) and General Dou Xian竇憲 (d. 92 ce) – both 187

Wicky W.K. Tse

were members of a consort family – respectively; Ban Gu took part in the latter one as a member of the general staff. Both campaigns achieved spectacular successes by invading deep into the northern Xiongnu’s domain and forcing them to retreat further northward. The northern Xiongnu henceforth disappeared from the Latter Han’s frontiers; the southern Xiongnu, for most of the time, served as a Dependent State (shu guo 屬國) and a main source of mercenaries of the Latter Han empire. The Xiongnu ceased to be a military threat, but the Latter Han did not enjoy peace for long, since the Qiang along the western frontiers became a new menace to the stability of the empire. Along the western frontiers, tension and rivalry between the Latter Han state and the Qiang tribes grew fierce, resulting in protracted warfare that intermittently lasted for over a century. The first armed conflict between the Latter Han and the Qiang broke out in 35 ce and ended with the suppression of the latter. Fights broke out again consecutively in 56, 57 and 58 ce; with a brief respite, another round began in 76 ce. These were all fights of low intensity, but the Latter Han could merely put them down without achieving any decisive victory; the peace was fragile. The situation turned out of control again in 86 ce, and not until three years later were the Qiang suppressed. In 92 ce, however, military confrontation resumed, and the Latter Han had to spend the following six years making the Qiang surrender. In addition to these confrontations, skirmishes happened intermittently. The hostility between the two sides kept escalating, and three large-scale wars broke out in the second century ce, bringing devastating effects to the empire. The first began in 107 ce and ended in 118 ce, the second lasted from 140 to 145 ce, and the third was fought through 159 to 169 ce (de Crespigny 1984: 76–172; Tse 2012: 236–303). The Qiang wars, which lasted over half of the course of the Latter Han dynasty, not only bankrupted the empire and caused heavy casualties but also deepened the political unease between the Guanzhong and Guandong regions, which further triggered the disintegration of the empire (Tse 2012: 259–303). The Qiang issue was further complicated by the composition of the so-called Qiang people, who, from the perspective of the Latter Han state, included both restive barbarians and imperial subjects. The military conflicts were therefore treated as both foreign wars and internal rebellions. The question of who were the Qiang was never easy to answer, as fully discussed by modern scholar Wang Mingke (Wang 98–99, 129–32; Tse 2012: 217–36). As the wars intensified, the Latter Han state put all the northwestern people who were on the opposite side under the name of Qiang, regardless of their ethnic origins.With such a disinterested, and even hostile, attitude towards the northwesterners, the Latter Han state further estranged them in two ways. Firstly, when facing the seemingly endless wars with the Qiang, some officials in Luoyang proposed, not once but thrice, to give up the war-trodden northwestern region. These proposals were all rejected after fierce court debates, but the advocacy of the idea and the wide support it gained were revealing. The eastern-based imperial court and officials did not take the northwest seriously as an integral part of the empire. Being a different empire, the Latter Han had no obligation to keep the Former Han territory intact. Secondly, although the Latter Han state formally rejected the plans of abandonment, it still ordered the local governments to carry out forcible evacuation as an alternative. The local authorities therefore withdrew and also pulled out the residents. The rationale behind the forcible withdrawal was to deny the enemy’s access to any resources by a kind of scorched-earth tactic (Tse 2012: 259–86). It might make sense in military strategy but was not to be easily accepted by the residents. Most important, the imperial army dispatched to carry out the plan ravaged the home and property of the residents during the relocation. It only resulted in increasing hatred and mistrust of the northwesterners against the imperial state, who were driven away from supporting the empire. This is no longer merely an external challenge facing the Latter Han but also an internal conflict upsetting the empire. 188

The Latter Han empire

Internal conflicts While the external warfare wreaked havoc in the frontier regions and incurred heavy costs for the Latter Han state, the internal political and social conflicts deeply threatened the stability of the empire. There were three kinds of internal conflicts troubled the Latter Han. The first was between the Guandong and Guanzhong regions, which was further complicated by the Qiang wars; the second was the power struggle among different factions in the imperial court, paralyzing the function of the government and undermining the imperial authority; the third was the large-scale uprising that broke out in 184 ce, which nearly toppled the dynasty and consequentially fostered the regional warlords. As noted in the preceding section, the Latter Han located its power base in the east, and the vision of the empire was also set accordingly. The northwestern frontier no longer maintained the strategic importance of flanking the imperial capital and serving as the springboard of power projection into the Western Regions; the northwest was now too remote for the imperial state in Luoyang. The northwestern people, especially the martial elite, also lost the superior political status they once enjoyed in the Former Han period. The new ruling elite of the Latter Han empire were eastern-based and trained in classics.They honored civil value over martial spirit and looked down on the military skills of the northwesterners as no more than a sign of being uncouth and uncivilized. Both the northwestern region and its people were thus placed in an inferior status. The northwestern martial elite could no longer attain high offices through military feats; some of them even had to enter government service through civil positions for the sake of pursuing future career advancement (Tse 2012: 194–213). The ruling elite’s contempt for the northwesterners was clearly manifested in the proposals of giving up the northwestern region in the midst of the tumultuous Qiang wars. The ensuing enforcement of destroying their home greatly disappointed and angered the northwesterners. The devastation of the Qiang wars further stirred the hatred and distrust of the northwestern warriors towards the imperial court, which finally turned into military resources for the ambitious Dong Zhuo. The tension between the eastern-based ruling elite and the northwestern military men only constituted a latent force to trouble the empire, if not intensified by the Qiang wars, but the fierce and bloody political infighting in the imperial court had consistently and explicitly staggered the dynastic rule. The core of the political struggle was the competition for gripping the imperial power. On one side, there were the empress dowagers and their male relatives, trying to control the young monarchs; on the opposite side were the emperor, who aimed to preserve his own power, and the eunuchs, who had established a very close-knit relationship with the throne that also served their own benefit. The imperial bureaucracy was also divided into different factions for supporting either of the two sides and for pursuing their own agendas. The nature of the Latter Han dynasty as a coalition of powerful clans determined the essence of the power struggle. Even the lesson of Wang Mang usurping the Former Han through intermarriage between his family and the emperors did not effectively prevent the Latter Han imperial house from granting power to the relatives of imperial consorts. Most, if not all, of the Latter Han empresses were from strong, wealthy and well-educated clans. Marriage with the imperial house allowed those clans to extend their power and influence. A direct result was that the male members of those clans and their protégés occupied important positions at different levels of the Latter Han officialdom; they also commanded a swath of retainers and followers as private forces.When there was a strong and mature emperor, like the first three rulers of the dynasty, the ambition and power of the consort families could be restrained. But beginning from the reign of the fourth emperor, a vicious cycle of power struggle set in. 189

Wicky W.K. Tse

The frequent occurrence of premature deaths of emperors and succession of young heirs became a salient feature of the Latter Han imperial politics. Starting from the fourth monarch, Emperor He (r. 89–105 ce), until the last ruler of the dynasty, there were ten sovereigns, and the eldest one ascended the throne at the age of 15, whereas the youngest one was an infant of only 100 days old; six empress dowagers therefore assumed regency. Hence, the following sequence of events became a recurrent theme, though in coincidence to a certain extent, during the Latter Han dynasty: a reigning emperor met untimely demise and left an underage heir apparent, sometimes even an infant, or none; the empress dowager and her male relatives, mostly her father or siblings, would enthrone a child emperor, for the sake of easy control, and assume regency; the emperor would plot with the eunuchs to restore his grip of power when he grew up; tension between the two sides intensified and eventually turned into a bloody purge. In addition, conflicts also broke out among the consort families, for they usually treated each other as competitors in accessing the imperial authority. Officials at different levels, for their own interests, also took sides in the power struggles. As a consequence, imperial policies and strategies were deeply affected and coloured by the factional strife. For example, Dou Xian, a brother of Empress Dowager Dou, was accused by his opponents of launching the northern expedition in 89 ce for his own ambition rather than for the interest of the empire. And Dou Xian was finally killed by the eunuchs plot after the death of the empress dowager; the historian Ban Gu, as a protégé of Dou Xian, was also executed with the downfall of his patron. The two most merciless purges happened in 168 ce and 189 ce. In the first incident, the eunuchs deployed the royal guards against the empress dowager’s father, Dou Wu 竇武 (d. 168 ce). Dou Wu committed suicide, and his followers and allies were either killed or arrested. In addition, the eunuchs launched an empire-wide proscription targeted at the officials and literati who were identified as dissenters (de Crespigny 1975: 1–36).The second incident began with the eunuchs’ murder of He Jin 何進 (d. 189 ce), a brother of the empress dowager, before He carried out his plan of killing all eunuchs. An official named Yuan Shao 袁 绍 (d. 202 ce) then led troops to break into the imperial palace and slaughtered most of the eunuchs, throwing the imperial capital into chaos, eventually providing the opportunity for Dong Zhuo to move in. The dire struggle between the consort families and the eunuchs not only cost human lives but also severely damaged the morale of officialdom and triggered political disorder. All these consequently devastated and discredited the imperial authority. Along with the power struggle came widespread corruption. Many unqualified or badly behaved aspirants and opportunists entered government service via the recommendation and/or appointment by the consort families and the eunuchs. In order to return a favour to the patrons and to pursue their own interests, these officials would put earning profits as their top priority at the expense of the subjects in their jurisdictions. Corruption quickly penetrated all levels of the imperial government and caused great suffering and discontent among the people.Voices against corruption became louder and louder. Seizing such an opportunity, the noble Dou Wu fashioned himself as an anticorruption hero via attacking the eunuchs and their underlings and thus earned a good reputation and wide support among the literati. The corruption, however, was not curbed, which was partly responsible for triggering a large-scale rebellion in 184 ce. The Yellow Turban Rebellion, widely known for the rebels wearing yellow headgear, was a mixture of religious and anti-dynastic movements.5 The general discontent against the corrupted imperial state played a crucial role in providing a hotbed for the spread of messianic belief and incentive of popular mobilization. Although the imperial troops quickly defeated the main force of the rebels, the aftermath was even more devastating.The remnants of the rebels lingered in several regions and turned into roaming bandits who severely upset the empire. Some of them 190

The Latter Han empire

finally surrendered and were transformed into private troops of the regional governors. These governors were the real winners of the rebellion, as their positions and power strengthened, their wealth increased, their armies were enlarged and their territories expanded in the midst of the suppression campaigns and afterward. When Dong Zhuo entered Luoyang and put the emperor under custody, the regional warlords immediately disavowed the imperial centre and established their de facto kingdoms, marking the collapse of the Latter Han empire and the prelude of the Three Kingdoms period (220–265 ce).

The end of antiquity In hindsight, the Latter Han dynasty is situated as a period of transition in several aspects between early and medieval China – as is designated by modern historiography. Behind the facade of being a continuation of the Former Han dynasty, there were dynamics of change in shaping the Latter Han period as different from its predecessors. The final collapse of the empire at the turn of the third century did not merely indicate the failure of a dynasty but also signified the end, or the beginning of the end, of antiquity in Chinese history. The transitions witnessed in the Latter Han period can be summarized in four aspects. The first three transitions were changes in political and military terms which were closely linked, and the fourth occurred in the intellectual world. They started at different times, but all were salient features for the period in question. The first transition marked the end of the first model of Chinese empire. Although the Latter Han only borrowed the name of Han and was not a real revival of the Former Han dynasty, as analyzed earlier, it did inherit the vast territories, an array of political ideas and institutions, and most importantly the form and ideal of empire from its predecessors – the Qin and Former Han dynasties. The ‘unification’, as claimed by traditional history, and establishment of the first empire by the Qin dynasty was significantly a product of the dynamics of empire-building that had started since the preceding Warring States period. Though the Qin as a unified empire was short-lived, it laid the foundation of imperial governance for its successors. The Former Han developed and consolidated the imperial system and made the ideal of universal empire become deeply ingrained in Chinese culture (Pines 2012: 19–25). The Latter Han then functioned on the same basis of imperial apparatus. In this sense, the Qin, Former Han, Wang Mang’s Xin, and Latter Han dynasties constituted the same set of early Chinese empires or the so-called first model of Chinese empire. Although sophisticated as the imperial system, it failed to overcome the challenges that finally dismantled the imperial authority of the ruling Latter Han dynasty. The collapse of the Latter Han opened a nearly four-century period of political disunion and ended the first imperial model in Chinese history. Although the Western Jin dynasty achieved a re-unification in the post-Latter Han period, the success was ephemeral. The dynamics of disintegration had already set in, and the short-lived Western Jin was not able to reverse the tide. It was not until the late sixth and early seventh centuries when the dust settled and the Sui and Tang dynasties rebuilt the universal empire that the second model of Chinese empire was established.6 The second transition was embodied in the adaptation of institutions which had crucially served as underpinnings in the building of the first empire. Among all the institutional changes, the abolition of universal conscription by the Latter Han was the most significant one. Universal conscription was a legacy of the Warring States period. To cope with the endless warfare, the Warring States needed to organize huge armies with a stable supply of manpower. The introduction of universal military service provided a solution, which allowed the commoners’ participation in warfare on an unprecedented scale (Tu 1990: 50–96; Lewis 1990: 54–67). 191

Wicky W.K. Tse

The system facilitated the birth of the Qin empire and henceforth was an essential institution during the Former Han. Compulsory enrollment and fulfillment of military service became an obligation to all able-bodied men, who were prepared to constitute the backbone of the imperial military force. The system was also an effective way to circulate imperially sponsored ideology and to forge people’s solidarity with the imperial state (Yates 2011: 360–64). Liu Xiu, however, for practical and pragmatic consideration over various factors, as shown in Mark Edward Lewis’s detailed study, formally abolished the system – though the system had been gradually transformed even in the Former Han times (Lewis 2000: 34–61). The recruitment and composition of military force thereafter changed, and the backbone of the Latter Han imperial army was no longer conscripts but professional soldiers, volunteers with military skills and foreign auxiliary troops. Due to the need arising from frequent external warfare and the ambition of rival factions to build loyal military forces, military officers at various levels were given a free hand to recruit and command their own troops, which was blamed by some critics in the last years of the dynasty as a crucial factor contributing to the decline and fall of the dynasty. The abolition of universal military service was not only a response of the Latter Han to the new circumstances, as Lewis has pointed out, but also a farewell to a military mechanism that had long buttressed the early Chinese empires.7 After the abolition, no regimes in imperial Chinese history ever fully resumed the system. Its far-reaching effects can be seen even in the early twentieth century, when modern Chinese historian Lei Haizong 雷海宗 (1902–1962) portrayed the abolition as a crucial step in the degeneration of Chinese martial spirit, for which his countrymen suffered in the foreign incursions since the nineteenth century (Lei 1940: 44–61, 125–59). The third transition was shown in the Latter Han’s turning inward rather than continuing the outward expansion of its predecessors. At least since the Warring States period, frontier states like Qin, Zhao,Yan and Chu were all expanding outward; the Qin empire, which unified all rival states, even launched a new wave of outward territorial expansion. Such a tide only met with interruption in the early decades of the Former Han dynasty, which learned a lesson from the sudden collapse of the Qin. But territorial expansionism gained momentum again during the reign of Emperor Wu, and the empire extended greatly in all directions until it reached certain natural barriers and strategic limits. The Former Han’s enterprise of territorial expansion also came hand in hand with the emergence of the Xiongnu hegemony over the steppes (Di Cosmo 2002: 161–205). The confrontation between the sedentary Chinese and nomadic empires was thus mixed with conflicts over controlling territories and attempts at projecting power into Central Asia. Even Wang Mang’s Xin dynasty followed the same path. Liu Xiu, however, was reluctant to engage in external warfare and adopted a policy of retrenchment, as mentioned earlier. Although several foreign military campaigns were launched, the Latter Han rulers basically did not divert from the domestic-oriented strategy formulated by the founding emperor. The Latter Han state finally closed the formal relationship with the Western Regions after intermittent cut-offs and connections, and there were even attempts to give up the northwestern region in the midst of the Qiang wars.The foreign relations of early Chinese empires also changed during the Latter Han period.The main thread running through the foreign relations of the Former Han was the bipolar struggle with the Xiongnu, but such a pattern ended in the early Latter Han and was transformed into a multi-polities system. The Xiongnu’s breakdown greatly changed and reshaped the political and diplomatic landscapes of the East Asian continent, for it released the room for new players. Facing the new circumstances, the Latter Han decided not to actively engage in foreign affairs, thus encouraging the growth of the newly emerged powers, which in turn wreaked havoc on the empire. The new powers like the Qiang, the Xianbei, and the Wuhuan, as well as the southern Xiongnu, all significantly took their part in the Chinese world 192

The Latter Han empire

during and after the collapse of the Latter Han. Some of their descendants even established their kingdoms in northern China during the fourth and fifth centuries. The fourth transition was seen in the rise of a new intellectual movement which gradually replaced the mainstream tradition of scholarship that developed in the Former Han period. In the late Warring States period, there was a tendency of unifying the intellectual world and putting it under state control. The compilation of the Springs and Autumns of Mr. Lü (Lüshi chunqiu 呂氏春秋), under the auspices of the Qin prime minister Lü Buwei 呂不韋 (ca. 292–235 bce), was an early example. The notorious Qin biblioclasm initiated by the chancellor Li Si shortly after the empire’s establishment was another attempt at state control. The Former Han imperial state also sponsored political ideologies that fit its needs. The so-called New-text Classicism/ Confucianism (so named for studying the texts that were written in ‘modern-day’, i.e., Qin-Han scripts, whereas the so-called Old text referred to pre-Qin scripts) finally outmatched its competitors, won imperial support and assumed the position of orthodox ideology.The overwhelming triumph of classicism, under the imperial auspices, gradually suffocated the intellectual world with rigid discipline of scholastic traditions and the mass production of canonical exegeses which were all purported to be the definitive one. Such an intellectual tradition reached its peak in the early Latter Han, which also helped fashion the image of the Confucianization of Latter Han in various aspects, such as law and familial relationship.8 The state-sponsored classicism became increasingly hardened and narrowed (Goldin 2007: 164–65). Meanwhile, it also began to decline. Having been brought up under a rigorous scholastic tradition of classicism, some of the well-educated elites began to shift their ideological concerns. They questioned the incumbent intellectual hegemony and attempted to search for a new order in the intellectual world. Such a reverse of the tide was especially clear and trenchant in the last years of the Latter Han as shown in the integration of the Old- and New-text classicism by masters such as Ma Rong 馬融 (79–166 ce) and Zheng Xuan 鄭玄 (127–200 ce), the new interpretation of the classics by the so-called Jingzhou school (Jingzhou xuepai 荊州學派) and the rise of abstruse learning (Xuanxue 玄學) (Qu and Wang 2013: 378–454). These new intellectual trends collectively dealt a great blow to the state-sanctioned classicism and heralded a new era of Chinese intellectual history (Yü 1985: 121–55). The Latter Han empire, on one hand, inherited the legacy from its predecessors; on the other hand, it also witnessed the end or reversal of some traditions of the early Chinese empires. The momentum of the early empires ended in the Latter Han; so did the age of antiquity. With the collapse of the Latter Han empire, the realm of China went through a long and winding road for nearly four centuries to search for a new outlet.

Acknowledgment This chapter was written under the financial support of the Early Career Scheme (project no. 25608215) from the Research Grants Council (RGC) of Hong Kong.

Notes 1 The only and also most updated general history of the Latter Han dynasty in English is de Crespigny 2016. My review of this book can be found in Tse 2017. 2 Hans Bielenstein provided a book-length study on the saga of Liu Xiu and his followers during the civil war; see his ‘The restoration of the Han dynasty: volume II, the civil war’, in Bielenstein 1959: 1–287. 3 The six commanderies were Longxi 隴西, Beidi 北地,Tianshui 天水, Anding 安定, Shang 上 and Xihe 西河. 4 For a comprehensive overview of the Latter Han’s northern strategy, see Crespigny 1984.

193

Wicky W.K. Tse 5 For the most detailed treatment of the Yellow Turban Rebellion in English, see Michaud 1958: 41–127. 6 For the classification of the first and second models of empire in Chinese history, see Tanigawa 1995: esp. 6–9. 7 Raimund Kolb, however, argues that Lewis’ emphasis on the significance of the abolition of conscription is too simplistic. See Kolb 2006: 435–64. 8 Paul Goldin has provided an updated discussion of the topic. See Goldin 2012: 1–31.

Works cited Ban Gu 班固 (32–92 ce) (1962) Han Shu 漢書, Beijing: Zhonghua shuju. Bielenstein, H. (1947) ‘The census of China during the period 2–742 A.D.’, in Bulletin of the Museum of Far Eastern Antiquities, 19: 125–63. Bielenstein, H. (1959) ‘The restoration of the Han dynasty: volume II, the civil war’, in Bulletin of the Museum of Far Eastern Antiquities, 31: 1–287. Bielenstein, H. (1986) ‘Wang Mang, the restoration of the Han dynasty, and Later Han’, in D.Twitchett and M. Loewe (eds.) The Cambridge History of China:The Ch’in and Han Empires, 221 B.C.–A.D. 220, vol. 1, Cambridge: Cambridge University Press: 223–90. Brashier, K. E. (2011) Ancestral Memory in Early China, Cambridge, MA: Harvard University Asia Center. Brown, M. (2007) The Politics of Mourning in Early China, Albany: State University of New York Press. de Crespigny, R. (1975) ‘Political protest in imperial China: the great proscription of Later Han, 167–184’, in Papers on Far Eastern History, 11, Canberra: Australia National University, Dept. of Far Eastern History: 1–36. de Crespigny, R. (1984) Northern Frontier:The Policies and Strategy of the Later Han Empire, Canberra: Faculty of Asian Studies, Australian National University. de Crespigny, R. (2016) Fire over Luoyang: A History of the Later Han Dynasty 23–220 ad, Leiden: Brill. Di Cosmo, N. (2002) Ancient China and Its Enemies: The Rise of Nomadic Power in East Asian History, Cambridge: Cambridge University Press. Fan,Ye 范曄 (398–445 ce) (2003) Hou-Han Shu 後漢書, Beijing: Zhonghua shuju. Goldin, P. (2007) ‘Xunzi and Early Han Philosophy’, in Harvard Journal of Asiatic Studies, 67.1 (June 2007): 135–66. Goldin, P. (2012) ‘Han law and the regulation of interpersonal relations: “the Confucianization of the law” revisited’, in Asia Major, 3rd series, 25, .1, Taipei: Academia Sinica: 1–31. Goodman, H. L. (1998) Ts’ao P’i Transcendent: The Political Culture of Dynasty-Founding in China at the End of the Han, Seattle: Scripta Serica. Kolb, R. T. (2006) ‘Excursions in Chinese Military History’, in Monumenta Serica 54: 435–64. Leban, C. (1978) ‘Managing heaven’s mandate: coded communication in the accession of Ts’ao P’ei, A.D. 220’, in D. T. Roy and Tsuen-hsuin Tsuen (eds.) Ancient China: Studies in Early Civilization, Hong Kong: The Chinese University Press: 315–42. Lei, Haizong. (1940) Zhongguo wenhua yu Zhongguo de bing 中國文化與中國的兵, Shanghai: Shangwu  yinshuguan. Lewis, M. E. (1990) Sanctioned Violence in Early China, Albany: State University of New York Press. Lewis, M. E. (2000) ‘The Han abolition of universal military service’, in H. van de Ven (ed.) Warfare in Chinese History, Leiden and Boston: Brill: 33–75. Michaud, P. (1958) ‘The yellow turbans’, in Monumenta Serica, vol. 17, Sankt Augustin: Monumenta Serica Institute: 41–127. Oba, O. (2001) ‘The ordinances on fords and passes. Excavated from Han tomb number 247, Zhangjiashan.’ Translated and edited by David Spafford, Robin D. S. Yates and Enno Giele; with Michael Nylan, in Asia Major, 3rd series, 14.2: 119–41. Pines,Y. (2012) The Everlasting Empire:The Political Culture of Ancient China and Its Imperial Legacy, Princeton: Princeton University Press. Pu Xianqun 卜憲群 (2002) Qin Han Guanliao Zhidu 秦漢官僚制度, Beijing: Shehui kexue wenxian chubanshe. Puett, M. (2010) ‘Centering the realm:Wang Mang, the Zhouli, and early Chinese statecraft’, in B. A. Elman and M. Kern (ed.) Statecraft and Classical Learning: The Rituals of Zhou in East Asian History, Leiden and Boston: Brill: 129–54. Qu Anquan 瞿安全 and Wang Kui 王奎 (2013) Jingzhou Xuepai Jiqi Yingxiang Yanjiu 荊州學派及其影響 研究, Wuhan: Hubei renmin chubanshe.

194

The Latter Han empire Tanigawa Michio 谷川道雄. (1995) Sekai Teikoku no Keisei世界帝国の形成, Tōkyō: Kōdansha. Tse, W. K. W. (2012) ‘Dynamics of disintegration: the Later Han empire (25–220 ce) & its northwestern frontier’, Ph.D. dissertation, University of Pennsylvania. Tse, W. K. W. (2017) ‘Review of Rafe de Crespigny, Fire over Luoyang: A History of the Later Han Dynasty, 23–220 ad’, in Journal of the Royal Asiatic Society, Series 3. doi:10.1017/S1356186317000074 Tu Cheng-sheng 杜正勝 (1990) Bianhu Qimin: Chuantong Zhengzhi Shehui Zhi Xingcheng 編戶齊民：傳 統政治社會之形成, Taipei: Lianjing chuban shiye gongsi. Yates, R. D. S. (2011) ‘Soldiers, scribes, and women: literacy among the lower orders in early China’, in Li Feng and D. P. Banner (ed.) Writing & Literacy in Early China: Studies from the Columbia Early China Seminar, Seattle: University of Washington Press: 339–69. Yü, Ying-shih  余英時 (1985) ‘Individualism and the Neo-Taoist movement in Wei-Chin China’, in D. Munro (ed.) Individualism and Holism: Studies in Confucian and Taoist Values, Ann Arbor: Center for Chinese Studies, University of Michigan: 121–55.

195

PART II

Topical studies

9 THE OLD CHINESE LANGUAGE

AXEL SCHUESSLERTHE OLD CHINESE LANGUAGE

Axel Schuessler

Old Chinese (OC, or Archaic Chinese, shànggǔ hànyǔ 上古漢語) is, in a broad sense, the language of texts and written material dating from the beginning of writing on Shāng dynasty oracle bones and bronze inscriptions (ca. 1200 bc) down to the beginning of the Qín-Hàn period (221 bc). More specifically, it is the language whose phonology and morphology are reconstructable for the early centuries of the Zhōu period after 1000 bc. OC belongs, together with its modern relatives and descendants like Guānhuà (Standard Chinese = Mandarin), Yuè (Cantonese),Wú, Gàn, Xiāng, Hakka (Kèjiā), and Mǐn to the Sinitic branch of the Sino-Tibetan language family (which includes Tibetan, Burmese, Kuki-Chin and other branches; see Matisoff 2003). Claims of wider affiliations do not find general acceptance and would in any case add nothing insightful to our understanding of Chinese. The Shāng dynasty language was already some form of Chinese, as the choice of graphs for homophones indicates. The only language in the area in which ‘mother’ and ‘don’t’ (母毋), ‘basket’ and ‘his, their’ and ‘probably’ (其), ‘head’ and ‘road’ (首道), and ‘axe handle’ and ‘carry’ and ‘river’ (柯何河) are all sets of (near) homophones can only be OC. (Compare: in how many languages are the words for ‘eye’, ‘I’, and ‘aye’ homophones?). OC was a monosyllabic language, notwithstanding some disyllabic words, mostly reduplications of one type or other, some animal names, and compounds. (For a general introduction and overview, see Norman 1988.) Some scholars hypothesize that some OC words had sesquisyllables, i.e. unstressed pre-syllables that left no trace in later Chinese. Unlike later Sinitic languages, OC had word initial consonant clusters like *kr-, *sm-, and voiceless sonorants. It was probably not yet a tonal language. OC had two laterals, *r and *l. By the Hàn period, especially its last centuries, the language had undergone significant changes so that it was already close to Middle Chinese: only one lateral /l/, probably no consonant clusters, possibly phonemic tones; however, even the language of late Hàn still had preserved an OC final -s.

Old Chinese phonology Since OC was not recorded in an alphabetic script, its sounds can only be indirectly inferred from (1) Middle Chinese (MC; = Karlgren’s ‘Ancient Chinese’, zhōnggǔ hànyǔ 中古漢語), which is more accurately called the Qièyùn System (QYS) because it is based on the rhyme dictionary Qièyùn 切韻 (ad 601) and the Sòng period rhyme tables like the Yùnjìng 韻鏡 (ad 199

Axel Schuessler

1161) (in this context, ‘rhyme’ is also spelt ‘rime’, following Chao Yuen Ren). (2) Phonetic series (xiéshēng series) that are sets of graphs which share a graphic element as phonetic. (3) Rhymes in the Shījīng (Book of Songs) and other poetry; these tend to agree with the phonetic series (see Baxter 1992 for the history of OC reconstruction). OC is therefore not a ‘reconstruction’ in the linguistic sense (comparing three or more related languages in order to arrive at a common proto-form) but philological interpretation of graphs and other material. Therefore OC reconstruction is to some extent a matter of judgement that depends on methods, assumptions, interpretations of the material, and on the cultural background and native language of the researcher and any other languages he may be familiar with. Therefore investigators working with this same material often arrive at different reconstructed OC forms. Note the OC phonological interpretations of the copula wéi (MC jiwi) 隹維惟 ‘to be, it is/ was’ > ‘only’ 唯: *di̯ wər Karlgren 1957 Li Fang-kuei 1974 *rəd *lul Schuessler 1974 *iuəi (?) Wang Li 1980 *ljuəj Schuessler 1987 Baxter 1992 *wjij *t(ə)-wij Sagart 1999 *k-lul Pān Wùyún 2000 etc. (cf. Schuessler 1974) *ɢʷi Zhèngzhāng 2003 Schuessler 2009 (cf. Baxter 1992) *wi *ɢʷij Baxter and Sagart 2014 (cf. Zhèngzhāng 2003) The choice of OC reconstruction is crucial for further work in text analysis, morphology, and comparative studies, because arguments and results could drift far apart. For example, Li compared his *rəd to Tibetan red-pa ‘to be’, while *wi is obviously cognate to Tibeto-Burman *wəy/*wi ‘to be’. Other versions of OC would show no connexion with other languages in the area. The point of departure for venturing back into the pre-Hàn language is Middle Chinese, as reflected in the rhyme dictionary Qièyùn (QY) and the Sòng period rhyme tables which systematize the QY. This work arranges graphs in phonological categories, first according to rhymes (yùn 韻); then subsections sort the graphs by MC tones: píng 平 ‘even tone’, shǎng 上 ‘rising tone’, qù 去 ‘departing tone’, rù 入 ‘entering tone’. In transcriptions, tone píng is not indicated, as it represents a normal unmarked syllable, nor is rù, which is the automatic tone of short-stopped syllables in final -k, -t, -p. The shǎng and qù tones are widely believed to derive from a segmental feature of the OC syllable. Karlgren indicated the MC shǎng and qù tones by a colon (:) and hyphen (-) respectively, thus 古 gǔ/kuo: and 故 gù/kuo- (here we will write QYS kuo´ and kuo’). The QY categories have by themselves no phonetic content. All the QY suggests is that, for example, the graphs 古五土 etc. share the same rhyme, whatever that might have been. To fill these phonologically empty categories, Karlgren compared reading pronunciations of graphs in several dialects, in Sino-Japanese, Sino-Vietnamese, and Sino-Korean, and arrived at their MC forms. For example, ‘antiquity’ 古 is gǔ in Standard Chinese and ko (or something similar) in other Sinitic languages; Karlgren concluded therefore that its MC pronunciation was kuo:. The much-debated QY does not represent a particular dialect of Middle Chinese or an actual living language. It was a list of reading pronunciations for the graphs in OC and other 200

The Old Chinese language

early texts, created by a committee that included in its work pronunciations from both northern and southern China, hence the minute phonological distinctions in the QY (a~â~ɐ~å etc.) that are thought by some scholars to be artificial. But as ever more detailed material on Sinitic languages and dialects is recorded, ever more QY distinctions seem to be validated. Occasional odd syllables that seem to be contrary to the overall QY system are reminders that there was more to the Chinese language(s) around ad 600 than what could be fitted into the framework of that system, e.g. 地 di’ (only QYS die or źji would be possible) and 冷 lɐŋ´ (no ɐ after l) are unique unorthodox syllables. Because the QY was a reading list for graphs encountered in earlier literature, and since it was certainly not a descriptive study of any Sinitic dialect, a term ‘Qieyun System’ (QYS) is more appropriate than ‘MC’ (Norman and Coblin 1995). From this it follows that the QYS cannot have been the ancestor of modern Sinitic languages (‘dialects’). Such a common ancestor would be Common Dialectal Chinese as reconstructed by Norman (2006, 2014), who used the traditional method of historical reconstruction. The QY has always played a dual role: (1) for going back into OC; and (2) for looking forward and explaining the development of modern dialects, i.e. being mistaken for an ancestral MC language. The latter has been questioned (also, modern dialects have not a few words that are not in the QY), the usefulness for the former is better grounded (the QY includes the graphs of the ancient classics and earlier texts). The QY is, after all, all we have as a basis from which to explore earlier stages of the language. Once the QYS had been established, the phonological categories of OC need to be determined with the help of graphs and rhymes, work that has largely been accomplished by Qīng dynasty scholars. The majority of graphs are composites of two or more elements, one of which serves, in the majority of items, as a phonetic element. Graphs with the same phonetic element constitute a ‘phonetic series’ or xiéshēng series; graphs within a series must originally in OC have been close in sound, with similar or identical rhymes and similar initial consonants (e.g. 且 qiě/ QYS tshja´, 祖 zǔ/tsuo´, 沮 jù/dzjwo´, 柤 zhā/tṣa, 助 zhù/dẓjwo’, all of the type OC *TSa). Rhymes in poetry, most prominently the Shījīng, agree closely with the categories revealed by the xiéshēng series. These phonetically unknown (empty) OC categories need to be reconciled (filled) with the phonemically reconstituted QYS forms. Usually, each OC rhyme category includes three or four QYS (and modern) rhymes, note qiě 且, just mentioned above. These four QYS ‘Divisions’ or ‘grades’ are: Div. I with simple QYS vowel or diphthong (gē 歌 QYS kâ); Div. IV is the front vowel counterpart to Div. I, with a medial -i- in the Sòng rhyme tables (jī 雞 QYS kiei); Div. II has a special vowel timbre (jiā 加 QYS ka, i.e. a vs. â); Div. III has a QYS medial ‘yod’ -j- (qí 奇 QYS gje). These divisions, and with them all syllables, belong to one of two fundamentally different types: Pulleyblank’s type B consists of all syllables of Div. III with the medial -j-, type A comprises all other Divisions with non-yod syllables. Although syllables of these two types had identical Old Chinese rhymes and similar or even identical initials, their evolution to Middle Chinese and modern Sinitic languages can be dramatically different. What points to an OC syllable *lə of type B, as in 台 ‘I, my’, became QYS jiɨ > yí, but *lə of type A 苔 ‘moss’ became QYS dậi > tái. This looks like Austroasiatic-type vowel warping, via intermediate jɨə and dɑə, respectively, where in syllable type B the vocalic onset has become high (*-ə >-ɨə, then eventually -i), in type A low (*-ə >-aə, then eventually -ai) (Schuessler 2006). Type B caused widespread palatalization of initial consonants (*t > tj- > QYS tśj-, *l- > j-, etc.); type A prevented this (*t > QYS t-). In type A words, OC *l- was delateralized to d-; OC *g- became a fricative ɣ-. What was presumably an s-prefix became a QYS voiced zj when preceding *l, *j, and perhaps *w in type B syllables (e.g. sì 食 in example 20 in the Morphology section of this chapter). 201

Axel Schuessler

Why many Old Chinese reconstructions? The OC rhyme *-a may serve as an illustration of why researchers arrive at different OC versions while using the same material. (This rhyme is relatively simple to reconstruct; others lead to more discussions and divergences.) An OC rhyme category is labeled with a typical member syllable; three OC categories are candidates for having ended in *-a: yú 魚 duó 鐸

QYS ŋjwo (Table 9.1A) includes QYS rhymes -uo, -jwo, -a, -ja QYS dâk (Table 9.1B) includes QYS rhymes -âk, -ak, -jak, -jäk, plus some yú rhymes gē 歌 or 戈 QYS kâ (Table 9.2A, 9.2B) includes QYS rhymes -â, -a, -je, -ja The first column in Tables 9.1A–B and 9.2A–B names the QYS Division, the second provides graph, QYS (Karlgren’s as amended by Li Fang-kuei, i.e. simplified and regularized), and gloss; then follow columns with OC reconstructions by a sample of investigators: Karlgren 1957, Wang Li, Li Fang-kuei; Baxter 1992; Baxter and Sagart 2014 (OC forms are from their 2011 website when missing in B&S 2014), Pān Wùyún (Zhèngzhāng’s OC is similar), and Schuessler 2009 ‘Minimal Old Chinese’ (OCM). The first group of rhymes in open syllables has the QYS vowels ‘a’, ‘o’ and ‘u’ (Table 9.1A), the second has QYS rhymes -â, -ja and -je (Table 9.2). One of these must represent OC *-a. The ‘a’ is dominant in Table 9.2A, the so-called gē category, while back vowels dominate the yú Table 9.1A  魚部 yúbù *-a, and Table 9.1B  鐸部 duóbù *-ak A I

QYS

古 kuo´ antiquity 孤 kuo alone 布 puo’ cloth II -a 家 ka family 瓜 kwa melon III 野 jiaᴮ wilderness 居 kjwo dwell 魚 ngjwo fish 瞿 kju’ anxious 無 mju not have B 鐸部 duóbù *-ak I 暮 muo’ evening 莫 mâk none 惡 ʔuo’ hate 惡 ʔâk bad II -a 亞 ʔaᴴ second III 借 tsja’ borrow 借 tsjäk 庶 śjwo’ all

Karlgren

Wáng Lì

Lǐ F-k.

B. ’92

B&S ’14/11

Pan

OCM

*ko *kwo *pwo

*ka *kua *pua

*kagx *kwag *pagh

*kaʔ *kʷa *pas

*kˤaʔ *kʷˤa *pˤa-s

*kaaʔ *kʷaa *paas

*kâʔ *kʷâ *pâh

*kå *kwå *di̯ å

*kea *koa *jya

*krag *kwrag *ragx

*kra *kʷra *ljAʔ

*kˤra *kʷˤra *lAʔ

*kraa *kʷraa *laʔ

*krâ *kʷrâ *laʔ

*ki̯ o *ngi̯ o *ki̯ wo *mi̯ wo

*kia *ngia *kiua *miua

*kjag *ngjag *kwjagh *mjag

*kja *ngja *kʷas *mja

*k(r)a *[r.ŋ]a ? *ma

*ka *ŋgla *kʷas *ma

*ka *ŋa *kʷah *ma

*mâg *mâk *ʔâg *ʔâk

*mak *mak *ak *ak

*magh *mak *ʔagh *ʔak

*maks *mak *ʔaks *ʔak

? *mˤak *ʔˤak-s *ʔˤak

*maags *maag *qaags *qaag

*mâkh *mâk *ʔâkh *ʔâk

*ʔăg *tsi̯ ăg *tsi̯ ăk *śi̯ ag

*eak *tsyak *tsyak *sjiak

*ʔragh *tsjiagh *tsjiak *sthjagh

*ʔraks *tsjAks *tsjAk *stjaks

*ʔˤrak-s *[ts]Ak-s *[ts]Ak *s-tak-s

*qraags? –*skag

*ʔrâkh *tsakh *tsak *lha(k)h?

202

*qhljags

The Old Chinese language Table 9.2  歌部 or 戈部 gēbù *-ai A

QYS

歌 kâ song 磨 muâ grind 過 kuâ’ pass over II -a 加 ka add 瓦 ngwa´ tile III 蛇 dźja snake 儀 ngi̯ ie dignity B I 个 kâ’ item 果 kuâ´ fruit II -a 踝 ɣwa´ ankle I

Karlgren Wáng Lì Lǐ Fang-k. Baxter 92 B&S ’14/11

Pan

OCM

*kâ *mwâ *kwâ

*kai *mai *kuai

*kar *mar *kwarh

*[k]ˤaj *mˤaj *kʷˤaj-s

*klaal *maal *klools

*kâi *mâi *kʷâih

*ka *ngwa

*keai *ngoai

*krar *kraj *ngwrarx *ngʷrajʔ

*kˤraj *C.ŋʷˤra[j]ʔ

*kraal *krâi *ŋʷraalʔ *ŋʷrâiʔ/*ŋrôiʔ

*d̑ ʻi̯ a

*djyai

*djar

*LjAj

*Cə.lAj

*ɢljal

*m-lai

*ngia

*ngiai

*ngjar

*ng(r)jaj

*ŋ(r)aj

*ŋral

*ŋai

*kâr *klwâr

*kai *kuai

*karh *kwarx

*kajs ? ? -ajʔ

*kˤa[r]-s *[k]ˤo[r]ʔ

*kaals *kloolʔ

*kâih *kʷâiʔ/*kôiʔ

*gwrarx

*gʷrajʔ ? *m-kˤo[r] *ɡroolʔ ʔ

*gʻlwar *goai

*kaj *maj *kʷajs

*gʷrâiʔ/*grôiʔ

group (Table 9.1A). Karlgren concluded that the gē category (Table 9.2) reflects original OC *-â. The yú category (Table 9.1) must be more back, hence his OC *-o. He reconstructed some words from the gē group with final *-âr, because they interchange in rhymes and/or phonetic series or word families with words in *-n (Table 9.2B).The QYS vowels of the several Divisions, and even within them, differed within an OC rhyme category; therefore Karlgren assumed that the vowels in the four Divisions (in the present examples only I, II, and III) reflect subtle timbre distinctions or different medials within an OC category, which still allowed rhyming. However, scholars have since concluded that the OC vowel in the yú category (Table 9.1A) must have been *-a. In phonetic series and word derivations, the QYS rhymes interchange occasionally with QYS -ak, which, all agree, reflects OC *-ak. In fact, exactly the same set of QYS rhymes from the yú group (Table 9.1A) are found also in the OC duó group (Table 9.1B), except that the latter is dominated by *-ak. Karlgren’s *-o and *-ak are difficult to imagine to be phonetically compatible. To explain the xiéshēng connexions between open vowels and *-ak in Table 9.1B, Karlgren concluded that there must have been a final consonant compatible with k, but weaker so that it disappeared by MC. Therefore he reconstructed not *-o but *-ag for those yú-like rhymes that interchange with *-ak, resulting in two OC sources for MC -a (*-å and *-ăg, as well as *-a) and MC -uo (*-o and *-âg), etc. But Li Fang-kuei reconstructed all yú-like graphs with OC *-ag across the board (Table 9.1A and 9.1B). Since the OC yú vowel was *-a, the gē rhymes (Table 9.2) must represent something else. Karlgren postulated final *-a, but also *-ar in cases where there were occasional rhyme and xiéshēng contacts with QYS final n (Table 9.2B). Li Fang-kuei extended Karlgren’s final *-r to all words in the gē rhyme group. This was theoretical conjecture. The sound of the final could just as well have been l, or i, or something else, or instead some final n might have derived from r (so, for example, Baxter and Sagart 2014). Other investigators like Wang Li 1980, and Baxter 1992 settled on *-ai for all words in the gē group. The rhyme *-ai fills a gap in their systems 203

Axel Schuessler

(that includes *-au and *-əi); and most interesting, some southern dialects have the diphthong -ai in archaic word forms of the gē group (e.g. wǒ 我 in Hakka ŋai). So -ai is the most plausible rhyme for the gē category – whatever earlier forms it might have derived from; some words have Tibetan cognates in -al (Schuessler 1974). Investigators since Karlgren have aimed for a simpler, more phonemic OC system. For example, Chao Yuen Ren pointed out that the *w or *u glide after labial initials in the Karlgren and Wang Li columns are not phonemic, since there are no syllables without this glide: Karlgren has only 布 *pwo and 無 mi̯ wo, but no *po and *mi̯ o; similarly Wang Li. In working out phonemic systems, Li Fang-kuei and others have eliminated such non-phonemic features; Li wrote *pagh and *mjag (the labial initial alone would explain the later rounding to QYS o and u).These back glides occur not just after labial but also after guttural initials. Pulleyblank’s and Li Fang-kuei’s solution was to assume labiovelar consonants, *kʷ- etc., which are found in many languages. Thus the QYS rounding of the vowel is accounted for exclusively by the initial consonants.This required a new interpretation of QYS syllables with acute initials (initial consonants other than ‘grave’ K and P) + medial w/u + dental finals, like duàn 斷 QYS tuân’ ‘cut’ and lún 輪 QYS luən ‘wheel’. There are no syllables of the type *TwaK, only *ToK, and no syllables of the type *ToT, only TuaT, in Karlgren’s and Li’s system. Thus such syllable types are in complementary distribution: *TuaT *ToK (no *ToT – no *TuaK) Following Yakhontov, the consensus is that here the QYS vocalisms derive from OC syllable types *ToT and *TuT (a gap in Karlgren’s and Li’s system), hence OC 斷 *tôns, 輪 *run. Div. II syllables had a more central QYS vowel timbre that eventually palatalized a preceding velar in some dialects (家 QYS ka ‘family’ > jiā). Karlgren’s solution was typical for a phonetician: his yú rhyme category had the basic vowel o, but the QYS had the Div. II vowel a. So he settled on a compromise å for words like 家 OC *kå and 瓜 OC *kwå ‘gourd’. As elsewhere, Wang Li moved the distinction to a medial glide in order to retain the uniform vowel a, the glide was e in ordinary (non-labialized) kāikǒu syllables (家 *kea), o in labialized hékǒu syllables (瓜 *koa). Eventually, internal reconstruction has led to the conclusion that QYS Div. II derived from medial *r for several reasons.The QYS initial ṣ- (also tṣ, dẓ) occurs only in connexion with Div. II and III syllables (e.g. 山 ṣan, never Div. I ˣṣân); then, ṣ- often interchanges with MC initial l in phonetic series and word families (shǐ 史 QYS ṣɨ´ vs. lì 吏 ljɨ’); and Div. II never occurs after initial l (one or two instances are striking exceptions: QYS Div. II 冷 lɐŋ´). Consequently, QYS retroflex sibilants and consonants, and by extension all Div. II syllables, must reflect traces of OC post-initial r. Therefore, since Yakhontov, Pulleyblank (1962, 1963), and Li Fang-kuei (1974), Div. II is reconstructed with OC r-clusters (家 *kra, 加 *krai, 山 *sran). Since OC medial *r now turns out to be responsible for QYS Div. II as well as retroflex consonants (QYS ṭ, ṣ < *tr-, *sr- etc.), these latter did not exist in OC. This OC medial *r solution, which is almost universally embraced, is a textbook example of successful internal reconstruction, guided by the principles of parsimony and naturalness (cognate Tibeto-Burman languages also have syllables with medial r). Now OC rhyme categories could be reconstructed with a single vowel, e.g. 姑 Div. I *ka, 家 Div. II *kra, 居 Div. III *kja. A recent consensus eliminates the medial -j- (type B = Div. III). It seems unnatural for a language to have slightly more than half of its vocabulary containing this medial glide (in Tibeto-Burman languages it is not nearly as ubiquitous, hence it seems to be a Sinitic innovation); conversely, type A with its delateralizing of *l (tái 苔 < *lə), etc., suggests 204

The Old Chinese language

the presence of a feature that blocked later palatalization. Opinions on this feature vary. Expanding on a suggestion by Norman (1994), Baxter and Sagart (2014), for example, believe that the initial consonants of type A syllables were pharyngealized, e.g. gū 姑 kuo < *kˤa ‘aunt’ (Div. I = type A) vs. jū 居 QYS kjwo < *ka ‘dwell’ (Div. III = type B). Others assume a distinction in vowel length; still others remain undecided and use a neutral symbol (e.g. OCM 居 *ka vs. 姑 *kâ).

Tones The rationale for Karlgren’s and others’ final -g, -d, and -b has already been mentioned (Table 9.1B). A voicing contrast in final consonants is alien to languages in the area.The problem for early investigators was how to separate, from the many MC open syllables, rhyme categories and xiéshēng series that had open OC syllables from those in Karlgren’s/Li’s final voiced consonants that have rhyme contacts with voiceless final stops. In the illustrations in Tables 9.1A (*-a) and 9.1B (*-ak group), Karlgren made a distinction and set up the yú group without a final -g (yú 魚 ŋjwo < *ŋjo), reserving the *-g for the duó group (mù 暮 muo’ < *mâg). In his system, many (but not all) MC open syllables derive exclusively from closed ones with OC final -g or -d, such as *-əg (zhī 之 category; there is no rhyme *-ə in his systems). Li Fang-kuei eliminated the problem by assuming only closed syllables for OC; these could only end in a voiceless, voiced, or nasal consonant or r, thus reconstructing final voiced consonants in 魚 *ngjag (Table 9.1A) just as in 暮 *magh (Table 9.1B). Karlgren excluded the MC tones from his considerations for OC. Wang Li also believed that information about early tones is so uncertain that he ignored them. He followed the Qīng scholar Duàn Yùcái by assuming that the qùshēng was recent and had no equivalent in OC. Therefore he made no distinction between Karlgren’s *-ag and *-ak, and final stop consonants and MC syllables rhyming with them in general, so that he wrote all of Karlgren’s *-ag as *-ak etc. (see the Wang Li column in Table 9.1B), musing that perhaps an OC long vowel might be responsible for the later loss of the final consonant in MC qùshēng. However, Haudricourt, Pulleyblank and others concluded that the qùshēng seems to have resulted from a lost final h or (ultimately) s. Eventually it was also shown that some dialects have a final glottal stop (or glottalization) where MC correlates have the shǎngshēng (Mei 1970). Thus QYS tones derived from OC segmental phonemes. Li Fang-kuei left the OC nature of later tones open but marked in his reconstruction later shǎngshēng with final *-x, qùshēng with final *-h. Baxter 1992 was among those who wrote OC *-ʔ and *-s outright as sources of later MC tones (syllables in píngshēng and rùshēng, i.e. in -k, -t, -p, are considered unmarked and ‘toneless’; consult the tables for illustrations. In OCM, I reserved final *-s for syllables which ended in *-ts or *-s in OC, while elsewhere I write Li’s -h because there is no evidence that other syllables ended, phonetically, in an -s in OC, while the s was still preserved during the Hàn period, as foreign transcriptions prove). When taking account of tones in the reconstruction of OC, it became apparent that only certain MC open syllables in qùshēng have rhyme and xiéshēng contacts with stop consonants. Since qùshēng can play a morphological role, just as the s-suffix in Tibetan, the explanation for these interchanges was one of *-k with *-ks (OCM *-kh, Li *-gh), *-t with *-ts (including *-ts QYS ji-. The reconstruction with *l- eliminated these voiced stops, reduced the OC phonemic inventory to a more plausible three-way manner system (t, th, d – rather than Karlgren’s t, tʻ, d, dʻ), and assumed the two laterals *r (> QYS l-) and *l-, as in the related Tibeto-Burman as well as other languages in the area (Tai). However, some compound graphs fall outside obvious patterns of phonetic composition. Here everything hinges on an investigator’s assumptions about the origin and nature of the Chinese script. There are two lines of interpretation, with gradations. The question is: did the developers of the script create also compound graphs that were semantic (as Chinese traditions teach), or were all compounds phonetic? 206

The Old Chinese language

According to this latter assumption, (almost) all compound graphs were originally phonetic in nature, even where this is no longer obvious by MC. This kind of interpretation attempts to reconcile phonetically incompatible graphic elements like gōng 公 QYS kuŋ ‘uncle, prince’ in zhōng 妐 QYS tśjwoŋ ‘father-in-law’ (QYS tś and k don’t mix in phonetic series) in one of two ways. One way supposes that a graphic element once had two or more additional obsolete readings (multivalence hypothesis). For example, zǐ 子 *tsjəgx (Li Fang-kuei’s OC) supposedly had an obsolete reading *grjəkw which explains its alleged phonetic role in hào 好 (Li *həgws) and some other graphs (Boltz 1994: 110ff). The other way attempts to reconcile incompatible QYS phonology by hypothesizing a common OC denominator from which the disparate MC readings are claimed to have derived, which effects the reconstructions of complex OC initial consonant clusters and syllables. Thus Baxter and Sagart reconstruct 妐 *t-qoŋ (based on 公 *C-qˤoŋ) (2014: 57), or set up wǔ 午 QYS ŋjwo´ ‘cyclical sign’ as OC *[m].qʰˤaʔ, and chǔ 杵 QYS tśhjwo´ ‘pestle’ as OC *t.qʰaʔ (2014: 129), with purely hypothetical initial configurations. Alternatively, the development of writing can be seen as a pragmatic process, where phonetic loans and semantic composites both, or in combination, were used to convey in writing a message in a way that a recipient could understand. Consequently, the OC writing is, like any writing system, an imperfect reflection of the spoken language and full of inconsistencies – it is not a writing system devised by a committee of modern linguists. It has generally been accepted, from Xǔ Shèn’s Shuōwén on down, that semantic compounds (huìyì 會意) are one of the six types of graphs (liù shū 六書). For example hǎo 好 ‘be good’ consists of nü 女 ‘woman’ + zǐ 子 ‘child’, neither of which bears any phonetic resemblance to hǎo. Similarly, the rhyme and meaning of 公 *klôŋ are sufficient to suggest that 妐 refers to *toŋ ‘father-in-law’, and the exact phonetic relationship (if any) of chǔ 杵 with wǔ 午 remains a matter of speculation. These are some of the approaches and difficulties about the reconstruction of OC. Nevertheless, there is some consensus that, beside the six vowels already mentioned, OC probably had the following consonants: Labials p ph b m ʔ h (x) Velars k kh g ŋ Laryngeals Labiovelars kʷ khʷ gʷ ŋʷ th d n Laterals r l Dental stops t Affricates ts tsh dz s (z?) j w Semivowels Initial consonants could form clusters with a following medial r, probably also l and j, perhaps also w, as well as with the prefix s-, and possibly others. A Sino-Tibetan medial w has disappeared at least by MC and is not reconstructable for OC (cf. tù 吐 *thâh > QYS thuo’ ‘to spit’ vs. Proto-Tibeto-Burman *twa; the u in QYS thuo’ is secondary and non-contrastive). Recent OC proposals postulate uvulars and even labio-uvulars (q, qʷ etc., e.g. Pān, Zhèngzhāng, Baxter & Sagart), but these are hypothetical assumptions, prompted by the desire to make the writing system, and with it the distribution of OC phonemes, more symmetrical (q, qh, ɢ, instead of ʔ and h). Other recent hypotheses lead to complex initial clusters and sesquisyllables with no supporting evidence in any known Sinitic language or philological source. The choice of a plausible OC reconstruction for one’s work is not easy. There tends to be an inverse correlation between speculative hypotheses, no matter how fascinating, and plausibility. The task of historical linguistics ought to be to identify the plausible among all the countless hypothetical possibilities (Sihler 2004: 225), with the help of time-tested criteria such as parsimony (Ockham’s Razor) and naturalness (as elaborated in Baxter 1992: 16–23). 207

Axel Schuessler

Returning to the multitude of proposals for the OC copula wéi 隹維 ‘to be’, let us take a different angle to reconstruction and start with the linguistic data we have: wéi Modern Standard Chinese Sin Sukchu (standard reading; Ming dynasty; Coblin 2001) vi [wi] ywi [yi] ‘Phags-pa Chinese (Mongol dynasty; Coblin 2007) (jiwi) Qièyùn 601 ad, not a phonetically attested form, but a reconstruction based on attested categories and dialects (Li1974) Old Northwest Chinese (400 ad; Coblin 1991) iui Common Dialectal Chinese (Norman 2006) wi2 wi Hàn Buddhist Transcriptional Dialect (Coblin 1982); transcribing Skt. vi OC ___ ? *wəy or *wi Tibeto-Burman ‘to be’ Word initial y, ji, or i (‘Phags-pa, Old Northwest Chinese, QYS) can be reconciled with Ming dynasty wi, common dialectal wi and Hàn period wi because the initial ji-, y-, i- is a Middle Chinese feature that corresponds to the Div. III medial j in syllables with other initial consonants. So the QYS ji- in jiwi is secondary. Down to Hàn times the word was wi; it probably could not have been astoundingly different earlier in OC; there are no clear phonological patterns to suggest otherwise. The large and disparate phonetic series 隹 suggests an OC *wi (or *wij if one insists). A complicating factor is that 隹 originally wrote a word zhuī, QYS tświ < *tui ‘a kind of dove’ (no textual example); therefore, some investigators assume some kind of dental or other initial consonant (note *di̯ wər, *rəd, *k-lul, *t(ə)-wij). Yet there is no systematic pattern for interchanges *t- ~ *w-; hence, this early Shāng period graphic choice is irregular, has unknown causes, ought to be set aside, and is at best a matter for speculation. OC *wi is not claimed on the basis of Tibeto-Burman; rather, Tibeto-Burman confirms it. In the end, any OC reconstruction needs to make sense in the larger picture of the SinoTibetan family to which it belongs. Cognate sets with related Tibeto-Burman languages can usually be identified already on the basis of the QYS (no need for OC). Thus QYS jiwi, Hàn wi is obviously to be connected with Tibeto-Burman *wəy or *wi ‘to be’ anyway. A plausible OC reconstruction cannot be out of line with these givens; the Sino-Tibetan word must have been *wi or something close to it. This excludes *di̯ wər, *rəd, *k-lul, *ɢʷij and similarly complex or un-Sinitic proposals. As usual, it seems that the mundane solution is the most plausible one.

Grammar Unlike Indo-European languages, OC had no grammatical inflection, no markings for gender, number, cases, persons, tenses, aspects, or moods. What determined the function and meaning of a word in a sentence was word class, word order, and grammatical words. Therefore much of the study of Chinese grammar is taken up by the study of these grammatical lexemes (see, for example, Pulleyblank 1995). OC has the word order Subject – Verb – [Indirect Object] – (Direct) Object, unlike TibetoBurman languages, but similar to languages in the area like Tai and Proto-Mon-Khmer. The subject need not be supplied as long as the context provides it: 1

飲鄉人酒 yìn xiāng rén jiǔ (let drink/country/people/wine) ‘[Nan Kuai] offered wine to drink to his country’s people’ (Zuǒzhuàn: Zhāo 12.8). 208

The Old Chinese language

There are some exceptions with preposed object pronouns in a negated sentence: 不吾見 bù wú jiàn ‘he does not see me’ (not the expected ˣbù jiàn wú); this is reminiscent of Tibeto-Burman. Chinese may properly be described as a topic-comment language, because any part of speech can be placed at the head of the sentence, like the object (patient) in this example: 2

戎狄是膺 Róng Dí shì yīng (Rong/Di/this/repress) ‘The Róng and Dí people, them [he] repressed’ (Shījīng 300,4; see Pulleyblank 1995: 70).

The modifier precedes the word modified, just as in English: gāo shān 高山 ‘high mountain’, wáng mìng 王命 ‘the king’s (royal) order’.The Shāng and early Zhōu language had the copula wéi 隹維惟 ‘to be’: 予惟小人 yú wéi xiǎo rén (I/to be/small/person) ‘I am an insignificant person’ (Shūjīng 27,9). After the loss of the copula by Mid-Zhōu, a predicative sentence was marked by the final particle yě 也: 我小人也 wǒ xiǎo rén yě (I/small/person/part.) ‘I am an insignificant person’. Word classification has engendered much discussion, because word class is not formally marked. Nevertheless, every word belongs inherently to one of several (notional) word classes, such as noun (quǎn 犬 ‘dog’), pronoun (wǒ 我 ‘I, me’) transitive verb (tr. jiàn 見 ‘see’), intransitive verb (intr. lái 來 ‘come’), stative verb (sv. hǎo 好 ‘be good’), copular verb (fēi 非 ‘is not’), auxiliary verb (kě 可 ‘can’), co-verb (yú 于 ‘[go] to’), particle (yě 也; bù 不 ‘not’). Verbs can unambiguously be identified because only they are negatable with bù ‘not’. The word class determines the grammatical role and meaning of a word in a sentence, not unlike English: ‘to run’ is inherently intr., as in ‘he runs fast’, but used transitively it has a causative meaning ‘he runs the car’, or it can be used as a noun: ‘this was a long run’. Some aspects of OC word classification are matters of debate because some types of words do not readily fit into a category, like abstract words which can occur as often as nouns as verbs, e.g. rén 仁 functions both as a noun ‘benevolence’, or as a verb ‘be benevolent’; however, syntax leaves no doubt as to which is intended in a given sentence. There is no argument that in a sentence quǎn shí ròu 犬食肉 ‘the dog eats the meat’, quǎn and ròu are nouns, shí ‘eat’ is the verb. A noun in a verbal position in a sentence means ‘to act as X’ or ‘treat like X’, thus 為臣而君 wéi chén ér jūn (do/minister/part./ruler) ‘to be a minister but act like a ruler . . .’ (Zuǒzhuàn, Xiāng 7/7, see Pulleyblank 1995: 26); or 吳王我 Wú wáng wǒ (Wú/king/me) ‘treat me the way the king of Wú was treated’ (Zuǒzhuàn, Dìng 10/7; see Pulleyblank 1995: 26). An intr. verb or stative verb followed by an object has either a causative or a putative meaning, as for instance yuǎn 遠 ‘be far’ (sv.): 3

不遠千里而來 bù yuǎn qiān lǐ ér lái (not/be far/thousand/miles/part./come) ‘You do not (make:) consider a thousand miles too far to come’ (Mèngzǐ 1,1; putative).

‘To come’ lái 來 is intransitive, but a following object makes the verb causative: 來之 lái zhī (come/it, them) ‘make them come’. A transitive verb without a direct object (expressed or implied) acquires either an intransitive meaning – 視之不見 shì zhī bù jiàn (look at/it/not/see tr.) ‘what he looks at he does not see (Lièzǐ 1, p. 3b) – or a passive meaning – 見於王 jiàn yú wáng (see tr./by/king) ‘he was (seen:) received by the king’ (Mèngzǐ 2B, 4), especially after certain auxiliary verbs: 可見 kě jiàn (can/ see tr.) ‘can be seen’. A multitude of grammatical words (particles, adverbs, co-verbs, auxiliary verbs, pronouns) constitute the field of OC grammar; these may indicate time, modality, aspect, plurality, et alia. 209

Axel Schuessler

In the texts some of these seem to be optional depending on context; we cannot know if the actual spoken language might have required them, like modern Sinitic languages. Here follow a few examples of such words: Negation. The particle bù 不 negates a verb; fēi 非 QYS pjwei, OC *pui or *pəi ‘is not’ is a fusion of bù 不 *bə + wéi 隹 *wi ‘to be’. The tr. verb wú 無 QYS mju, OC *ma ‘have not, there is no’ is the negative counterpart of yǒu 有 *wəʔ ‘have, there is’. Especially in the earlier literature occur other negatives that seem to be synonyms of wú, such as, among other rarer items, mǐ 靡 *maiʔ, wáng 亡 *maŋ, and wǎng 罔 *maŋʔ. The prohibitive ‘don’t’ wú 毋 *mə was later replaced by the common wú *ma. The verb wú 無 is related to Tibeto-Burman *ma ‘not’; southern Sinitic languages still use it instead of bù. Bù is a (northern) Chinese innovation. Particles. Sentence (or phrase) final yě 也 *laʔ has already been mentioned. Sentence final hú 乎 *ɦa changes a declarative sentence to an interrogative one: 信乎 xìn hú ‘is it true?’ (Mèngzǐ 5A, 9). The ‘pronominal substitute for the head of a noun phrase’ (Pulleyblank 1995) is zhě 者 *taʔ, as in 4

故失之者死, 得之者生 gù shī zhī zhě sǐ, dé zhī zhě shēng (therefore/lose/it/zhě/die // get/it/zhě/live) ‘Therefore he who loses it [i.e. the way of heaven] dies, he who gains it lives’ (Lǐjì, Lǐyùn 3).

Pronouns.Widely used is zhī 之 *tə, which is a general third-person/demonstrative pronoun in the earliest texts, e.g. 之子 zhī zǐ ‘this gentleman’ (Shījīng 229,7), zhī wǔ yuè 之五月 ‘this fifth month’ (Oracle Bone Inscriptions, Bǐngbiàn 98,7, Takashima). Later in classical texts zhī is restricted to two specialized functions: (1) it is the common object pronoun ‘him, her, it, them’, as in 孔子聞之 Kǒngzǐ wén zhī ‘Confucius heard of it’ (Zhuāngzǐ 6,63); (2) zhī functions as genitive marker, as in 彼之怒 bǐ zhī nù ‘the anger of these, their anger’ (Shūjīng 26,2). Co-verbs. In Shāng and early Zhōu texts yú 于 *wa can be a verb ‘to go’: 于以采蘋 yú yǐ cài pín ‘she goes to gather the pin plant’ (Shījīng 15,1); 王于伐楚伯 wáng yú fá Chǔ bó ‘the king went to attack the prince of Chǔ’ (Bronze Inscription, Chou Fa-kao #1377); then yú functioned like a general preposition ‘to, at, from’ etc.: 或降于阿 huò jiàng yú ē ‘some (cattle) descended from the slope’ (Shījīng 190,2). By the mid to late Zhōu period, the common ‘preposition’ yú 於 *ʔa takes its place. Verbs and auxiliary verbs express time, aspect, or mood. Jì 既 ‘to complete’ indicates completed action: 既克商二年 jì kè Shāng èr nián ‘two years after he had defeated Shāng. . . ’ (Shūjīng 26,1). Qí 其 *gə (perhaps originally ‘to anticipate, expect’) has often been called a ‘modal particle’; it expresses probability, ‘should, will, probably’: 王其罔害 wáng qí wǎng hài ‘the king will probably suffer no harm’ (Shūjīng 26,10).

Morphology OC morphology is limited to deriving new words from roots or other words. Words that are apparently somehow related can be grouped into word families (Karlgren 1933; also Wáng Lì 1982; Schuessler 2007), e.g. qiáng 強 *gaŋ ~ jìng 勁 *keŋs ~ gěng 梗 *krâŋʔ ‘strong’, where the functions of the QYS medial -j-, the difference in vowel (*a vs. *e) and the medial *-r- are not (yet) understood (theoretically dialect interference could also have played a role). But the meaning of some derivational morphemes is becoming clear. The most common derivational device is the QYS qùshēng, which is now widely believed to go back to an OC final *-s (it appears that phonetically it was realized as *-h after velars and vowels already in OC). Its function has been difficult to define. Trying to determine this by studying words with that tone in all of Shāng, Zhōu, and even Hàn dynasty literature indicates 210

The Old Chinese language

that it can derive practically any word from any other; Downer (1959) has come up with a long list of Indo-European-like word class changes for which this tone is supposedly responsible, and Baxter and Sagart (2014) partially replicate Downer’s list but admit that the meaning of this morpheme is mostly not understood. In order to identify its function, it is necessary to study the earliest material and to analyze those qùshēng words that contrast with closely related items without this feature, ideally minimal pairs.The picture that emerges is clear.The s-suffix has two major functions in early OC: (a) to increase valence (intr./s.v. > tr., tr. > ditr., causative), and (b) to form passives. These functions are not grammatical, because the derivatives are new words with their own unpredictable meanings and word class. For example: 5

hǎo 好 *hûʔ sv. ‘be good, fine’ as in 好也 hǎo yě (be good/particle) ‘it is good’ + *-s (or *-h) > hào 好 *hûh tr. ‘be fond of, love’, as in 好貨 hàu huò ‘be fond of material possessions’, or 好之 hào zhī ‘be fond of it’.

Note that this transitive verb has a word class and meaning which is different from the syntactical (and predictable) transitive use of intr. hǎo 好 in 好之 hǎo zhī ‘make/consider it good’ (see earlier). In the next example, ‘keep away from’ (< ‘make to be far away, put a distance between’) is not the same as the grammatical causative/putative of yuǎn 遠 in example (3): 6 yuǎn 遠 *wanʔ ‘be far’ + *-s > yuàn 遠 *wans ‘to keep at a distance, keep away from’, as in 君子遠庖廚也 jūn zǐ yuàn páo chú yě ‘a superior man keeps away from his butchering house and kitchen’ (Mèngzǐ 1A, 7). 7 yī 衣 *ʔəi n. ‘clothes’ + *-s > yì 衣 *ʔəis tr. ‘to wear’, as in 衣衣 yì yī (wear tr./clothes) ‘to wear clothes’ (Yì Zhōushū 37,9). Through word order, this verb can be shown to be also ditransitive/causative: 8

載衣之裼 zài yì zhī tì ‘then they dressed them (the babies) in wrappers’ (Shījīng 189,9).

When added to a transitive verb, the meaning becomes ditransitive and causative. Unlike the causatives of syntactic position, the s-suffix causatives tend to be permissive, in that the agent of the sentence ‘lets’ others do something, he does not ‘make’ (force) them to do it. 9

yǐn 飲 *ʔəmʔ tr. vb. ‘to drink’, as in 飲酒 yǐn jiǔ ‘drink wine’ +*-s > yìn 飲 *ʔəms ditr./caus. ‘let someone drink something’; see example (1).

This yǐn fits into a pattern where the simplex is introvert, marked with a final *ʔ (later shǎngshāng), the distransitive/causative derivative in *s is extrovert (drink wine oneself vs. let others drink it). This shows that *-ʔ is a morpheme in such pairs and many other words, and not always part of the word root. Other such pairs include 10 mǎi 買 *mrêʔ tr. ‘to buy something’ > mài 賣 *mrêh ditr. (‘let someone buy something’ =) ‘to sell something to someone’. 11 shì 視 *giʔ tr. ‘to look at something’ > shì 示 *gih ditr. (‘let someone look at something’ =) ‘show someone something’. 211

Axel Schuessler

The second function of the s-suffix that emerges when studying early OC (near) minimal pairs is the formation of passive derivatives. This is, again, not a grammatical passive (for this is expressed by word order) but creates new words with their own inherent word class. Thus from transitive wēi 威 *ʔui ‘to frighten’ is derived transitive wèi 畏 *ʔuis ‘to fear’, which is etymologically (not grammatically) passive ‘be frightened’, but is now a new tr. verb. These passive derivations are often nouns and can be paraphrased by ‘what has been X-ed’: 12 zhī 織 *tək ‘to weave’ tr.: 織布 zhī bù ‘to weave cloth’ (Mengzi 3A,4) +*-s > zhì 織 *təks (*təkh) n. ‘what has been woven’ (not: ‘what is weaving’) = ‘(elaborately) woven cloth’, as in: 士不衣織 shì bù yì zhì ‘an official does not wear (fancy) cloth’ (Lǐjì,Yùzǎo II,8). Written Tibetan has an exact parallel: ’thag-pa tr. ‘to weave’, thags n. ‘texture, web’ (< ‘something woven’, pf. pass.). In Tibetan, final -s marks the past/perfect tense, but is probably still the same Sino-Tibetan morpheme as the OC one under consideration. (Generally, morphemes often play different roles in different languages; for example, an Indo-European s-suffix attached to Greek verbs marks the aorist, in Latin perfect tense). 13 wén 聞 *mən tr. ‘to hear, hear about’: 聞汝眾言 wén rǔ zhòng yán ‘I have heard the words of you all’ (Shūjīng 10,2). – 我聞其聲 wǒ wén qí shēng ‘I hear his voice’ (Shījīng 199, 3) +*-s > wèn 聞 *məns intr. ‘be heard’ (pass.): 聲聞于外 shēng wèn yú wài ‘the [instruments’] sound is heard outside’ (Shījīng 229,5).Then ‘heard about’ > ‘renowned, famous, fame’. 14 zhāng 張 *traŋ tr. ‘to stretch’ + *s > zhàng, *traŋh (pass.) ‘bloated, conceited’, 張; ‘curtain’ 帳 (< something stretched [passive] – not something that is doing the stretching [active]). These meanings and functions cut across Indo-European-like word class categories. What combines zhì 織 *təks ‘cloth’ n., wèn 聞 *məns intr. ‘famous’, and wèi 畏 *ʔuis tr. ‘to fear’ is not word class, it is an etymological passive; similarly, yì 衣 *ʔəih tr. ‘to wear’ and mài 賣 *mrêh ditr. ‘to sell’ mark increase in valence, not a verb class. By Hàn time, many of the s-suffix derivations are nouns, but the s does not form nouns, it is that the passive or causative derivations happen to turn out to be inherently nouns. Of course one could prefer to cut the categories in different ways and define derivation of nouns as one of several functions of qùshēng; in Tibeto-Burman languages -s apparently forms nouns as well (among others; Matisoff 2003: 465ff.), but the alternative view here is more economical. Derivations with final *ʔ (later shǎngshēng) are introvert and active (not extrovert and passive; see examples 9–11): 15 zhāng 張 *traŋ tr. ‘to stretch’ + *-ʔ > zhǎng 張 *traŋʔ intr. ‘to grow tall; grown up’ (introvert, active). 16 dēng 登 *tə̂ŋ intr. ‘to ascend’ + *-ʔ > děng 等 *tə̂ŋʔ n. ‘step of stairs’ (< what is ascending, going up – active, i.e. not ‘what is ascended’). 212

The Old Chinese language

17 zhī 之 *tə ‘to go’ + *-ʔ > zhǐ 止 *təʔ n. ‘foot’ (< what is doing the going – active). Some common intransitive verbs that imply, like zhǎng 張, a change of state also have final *ʔ: sǐ 死*siʔ ‘to die’, qǐ 起 *khəʔ ‘to rise’, zuò 坐 *dzoiʔ ‘to sit’, and others. The s-suffix may perhaps originally have started out as a marker for extroversion, *-ʔ for introversion. Another group of words with final *ʔ are particles which occur at the end of a declarative sentence or phrase: zhě 者 *taʔ, yě 也 *laʔ, yǐ 矣 *ləʔ, yǐ 已 *ləʔ. In contrast, final interrogative particles and interrogative pronouns do not end in *ʔ, for instance hú 乎 *ɦâ (or *â or *gâ) final particle, hú 胡 *gâ ‘why, how’, hé 何 *gâi ‘what’. Some words have counterparts with final *ʔ which can stand at the end of a phrase or sentence, or can stand alone: bù 不 *pə + vb.‘not’, but fǒu 否 *pəʔ is sentence final ‘not, . . . or not’; wéi 隹維 *wi copula ‘is, it is’, standing alone wěi 唯 *wiʔ ‘yes’ (< ‘it is.’) (Lúnyǔ 4,15). Wú 吾 *ŋâ ‘I, my’ must always precede another word, as in 不吾見 bù wú jiàn (*pə ŋâ kêns) ‘he does not see me’, while wǒ 我 *ŋâiʔ ‘I, we’ can also stand alone or be the last word of a sentence or phrase (where it is by default an object); thus when changed to a positive statement, ‘he sees me’ must be 見我 jiàn wǒ (*kêns ŋâiʔ). In these cases *ʔ was originally not strictly a morpheme but a positionally conditioned feature marking the end of a declarative statement or part of speech (perhaps an abrupt end of articulation), whereas when followed by another word, by continued speech, or by continuation with an anticipated answer, words are employed that are not marked for termination (unless the word has an inherent final *ʔ, as rǔ 汝 *naʔ ‘you’). Voicing of the initial consonant forms words with decreased valence (tr. > intr., sv.); it is thus the opposite of the s-suffix functions, e.g. 18 jiàn 見 *kêns ‘to see’ tr. (the *-s is not a suffix) + voicing > xiàn 現 *gêns ‘to appear’ intr., e.g. 見於君 xiàn yú jūn ‘to appear before the prince’ (Mòzǐ 4 zhōng, 5). Xiàn is an intransitive verb, but not the intransitive counterpart to transitive jiàn ‘to see’ (see discussion of jiàn after example 3). 19 zhāng 張 *traŋ ‘to stretch’ tr. + voicing > cháng 長 *draŋ ‘be long’ sv. Derived from cháng 長 *draŋ is zhàng 長 *draŋh ‘measure of length’ (Lǐjì); here the suffix is neither valence increasing nor passive, but the late general purpose morpheme (the Lǐjì is a late Zhōu text). The QYS voiced initial is normally projected back into OC. But some proposals for OC (Baxter & Sagart 2014) derive them from an OC nasal prefix *N- (xiàn 現 < *N-kêns) or *m-, because some words have Common Mǐn ‘softened’ initials where Hmong-Mien (Miao-Yao) loans show pre-nasalization. On the other hand, in the related Tibeto-Burman languages, voicing is common among related words without apparent nasalization. A widely recognized prefix is the Sino-Tibetan causative *s- (Matisoff 2003: 100 ff.; Handel 2012), e.g. 20 shí 食 *mlək ‘to eat’ tr. + s- > sì 食 *zləkh (< *s-lək-s) ‘to feed’ tr. (ditr.?). 213

Axel Schuessler

This set has an exact Tibeto-Burman equivalent: Proto-Tibeto-Burman *m-lyak ‘to lick, eat’, and *s-lyak ‘feed’. The OC final *-s seems redundant; perhaps the Sino-Tibetan prefix was not understood anymore. In most words with a nasal (and also other sonorant) initial, the s-prefix was lost after devoicing the initial consonant (usually indicated by *h in transcriptions), as in xuè/miè 烕 *hmet ‘cause destruction’, derived from miè 滅 *met ‘destroy, extinguish’. Because both verbs are transitive and are often glossed ‘extinguish, destroy’, the causative meaning of xuè can be missed, yet it is clear in Shījīng Ode 192,8 (Mei Tsu-Lin 2012: 10): 燎之方揚 liáo zhī fāng yáng 寧或滅之 níng huò miè zhī 赫赫宗周 hè hè Zōng Zhōu 褒姒烕之 Bāo Sì miè (xuè) zhī

When the fire is just flaming high how can anyone extinguish it? (*met) The majestic Zōng Zhōu, (lady) Bāo Sì has caused its ruin (*hmet).

The essence of causativity is indirectness; someone extinguishing a fire with a bucket with water amounts to direct action. Bāo Sì did not destroy Zhōu herself with a battle ax or her army but was the cause for others to do it; therefore, *hmet is causative. These are some of the most obvious morphemes. Others include a nominalizing k-prefix (Sagart 1999) and a Sino-Tibetan nominalizing n-suffix (Matisoff 2003: 444 ff; Schuessler 2007: 75f.).

Evolution and dialects The language of the Shāng oracle-bone and Western (early) Zhōu bronze inscriptions must have undergone changes by the later Mid-Zhōu, Zhànguó, and the classical periods. The writing system allows only glimpses of these. Baxter (1992) points out that wén 聞 ‘to hear’ was earlier written with the phonetic hūn 昏, from which he deduces a change from *mun to *mən. Or possibly some words in QYS -i ended at some early period of OC in final *-r or *-l (Baxter and Sagart 2014). The OC dialect of the early classics Shījīng and Shūjīng, as well as of the phonetic series, had merged the final *ə and *o after labial initials, whereas the later QYS and modern Sinitic languages are descended from a strain that distinguished these vowels; therefore today’s mǔ 母*məʔ/*moʔ and měi 每 *məʔ belong to the same xiéshēng series (Baxter 1992: 465). A few grammatical changes and differences in vocabulary can be detected over time.We have already noted the copula wéi 隹 which disappeared by the end of the Western Zhou period; afterwards, a predicative sentence was marked by the final particle yě 也. Excavated late Zhōu documents show that in different regions of China other particles were used instead of yě. ‘All the’ was expressed with duō 多 ‘many’ in the earlier language (多士 duō shì ‘the many officials’ i.e. ‘all the officials, the officials’ pl.), later zhū 諸 ‘all’ is used (諸侯 zhū hóu ‘all the nobility’). ‘Head’ tóu 頭 replaced shǒu 首, which survives only in special expressions. Any language is apt to lose vocabulary over the course of a thousand years, yet OC is usually viewed as monolithic and static because of the conservative writing system (Handel 2013). An obsolete word for ‘blood’, huāng 衁 *hmâŋ, happens to occur once in the classical text Zuǒzhuàn and was therefore perpetuated in later dictionaries with sound glosses. The graph 眔 (‘eye’ with lines = drops below) was obviously created for ‘tears/to weep’, but no such word has survived. We should expect excavated texts to contain graphs that cannot be connected with words that are known from the received texts. 214

The Old Chinese language

Works cited Baxter, William H. III (1992) A Handbook of Old Chinese Phonology. Berlin, New York: Mouton de Gruyter. Baxter, William and Laurent Sagart (2014) Old Chinese. A New Reconstruction. Oxford, New York: Oxford University Press. Boltz, William G. (1994). The Origin and Early Development of the Chinese Writing System. New Haven, CT: The American Oriental Society. Chou Fa-kao (Zhōu Fǎgāo) 周法高 (1974). Jīnwén gǔlín 金文詁林. Hong Kong: Zhōngwén Dàxué chūbǎn. Coblin, W. South (1982) ‘Notes on the dialect of the Han Buddhist transcriptions.’ In: Proceedings of the International Conference on Sinology. Taipei: Academia Sinica. Coblin,W. South (1991) ‘Studies in old Northwest Chinese.’ Journal of Chinese Linguistics Monograph Series Number 4. Coblin, W. South (2001) ‘Phags-pa Chinese and the standard reading pronunciation of early ming: a comparative study.’ Language and Linguistics 2.2: 1–62. Coblin, W. South (2007) A Handbook of ‘Phags-pa Chinese. Honolulu: Hawai’i Univ. Press. Downer, G. B. (1959) ‘Derivation by tone-change in classical Chinese.’ British School of Oriental and African Studies 22: 258–290. Handel, Zev (2012) ‘Valence-changing prefixes and voicing alternation in old Chinese and Proto-SinoTibetan: reconstructing *s- and *N- Prefixes.’ Language and Linguistics 13.1: 61–82. Handel, Zev (2013) ‘Fuzzy word identification: a case study from the Oracle Bone inscriptions.’ Bulletin of Chinese Linguistics 7.2: 1–27. Karlgren, Bernhard (1933) ‘Word families in Chinese.’ Bulletin of the Museum of Far Eastern Antiquities 5: 9–120. Karlgren, Bernhard (1957) Grammata serica recensa. Stockholm: Museum of Far Eastern Antiquities. Li, Fang-kuei (1974–75) ‘Fang-kuei Li: studies on Archaic Chinese’, translated by Gilbert Mattos, Monumenta Serica 31: 219–287. – Translation of Li (1971) ‘Shànggǔ yīn yánjiù 上古音研究 (Studies on Archaic Chinese phonology)’, Tsing Hua J. of Chin. Stud. n.s. 9: 1–61. Matisoff, James A. (2003) Handbook of Proto-Tibeto-Burman. Berkeley: University of California Press. Mei, Tsu-Lin (1970) ‘Tone and prosody in Middle Chinese and the origin of the rising tone’, Harvard Journal of Asiatic Studies 30: 86–110. Mei, Tsu-Lin (2012) ‘Causative *s- and Nominalizing *-s in Old Chinese and Related Matters in ProtoSino-Tibetan’, Language and Linguistics 13.1: 1–28. Norman, Jerry (1988) Chinese. Cambridge, New York et alia: Cambridge University Press. Norman, Jerry (1994) ‘Pharyngealization in Early Chinese.’ Journal of the American Oriental Society 114: 397–408. Norman, Jerry (2006) ‘Common Dialectal Chinese.’ In: David Prager Branner, ed. The Chinese Rime Tables: Linguistic Philosophy and Historical-Comparative Phonology. Amsterdam/Philadelphia: John Benjamins, pp. 233–254. Norman, Jerry (2014) ‘A Model for Chinese Dialect Evolution.’ In Richard VanNess Simmons and Newell Ann Van Auken, eds. Studies in Chinese and Sino-Tibetan Linguistics: Dialect, Phonology, Transcription and Text. Taipei: Institute of Linguistics, Academia Sinica, pp. 1–26. Norman, Jerry and South Coblin (1995) ‘A new approach to Chinese historical linguistics’, Journal of the American Oriental Society 115.4: 576–584. Pān Wùyún 潘悟云 (2000) Hànyǔ lìshǐ yīnyùnxué 漢語歷史音韻學. Shanghai: Shanghai Educational Publishing House. Pulleyblank, E. G. (1962) ‘The consonantal system of Old Chinese.’ Asia Major n.s. 9: 58–144, 206–265. Pulleyblank, E. G. (1963) ‘An interpretation of the vowel systems of Old Chinese and Written Burmese.’ Asia Major n.s. 10: 200–221. Pulleyblank, E. G. (1995) Outline of Classical Chinese Grammar.Vancouver, bc: Ubc Press. Sagart, Laurent (1999) The Roots of Old Chinese. Amsterdam/Philadelphia: John Benjamins. Schuessler, Axel (1974) ‘Final -l in Archaic Chinese.’ Journal of Chinese Linguistics 2.1: 78–87. Schuessler, Axel (1987) Dictionary of Early Zhou Chinese. Honolulu: University of Hawai’i Press. Schuessler, Axel (2006) ‘The Qièyùn System “Divisions” as a Result of Vowel Warping.’ In David Prager Branner, ed. The Chinese Rime Tables: Linguistic Philosophy and Historical-Comparative Phonology, Amsterdam/Philadelphia: John Benjamins, pp. 83–96. Schuessler, Axel (2007) Abc Etymological Dictionary of Old Chinese. Honolulu: University of Hawai’i Press.

215

Axel Schuessler Schuessler, Axel (2009) Minimal Old Chinese and Later Han Chinese: A Companion to Grammata Serica Recensa. Honolulu: University of Hawai’i Press. Sihler, Andrew L. (2004). Review of “ Pre-Indo-European. By Winfred P. Lehmann. (Journal of IndoEuropean Studies Monograph Series, 41). Washington, DC: Institute for the Study of Man, 2002.” Diachronica 21.1: 214–226. Takashima, Ken-ichi (2010). Studies of Fascicle Three of Inscriptions from the Yin Ruins. 2 vols. Taipei: Academia Sinica. Wáng, Lì 王力 (1980) Hànyǔ shǐgǎo 漢語史搞. Beijing: Kēxué chūbǎnshè. Wáng, Lì 王力 (1982) Tóngyuán zìdiǎn 同源字典. Beijing: Shāngwù yìnshūguǎn. Zhèngzhāng Shàngfāng 鄭張尚芳 (2003) Shànggǔ yīnxì 上古音系 (Old Chinese Phonology). Shanghai: Shanghai Educational Publishing House.

216

10 EARLY CHINESE WRITING

LUO XINHUIEARLY CHINESE WRITING

Luo Xinhui; tr. Zachary Hershey and Paul R. Goldin

It is generally acknowledged that the earliest mature script in China appeared during the Shang dynasty (sixteenth–eleventh century bce). The Chinese script has undergone significant change since the earliest mature characters, namely the oracle-bone script, but has been in continuous use to the present day. Characters held a special significance throughout the development of Chinese history. Differences among the various dialects of China are substantial, yet the stability and unity of the script has boosted a sense of cultural unity. After the appearance of a mature script, writing saw extensive use in numerous fields. In areas such as religious belief, state management, the collection and dissemination of knowledge, and daily life, writing was employed to great effect. The vast number of classical texts stands as a testament to the great importance that the Chinese people have invested in writing, from the oracle-bone script and bronze inscriptions to the Confucian Classics, the twenty-four dynastic histories, and the Complete Writings of the Four Treasuries (Siku quanshu 四庫全書, 1773–82). The present chapter will address four main topics: (1) the formation and development of Chinese characters; (2) the types and characteristics of Chinese characters; (3) writing in early China; (4) general reflections.

The formation and development of Chinese characters Regarding the formation of Chinese characters, there are several interesting ancient sayings, such as: “In distant antiquity, society was managed by means of knotted rope. The sages of later generations replaced this with documents and inscriptions” (Zhouyi 周易, “Xici” commentary 繫辭傳). This means that the people of distant antiquity relied on knotting rope to record events, but later, sages invented writing. Ancient people believed that a person by the name of Cang Jie 蒼頡 invented writing: “The Yellow Emperor’s scribe, Cang Jie, observed the traces of the tracks of the birds and beasts, realized the possibility of differentiating the connections, and first created documents and inscriptions” (Preface to Shuowen jiezi 説文解字; also Han Feizi 韓 非子, “Wudu” 五蠹 chapter). Thus one explanation for the appearance of the script is that it arose from observation of nature. Before the appearance of a mature script, the ancients probably used drawings and symbols to record events or disseminate information (Qiu Xigui 2000: 2–12). Such symbols are attested in great numbers at Neolithic archaeological sites. Collections of more than ten such symbols 217

Luo Xinhui

arranged in a row have been discovered at Jiahu 賈湖, a site belonging to the Peiligang 裴李 崗 Culture (ca. 6000 bce). These are the earliest known incised symbols; some were incised into turtle plastrons and others into pottery or stone. The most representative incised symbols from the Neolithic period are the great numbers of symbols incised on pottery from the Yangshao 仰韶 Culture (ca. 4000 bce). These symbols have a relatively stable incision pattern, with the majority incised with a finely sharpened instrument on a black stripe bordering the opening of the vessel. Each vessel has but one incised symbol, and the rate of repeated symbols is relatively high (Wang Zhijun 1980). This is obviously a system of recording some kind of designated meaning. Among the symbols discovered from the Neolithic period, numerous scholars believe that the red symbols discovered at Taosi 陶寺 (ca. 2600–2000 bce) are characters. A pottery hu 壺 vessel unearthed at Taosi has a character written in vermilion on the front and rear surfaces, and some scholars believe that the character on the front is the character wen 文 (Xie Xigong 2007: 620–623). The recurrence of incised or painted symbols during the Neolithic period was a driving force in the ultimate emergence of the script. The oracle-bone script is the earliest known mature form of Chinese writing. It is considered mature because each character corresponds to a sememe in the language, and it is systematically ancestral to the modern writing system.The oracle-bone texts are writings from the Late Shang (ca. 1200 bce), but the creation of the script may date to around 1700 bce. Since the discovery of inscribed turtle plastrons and animal bones at Yinxu 殷墟 (near present-day Anyang 安陽, Henan 河南 Province) toward the end of the nineteenth century, around 150,000 plastrons and bones have been discovered, containing more than 4,500 distinct characters, of which more than 1,000 have been deciphered (Wang Yuxin and Yang Shengnan 1999). The vast majority of characters on plastrons and bone were incised; an exceedingly small number were painted with a brush. The artifacts recovered from Yinxu indicate that the tools used for these incisions were probably bronze knives or awls. Many oracle-bone characters are pictographs with the following main characteristics: 1

2

The orientation of the characters had not yet completely stabilized: they could be written forwards, backwards, or even on their side. For example, the character gui 龜 could be written forwards ( ) or sideways ( ). Each character had variants, and they could exist in complex and simple forms. As long as the important elements of the character were the same, even if there were additions, omissions, or general changes to basic structure, it was still the same character. For example, the character wu 物 could be written , but also, by shifting the components, as . Another example: the character yang 羊 could be written , or . Such characteristics suggest an early stage of systematization.

During the Shang and Western Zhou periods (eleventh century–771 bce), inscriptions were cast into bronze vessels. This is called jinwen 金文, literally “bronze writing,” in Chinese. Bronze vessels used in ritual sacrifices were often adorned with mysterious patterns, and inscriptions on bronze vessels revealed the authority of the text. Shang-era bronze inscriptions are relatively rare; mainly, they state the names of the invoked ancestors, as well as that of the vessel’s sponsor and his lineage. Long inscriptions came into fashion during the Western Zhou and were still in favor during the Spring and Autumn period (770–476 bce), but their popularity slowly declined. The characteristics of Western Zhou bronze inscriptions are: 1 Some characters’ pictorial characteristics are stronger, perhaps because characters were deployed with a tendency of revivifying antiquity in order to express the elegance and 218

Early Chinese writing

2

3

gravity of bronze vessels. Examples include the xiang 象 (elephant) in the Xiangzu Xin ding 象且（祖）辛鼎 inscription, written as an ornate elephant , and the niu 牛 (ox) on the Niu ding 牛鼎 vessel, held in the collections of the University of Pennsylvania Museum of Archaeology and Anthropology, which is written as an oxhead . During the Spring and Autumn and Warring States periods, however, lineage emblems (which were sometimes highly pictorial) disappeared. As compared to the oracle-bone script, bronze inscriptions took a step toward standardization, as in the case of the graphic component 彳, which, in the oracle-bone script, was written on the left or right side with equal probability. In bronze inscriptions, this component is found almost invariably on the left side. The structure of characters gradually moved toward unification, and the format and size of the characters slowly progressed toward order. For example, some characters that were written horizontally and occupied greater area in the oracle-bone script tended to be reoriented in a writing style more conducive to vertical formatting (Chen Weizhan and Tang Yuming 2009: 65).

During the Warring States period (475–221 bce), a great variety of writing styles emerged. In addition to inscriptions on bronze vessels, numerous texts were written on bamboo strips, jade stone, and seals.Texts have recorded that bamboo strip documents were already in use during the Shang era. The “Many Officers” (“Duoshi” 多士) section of the Book of Documents states: “Only the ancestors of the Yin people have ce and dian.” In the oracle-bone script, ce 册 is written and dian 典 is written , portraying bamboo strips strung together to form a manuscript. The references to a maker of ce-records named Ban 作冊般 on the Zuoce Ban yan 作冊般甗 and Zuoce Ban yuan 作冊般黿 vessels (e.g., Li Xueqin 2005; Zhu Fenghan 2005) are even stronger indications that certain late Shang scribes were tasked with producing texts on bamboo. Characters on bamboo strips were probably written with brushes. The Chu tombs at Changtaiguan 長台關 and Baoshan 包山 as well as the Qin tombs at Shuihudi 睡虎地 have all yielded such brushes, and ink blocks have been discovered in the Chu tombs at Jiudian 九店 in Hubei 湖 北 Province. Warring States writing exhibits regional differences, which led to the phenomenon of variant forms. For example, the character ma 馬 was written in Chu as , in the Three Jin area as , and in Qin as . In addition,Warring States characters also tend to use abbreviated forms, omitting strokes, such as 馬written as or wei 為 as . However, sometimes stylistic strokes were added, such as shang 上written as or da 大 as (Chen Weizhan and Tang Yuming 2009: 83). During the Spring and Autumn and Warring States periods, political movements and the exchange of goods were more complex. Seals were created in such situations as a proof of transaction. Xi 璽 seals from the Eastern Zhou, especially the Warring States, are vast in number; over 6,000 have been discovered to date. Ministerial xi seals from the Warring States were proof of political authority, and in the event of a transfer of position or death of a minister, the seals needed to be submitted to a higher authority (Cao Jinyan 1996: 6). Ministerial xi seals are of significant value for the study of ministerial organization and geography. The wording of the inscriptions on personal seals often was influenced by the many prominent philosophers of the Warring States, employing such phrases as zheng xing wu si 正行无私 (Act uprightly, without self-interest) or si yan jing shi 思言敬事 (Speak thoughtfully and serve respectfully). Auspicious phrases were also popular, such as churu da ji 出入大吉 (Great auspiciousness, whether coming or going) or yi you qianwan 宜有千万 (May you have thousands upon thousands). This type of seal was primarily worn on the person, probably for the purpose of warding off malevolent influences. Because of the limitations of the area of the seal, characters on the seals are greatly 219

Luo Xinhui

altered or omitted to save space. Shifting and interspersing components or reducing the number of strokes are common phenomena in seal-script characters. The First Emperor of Qin implemented the unification of the script, which had an extreme impact on the standardization of Chinese characters. (For studies of this complex process, see Galambos 2006 and Chen Zhaorong 2003). During the Warring States period, even though each state used the same basic system, the prevalence of variants was serious and unconducive to communication. The First Emperor unified the Chinese world and standardized the seal script, while moving toward simplification by creating the so-called clerical script (lishu 隸書), which is very close to the modern writing system.

The basic types and characteristics of Chinese characters Chinese characters are a logographic script with the basic characteristic of having meaning contained in their form, so that meaning and form are intimately related. Regarding the origin of the structure of Chinese characters, one Han-dynasty view invokes the “six categories” (liushu 六書): 古者，八歲入小學，故周官保氏掌養國子，教之六書，謂象形、象事、 象意、象聲、轉注、假借，造字之本也。 The ancients entered into early schooling by the age of eight. Thus, according to the Rites of Zhou, the Tutor took charge of raising the children of the state and taught them the six categories, which refer to images of the form, images of the matter, images of the significance, images of the sound, derivative cognates, and phonetic loans. These are the basis of the creation of characters. This is the earliest systematic theory of the structure of Chinese characters. “Images of the form” are pictographic characters: they copy the form of an object in order to express it in writing. For example, the oracle-bone form of xiang 象, meaning “elephant,” is , with a long trunk evoking an elephant. Yang 羊, meaning “sheep, ovicaprid,” is , like the curved horns of a ram. “Images of the matter” (sometimes called zhishi 指事, “indicating the matter”) are a more abstract way of creating characters: when it is impossible to use the likeness of the subject to create a character, an abstract symbol is employed instead. The distinguishing feature of such characters is the use of abstract symbols to provide a prompt by adding a semantic symbol to pictographs. For example, the characters shang 上 (above) and xia 下 (below) were written as and , respectively: the long horizontal line represents a line of reference, and the use of the slightly shorter line above or below it conveys the idea of “above” or “below.” Additionally, xiong 凶 (inauspicious) is composed of 凵, which represents a deep pit, with 乂 inside it, which represents a dangerous object inside the pit; these two components combine to represent “precariousness” or “danger.” “Images of the significance” (sometimes called huiyi 會意, “combining the significance”) are compound graphs that combine two or more simpler characters. For example, lei 泪 (“tear, teardrop”) is made up of shui 水 and mu 目, which mean “water” and “eye,” respectively. Also, xiu 休 (“to rest”) is composed of ren 人 and mu 木, meaning “person” and “tree,” respectively. A person leaning against a tree is “resting.” “Images of the sound” are characters that combine semantic and phonetic components. One part of the character represents the type of the object while another indicates the pronunciation. For example, zhu 珠 (pearl) is composed of yu 玉 (jade, gem) to invoke the idea of a precious 220

Early Chinese writing

stone and zhu 朱 (vermilion), which provides a phonetic hint. Another example is jiang 江 (river) which is composed of shui 水, to invoke the idea of water, and gong 工 (work), which provides a phonetic hint. (In Old Chinese, the phonetic similarity between 江, *kˤroŋ, and 工, *kˤoŋ, would have been clearer than in Modern Mandarin. Reconstructions are based on Baxter and Sagart 2014.) The underlying word pairs, such as “pearl” and “vermilion” or “river” and “work,” may or may not be truly cognate; the writing system discloses only that they are homophones or near homophones. “Derivative cognates” are characters that have identical radicals, are pronounced similarly, have similar meanings, and can be used to explain each other. For example, kao 考 (deceased father) and lao 老 (aged) are both classified under the radical 老; the explanation for kao in Shuowen jiezi is lao, and the one for lao is kao. Phonetic loans are characters that are used to write homophones or near homophones. Phonetic loan characters were introduced by harried scribes who needed to find graphs for more and more words. Some may also be attributable to the habits of specific periods or regions. Examples include tang 湯 written as dang 蕩 and pan 叛 written as pan 畔. Chinese characters combine shape, sound, and meaning into a single image. Because Chinese characters have a semantic component, they can convey more information in less space than purely phonetic writing systems. (A case in point: the original Chinese version of this chapter comprised a mere seven pages in typescript, whereas the English translation required twenty-three.)

The content of writing After the creation of the script, the practice of writing spread widely. It should be noted that the earliest forms of Chinese writing almost always bespeak a ritual context. Over time, writing expanded beyond its religious domain and was used to record people’s affairs, their history, and their thought. (The clearest account of this process is Li and Branner 2011; for a different viewpoint, see Bagley 2004 and Wang Haicheng 2014.) Unlike the ancient Egyptians and Mesopotamians, who inscribed stone tablets or recorded important events on the walls of temples, tombs, and palaces, the Shang mainly inscribed turtle plastrons and animal bones. Oracle-bone texts are predominantly records of divination, which was one method for humans to communicate with spirits. In order to receive revelations from deceased ancestors and other spirits, the King of Shang and the nobility conducted divinations. Oracle-bone inscriptions cover the nine reigns from Wuding 武丁 to Zhòu 紂. The topics include: (1) sacrifices and entreaties to ancestral and nature spirits; (2) meteorological phenomena such as wind, rain, thunder, and lightning; (3) harvests; (4) military actions, such as conflicts with other states; (5) royal affairs, such as the king’s hunts, illnesses, dreams, and the birth of heirs; (6) predicting the auspicious or inauspicious events of a given ten-day cycle (Chen Mengjia 1988: 636). Oracle-bone texts have numerous records of prognostications regarding agricultural harvests, such as shounian 受年 (receiving the harvest) and shouhe 受禾 (receiving grain). There are also many related to the Shang kings’ inquiries about illnesses, for example: “To be divined: Will the king not have a malady of the eye?” (Guo Moruo 1978–83: 456 recto). Royal dreams also required divinations, for example: “To be divined: The king dreamed of Lady Hao; is it not evil?” (Ibid: 17380).This was an inquiry into whether the dream would bring about inauspicious events. Zhou oracle bones have also been discovered, including those discovered in 1977 at the Zhouyuan 周原 site at Fengchu 鳳雛, Shaanxi 陝西 Province, and those discovered in 2003 at the Zhougongmiao 周公廟 site at Qishan 岐山, Shaanxi Province.The Zhouyuan oracle bones 221

Luo Xinhui

have attracted scholarly attention, but there are conflicting opinions about their origin. They seem to represent a tradition distinct from that of Yinxu, because the calligraphy and the style of the drill holes are completely different, yet they contain records of sacrifices to Shang kings such as Tang the Successful 成湯 and Taijia 太甲, which the Zhou would not be expected to perform. The problem remains unsolved. The majority of surviving writing from the Western Zhou period comes in the form of bronze inscriptions. The content of bronze inscriptions also has much to do with religious sacrifices; however, a significant portion of these inscriptions also reflects the social history of the Zhou as well as its system of decrees and regulations. The content of Zhou period bronze inscriptions is quite rich: They record the royal dynasty’s government planning, kings’ activities, instructions regarding sacrificial procedure, feasting, hunting, campaigns against neighboring states, political uprisings, rewards and investitures, the purchase and sale of slaves, the transfer of lands, law and lawsuits, oaths and covenants, as well as family history, marriage, and so on. (Ma Chengyuan 2003: 350). Bronze inscriptions record numerous instances of sacrificial ceremonies; for example, the Xian gui 鮮簋 inscription records the di 禘 sacrifice conducted by King Mu of Zhou 周穆王 and his sacrifices to his father, King Zhao 昭王. Numerous other inscriptions sing the praises of ancestors. The “Ji tong” 祭統 chapter of the Liji 禮記 states: 銘者，論譔其先祖之有德善，功烈、勳勞、慶賞、聲名，列於天下。 An inscription arranges and compiles the virtue and goodness of one’s ancestors, so that their merit, glory, rewards, and reputation are displayed throughout the world. The goal is to recount the ancestors’ virtue in order to encourage later generations to follow their example and also to display the special powers afforded by relying on the ancestors’ mysterious boon, thereby elevating the status of the sponsor of the vessel. Western Zhou bronze inscriptions also recorded important historical events. For example, the Li gui 利簋 inscription records King Wu of Zhou’s victory over the Shang, and the He zun 何尊 inscription relates that King Cheng moved the capital. In the He zun inscription, Luoyi 洛 邑 is called zhongguo 中國 (the central region), which is sometimes said to be the earliest known use of the term that is now used to refer to China (see the Introduction for an alternative view). Bronze inscriptions record numerous investiture ceremonies; among these are records of ministers’ receipt of land, official positions, and other similar forms of recognition, as well as records of various kinds of awards and appointments. The famous Greater Yu ding 大盂鼎 inscription, from the time of King Kang 康王, records the king’s command for Yu to inherit the responsibilities of his grandfather, Nangong 南公, to preside over military affairs, manage legal suits, and assist the king in governing the state. The king also instructed Yu to learn from the Shang kings’ loss of their state because of excessive drinking; he must revere Heaven and be diligent in government affairs. These instructions are roughly the same as those found in the “Jiu Gao” 酒 誥 chapter of the Book of Documents.The Lai ding 逨鼎, unearthed in 2003 at Yangjiacun 楊家村, Shaanxi Province, as part of a Western Zhou hoard, records the king’s conferring a title on Lai and encouraging him to take on the responsibilities of his deceased grandfather, assist the former kings, and stabilize the state of Zhou. The teachings or admonitions of the kings are commonly found in the Documents and had a guiding importance in Zhou-era governance. Their content 222

Early Chinese writing

can be compared with the inscriptions on such bronze vessels. Bronze inscriptions also recorded Zhou military expeditions; for example, the Shi Qiang pan 史墙盤 records the southern expedition of King Zhao, the Yu ding 禹鼎 records King Li’s 厲王 suppression of the so-called Huai Yi 淮夷 (peoples in the Huai River area), and the Guoji Zibai pan 虢季子白盤 records the king’s orders for Zibai to make war on the infamous Xianyun 玁狁 to the north and northwest of the Luo River 洛水. During the Spring and Autumn period, the House of Zhou declined, and the majority of bronzes discovered from this period are from the territorial states. Vessels made as part of a daughter’s or other female relative’s dowry and vessels made for entertainment were common during this period. For example, a set of bronze vessels discovered in a tomb at Liangdaicun 梁带村, Shaanxi Province, includes six tiny and elegant vessels which are considered to be the tomb occupant’s “toys” from his life as a nobleman in Rui 芮. However, inscriptions involving the conferring of titles all but disappeared in the Spring and Autumn period. Vessel sponsors referred to themselves as the son or grandson of a renowned ancestor, flaunting their descent in order to assert themselves. Inscriptions on such vessels as the Qingong gui 秦公簋 and Jingong pan 晉公盤 exalt the receipt of “Heaven’s Mandate” (tianming 天命) by the sponsors’ respective ancestors, never mentioning the receipt of the Mandate by King Wen or Wu, as would have been common in the Western Zhou (cf. Mattos 1997). Warring States period bronze inscriptions were even shorter, and it was typical to record such information as the name of the workshop where the vessel was cast and the official title of the individual supervising the production. During the Warring States period, the content of writing became richer, and, to judge from excavated bamboo documents, texts included legal documents, philosophical writings, historical records, stories of Zhou, and divination or sacrifice. Most bamboo manuscripts have been found in Chu and Qin, presumably because the climate is more conducive to their preservation. The Qin documents from Shuihudi have particular scholarly significance. They were unearthed in 1975 and include approximately 1,100 slips. The tomb occupant was a low-level Qin official named Xi 喜 (Happy). Many of these documents are related to the Qin legal structure, with excerpts from government documents, which have provided valuable insights into the details of Qin law. For example, a text which was named Eighteen Qin Statutes (Qinlü shiba zhong 秦律十 八種) by the modern editors of the text contains statutes on such topics as agriculture, animal husbandry, and grain storage. These legal documents pertain to water conservancy, the protection of forests, raising oxen and horses, and stockpiling grain. Answers to Questions About the Law (Falü dawen 法律答問) explains Qin statutes (and their application, especially in penal matters), also shedding light on the legislative process. Annals (Biannian ji 編年紀), which contains fiftythree slips, was placed beneath the head of the tomb occupant and records important events in the Qin conquest of the six eastern states and highlights in the life of the tomb occupant, Xi, between the first year of the reign of King Zhao of Qin 秦昭王 (306 bce) and the thirtieth year of the First Emperor (217 bce). Intellectual history is not absent from the excavated bamboo documents. For example, the approximately 730 Chu slips excavated at Guodian 郭店, Hubei 湖北 Province, in 1993, include such Confucian texts as The Five Forms of Conduct (Wuxing 五行), Lord Mu of Lu Asked Zisi (Lu Mugong wen Zisi 魯穆公問子思), and Human Nature Emerges from the Endowment (Xing zi ming chu 性自命出). The Way of Tang and Yu (Tang Yu zhi dao 唐虞之道) is a document which explains the ancient practice of abdicating the throne. The bamboo documents also include three editions of the Laozi 老子, which are the earliest known manuscript copies of that text. Taiyi Engendered Water (Taiyi sheng shui 太一生水) discusses the origin of heaven and earth at the hands of the god Taiyi. 223

Luo Xinhui

Another subset of the bamboo documents contains information about the sacrificial practices. Manuscripts from Baoshan 包山, Xincai 新蔡, Wangshan 望山, Tianxingguan 天星觀, Qinjiazui 秦家嘴, and Yancang 嚴倉 contain a significant number of slips describing divination, sacrifice, and prayer (e.g., Kalinowski 2009, 2010; Poo 1998: 69–101). The identity and status of the occupants of these tombs were not entirely the same. The occupants of the Qinjiazui tomb did not have high social standing; they belonged to the ranks of men of service (shi 士) or commoners.The occupants of the other five tombs were all highly ranked Chu aristocrats.The texts on divination and sacrifice were, in all five cases, occasioned by the occupant’s illness, and thus their features are similar. These manuscripts are valuable for understanding beliefs at different levels of society in the Warring States. There are also inscriptions on stone. Research suggests that stone inscriptions had already appeared by the Shang period; however, they did not become widespread until after the Qin. The so-called stone drums (shigu 石鼓 – so named for their shape, not their function) currently held in the Palace Museum in Beijing were discovered in Fengxiang 鳳翔, Shaanxi Province, during the Tang dynasty. These Qin inscriptions sing the praises of the beauty of the fields and the abundance of the hunt, with great literary merit. In addition, there are the famous stele inscriptions by the First Emperor of Qin (Kern 2000). These are in verse, a report to heaven, as well as a message for mundane audiences, that the great task of unification has been completed. An inscribed jade tablet called Qin Yin daobing yuban 秦駰禱病玉版, currently in the Shanghai Museum, records a prayer to “the great Mount Hua” 華大山 by a man from Qin by the name of Yin, expressing his anxiety after praying to heaven and earth, the four directions, the mountains and rivers, and the ancestors for his diseases, but not receiving any response. Through detailed analysis, the religious substance in Shang-era texts becomes evident, in as much as they all pertain to interactions between humans and spirits. Though many Western Zhou documents also pertain to religious ceremonies, they all address contemporary audiences (regulating, encouraging, or admonishing them), regardless of whether they were stored in a repository or inscribed on a ritual vessel. Writing increasingly pertained to social realities, and a large number of new topics, beyond the scope of religious practice, emerged. Writing and the compilation of documents are intimately tied to scribes. The “Spring Offices” (“Chunguan” 春官) section of the Rites of Zhou (Zhouli 周禮) records the duties of the “five scribes,” which mainly involve: (1) recording the programs used during sacrificial and court ceremonies; (2) recording royal edicts; (3) handling the proposal submitted to the court by the territorial lords; (4) managing documents and legal codes handed down from previous generations; and (5) disseminating the writing system.

Reflections on script and writing in early China Research into early Chinese writing has had a profound impact on inquiries into the history and culture. In the 1920s, following the discovery of and research into the oracle-bone script, scholars discovered that the records in the “Yu Xia shu” 虞夏書 section of the Book of Documents were not historical reality, and the history of early China needed to be reexamined (Gu Jiegang 1982). Since then, the use of new materials in exploring the first stages of early Chinese society has not ceased. However, the use of new materials in the research of ancient society poses numerous problems and challenges. First, interpreting graphs remains one of the most difficult problems. In oracle-bone research, the decipherment of characters is still a weak link. Even for some frequently used characters, scholars have been unable to confirm exactly what word is meant. For example, the oracle-bone 224

Early Chinese writing

script has the two ostensibly related forms and ; in the past, these were both assumed to be composed of mu 目 (eye) and ren 人 (person). Since the structure of the two are similar, they were taken to be variants of the same character and explained as jian 見 (to see). However, some scholars have pointed out that the ren 人 components of the two forms are different, with one standing and one kneeling, suggesting that corresponds to jian 見 and to wang 望 (to gaze) (Zhang Guiguang 1982). Another scholar (Qiu Xigui 2012)has suggested that may be interpreted as shi 視 (to view). However, yet another scholar has pointed out that the oracle-bone script already has another form of shi 視, namely , so it is inappropriate to interpret as shi 視 (Zhao Cheng 2006: 935).These two forms are commonly seen in oracle-bone texts, but even today, it is still impossible to determine which modern character they correspond to. The situation for bronze inscriptions is similar. The inscription on the famous Si Mu Wu ding 司母戊鼎, held in the National Museum of China, reads: To which modern character does correspond? In the past, scholars interpreted the inscription as si Mu Wu 司母戊, interpreting si 司 as “sacrifice” and Wu 戊 as a posthumous name; thus si Mu Wu 司母戊 means “sacrificing to Mother Wu.” This would mean that the structure of si Mu Wu 司母戊 is verbobject, but the vast majority of inscriptions on bronze vessels consist of the name of the tomb occupant; verb-object constructions are rare. Additionally, if si 司 is interpreted as “sacrifice,” then the posthumous name is Mother Wu, so the vessel would have been made for a deceased mother by her living son and should not have been placed in the tomb of the deceased. Similarly, in 1976, a bronze vessel with an inscription reading si Mu Xin 司母辛was discovered in the tomb of the consort of King Wuding 武丁 of Shang, Lady Hao 婦好, at Yinxu. If this is construed as a descendant sacrificing to Mother Xin 母辛, then it does not fit, because the vessel is a grave good, not a vessel to be used by descendants. Because of this, reading the character as si 司 is problematic. Some scholars provided a new interpretation, reading the character as hou 后 (queen), because in bronze inscriptions, characters are often written reversed; in this case, Hou Mu Wu 后母戊 would mean “Royal Mother Wu.” The problem is that in the oracle-bone and bronze inscriptions, hou 后 usually does not mean “queen.” Scholars have also suggested readings such as si 㚸 (of uncertain meaning), and a consensus has yet to be reached. The periodization of palaeographic material presents other difficulties. Dong Zuobin 董作 賓 (1895–1963) suggested a five-era periodization for oracle-bone inscriptions in 1933: (1) Pan Geng to Wu Ding; (2) Zu Geng and Zu Jia; (3) Lin Xin and Kang Ding; (4) Wu Yi and Wen Ding; (5) Di Yi and Di Xin (Zhòu). However, in the 1950s, Chen Mengjia 陳夢家 realized that divisions according to royal reigns are flawed and created a system of organization based on the names of the diviners, with names such as the “Bin 賓 group,” “Shi group,” and “Zi 子 group.” In the 1970s, Li Xueqin 李學勤 used characteristics such as the calligraphy and form of characters to suggest that the divinations from Yinxu could be divided into seven groups: Bin 賓, Shi , Chu 出, He 何, Li 歷, Huang 黃, and “Unnamed” 無名. However, relying on calligraphy to classify oracle bones complicates the process and makes it more difficult for a classification to gain acceptance (Zhao Cheng 2006: 1007). Periodization is also a problem for bronze inscriptions. Guo Moruo 郭沫若 relied on clues from recorded dates, connections among inscriptions related to historical events, literary style, and calligraphy, as well as the decorative patterns, physical structure, and provenance of objects to postulate an overall periodization (Guo Moruo 1957).This is the standard method of periodization today. For one example of inference from personal names and appellations, the inscription zuo Zhougong yi 作周公彝 on the Xinghou gui 井(邢）侯簋 means that this vessel was made for the Duke of Zhou, which implies that it was produced during the reign of King Kang. The Li gui 利簋 inscription records, “[King] Wu defeated Shang; it was on the morning of jiazi 225

Luo Xinhui

day” 武征商，唯甲子朝, suggesting that it was produced during the reign of King Wu. Other phrases, such as guang qi X shen 廣啟某身 (“[the ancestors] broaden and open my person,” said by a descendant who claims to be enlightened by his ancestors’ deeds) or qi yan zai shang 其嚴在 上 (“[the ancestors] reside above”) did not appear until after the mid-Western Zhou, providing hints for the dates of inscriptions containing them. Using such clues, the approximate date of an inscribed bronze vessel can often be determined, but because of situational limitations, the vast majority of vessels can only be said to have been produced during the reign of a certain king or during a certain period, with no likelihood of improving these estimates. Bamboo and silk documents also bear on our understanding of the compilation and transmission of ancient texts (e.g., Kern 2002, 2005a, 2005b, 2007). For example, the Tsinghua 清華 University collection of Warring States manuscripts includes a text whose content is roughly the same as the “Jinteng” 金縢 chapter of the Documents, with minor differences. After comparing the received version of the text with the bamboo slip “Jinteng,” some scholars believe that the two belong to distinct traditions of transmission. The transmission of texts in ancient times must have been complicated. The Tsinghua University bamboo slips also contain short texts called Yin Gao 尹誥 and Fuyue zhi Ming 傅說之命, which the organizer of the slips believes are equivalent to the “Xian you yide” 咸有一德 and “Yue ming” 說命 chapters of the ancient-script (guwen 古文) edition of the Documents, respectively (Li Xueqin 2012: 121). However, determining whether these two texts are from the ancient-script edition requires further research. Adding to the challenges is the fact that the context and provenience of looted artifacts like the Tsinghua University manuscripts are irrevocably lost (Goldin 2013). Some received documents record that Confucius edited the Odes and commenced their chain of transmission. After the appearance of Shilun 詩論 (Discourse on the Odes), a text in the Shanghai Museum collection, scholars pointed out that it alludes to an ancient edition of the Odes that circulated before Confucius. The transmission of the Odes was also not as simple as is suggested by the received texts (Chao Fulin 2013: 426). Excavated texts have revealed careful and subtle ancient distinctions in the writing system. For example, the Shuowen jiezi 説文解字, by Xu Shen 許慎 (ca. ad 55–ca. 149), glosses zhui 追 as zhu 逐, and zhu as zhui, without explaining the difference between the two. (Both words mean “to chase, to pursue,” but are not cognate.) In oracle-bone inscriptions, pursuing other people is always called zhui, but pursuing game is always called zhu. The shape of the characters reflects this distinction: zhu is written , suggesting “tracking a boar,” and zhui is written, , suggesting “chasing a host of people.” In later times, this distinction was elided (Yang Shuda 1954). Similarly, yong 勇 is written today as a combination of 甬 and 力, expressing courageous energy. In the Guodian manuscripts, however, there is a different yong written, , expressing a courageous heart, as when Confucians say that “to know shame is to draw near to courage” 知恥近乎勇 (from “Zhongyong” 中庸), referring to an internal courage (Pang Pu 2000; also Pang Pu 2009). These two graphs demonstrate that the ancients carefully distinguished between internal and external courage. Finally, excavated texts have also resolved a number of previously intractable questions. Three batches of excavated texts related to the Laozi have already been published: the silk Laozi from Mawangdui (two closely related editions); the three bamboo manuscripts of Laozi from Guodian; and the newly published bamboo edition of the Laozi from the Peking University collection. The three copies unearthed from Guodian are currently the earliest known manuscript copies of the Laozi, and the content differs from both the received version of the text, with commentary by Wang Bi 王弼 (226–249), and the Mawangdui versions. The different editions of the text have clear divergences in both compilation and philosophical content, demonstrating the complex circumstances of the transmission of the Laozi.The Guodian Laozi 226

Early Chinese writing

has shed light on the date of the Laozi as a text (e.g., Cook 2012: I, 195–216). Previously, there were those who attributed the text to the Warring States; others attributed it to the end of the Spring and Autumn period, and still others to the Western Han.The Guodian Laozi proves that the earliest examples of Laozi-related material date to no later than the mid-Warring States period.

Works cited Bagley, Robert W. 2004. “Anyang Writing and the Origin of the Chinese Writing System,” in Stephen D. Houston (ed.), The First Writing: Script Invention as History and Process, Cambridge: Cambridge University Press. Baxter,William H., and Laurent Sagart. 2014. Old Chinese: A New Reconstruction, Oxford: Oxford University Press. Cao, Jinyan 曹錦炎. 1996. Gu xi tonglun 古璽通論, Shanghai: Shanghai shuhua chubanshe. Chao, Fulin 晁福林. 2013. Shangbojian “Shilun” zonghe yanjiu 上博簡《詩論》綜合研究, Beijing: Shangwu yinshuguan. Chen, Mengjia 陳夢家. 1988. Yinxu buci zongshu 殷虛卜辭綜述, Beijing: Zhonghua shuju. Chen, Weizhan 陳煒湛 and Tang Yuming 唐鈺明 (eds.). 2009. Guwenzixue gangyao 古文字學綱要, Guangzhou: Sun Yat-sen University Press. Chen, Zhaorong 陳昭容. 2003. Qinxi wenzi yanjiu: Cong Hanzi shi de jiaodu kaocha 秦系文字研究：從漢 字史的角度考察, Taipei: Zhongyang Yanjiuyuan Lishi Yuyan Yanjiusuo. Cook, Scott. 2012. The Bamboo Texts of Guodian: A Study and Complete Translation, Ithaca, NY: East Asia Program, Cornell University. Galambos, Imre. 2006. Orthography of Early Chinese Writing: Evidence from Newly Excavated Manuscripts, Budapest: Department of East Asian Studies, Eötvös Loránd University. Goldin, Paul R. 2013. “Heng xian and the Problem of Studying Looted Artifacts,” Dao 12.2: 153–60. Gu, Jiegang 顧頡剛. 1982. “Yu Qian Xuantong xiansheng lun gushi shu” 與錢玄同先生論古史書 (1923), in Gushi bian 古史辨, rpt., Shanghai: Shanghai guji. I, 59–66. Guo, Moruo 郭沫若. 1957. Liang Zhou jinwenci daxi tulu kaoshi 兩周金文辭大系圖錄考釋. Revised ed., Beijing: Kexue chubanshe. Guo, Moruo (chief ed). 1978–83. Jiaguwen heji 甲骨文合集, Beijing: Zhonghua shuju. Kalinowski, Marc. 2009. “Diviners and Astrologers under the Eastern Zhou: Transmitted Texts and Recent Archaeological Discoveries,” tr. Margaret McIntosh, in John Lagerwey and Marc Kalinowski (eds.) Early Chinese Religion, Part One: Shang through Han (1250 bc-220 ad), Leiden: Brill. Kalinowski, Marc. 2010. “Divination and Astrology: Received Texts and Excavated Manuscripts,” in Michael Nylan and Michael Loewe (eds.) China’s Early Empires: A Re-Appraisal, Cambridge: Cambridge University Press. Kern, Martin. 2000. The Stele Inscriptions of Ch'in Shih-huang: Text and Ritual in Early Chinese Imperial Representation, New Haven: American Oriental Society. Kern, Martin. 2002. “Methodological Reflections on the Analysis of Textual Variants and the Modes of Manuscript Production in Early China,” Journal of East Asian Archaeology 4: 143–81. Kern, Martin. 2005a “The Odes in Excavated Manuscripts,” in Martin Kern (ed.) Text and Ritual in Early China, Seattle and London: University of Washington Press. Kern, Martin. 2005b. “Quotation and the Confucian Canon in Early Chinese Manuscripts:The Case of ‘Zi yi’ (Black Robes),” Asiatische Studien 59.1: 293–332. Kern, Martin. 2007. “Excavated Manuscripts and Their Socratic Pleasures: Newly Discovered Challenges in Reading the ‘Airs of the States,’ ” Asiatische Studien 61.3: 775–93. Li, Feng 李峰 and Branner, David P. (eds). 2011. Writing and Literacy in Early China: Studies from the Columbia Early China Seminar. Seattle: University of Washington Press. Li, Xueqin. 2005. “Zuoce Ban tongyuan kaoshi”作冊般銅黿考釋, Zhongguo lishi wenwu 1: 4–5. Li, Xueqin. 2012. Qinghua Daxue cang Zhanguo zhujian 清華大學藏戰國竹簡 vol. 3, Shanghai: Zhongxi shuju. Ma, Chengyuan 馬承源. 2003. Zhongguo qingtong qi 中國青銅器, Shanghai: Shanghai guji. Mattos, Gilbert L. 1997. “Eastern Zhou Bronze Inscriptions,” in Edward L. Shaughnessy (ed.), New Sources of Early Chinese History: An Introduction to the Reading of Inscriptions and Manuscripts, Berkeley, CA: Society for the Study of Early China.

227

Luo Xinhui Pang, Pu 龐樸. 2000. “Yingyan shu shuo – Guodian Chujian Zhongshan sanqi xinpang wenzi shishuo” 郢 燕書說 – 郭店楚簡中山三器心旁文字試說, in Guodian chujian guoji xueshu yantaohui lunwenji 郭店 楚簡國際學術研討會論文集, Wuhan: Hubei renmin. Pang, Pu. 2009. “Some Conjectures Concerning the Character ren,” tr. William Crawford, Contemporary Chinese Thought 40.4: 59–66. Poo, Mu-chou. 1998. In Search of Personal Welfare: A View of Ancient Chinese Religion, Albany: State University of New York Press. Qiu, Xigui. 2000. Chinese Writing, tr. Gilbert L. Mattos and Jerry Norman, Berkeley, CA: Society for the Study of Early China. Qiu, Xigui 裘錫圭. 2012. “Jiaguwen zhong de jian yu shi” 甲骨文中的見與視, in Qiu Xigui xueshu wenji (Jiagu juan) 裘錫圭學術文集（甲骨卷）, Shanghai: Fudan Daxue chubanshe, 444–48. Wang, Haicheng. 2014. Writing and the Ancient State: Early China in Comparative Perspective, Cambridge: Cambridge University Press. Wang, Yuxin 王宇信 and Yang Shengnan 楊升南 (eds.). 1999. Jiaguxue yibai nian 甲骨學一百年. Beijing: Shehui kexue wenxian chubanshe. Wang, Zhijun 王志俊. 1980. “Guanzhong diqu Yangshao wenhua kehua fuhao zongshu” 關中地區仰韶 文化刻劃符號综述, Kaogu yu wenwu 考古與文物 3: 14–21. Xie, Xigong 解希恭. 2007. Xiangfen Taosi yizhi yanjiu 襄汾陶寺遺址研究, Beijing: Kexue chubanshe. Yang, Shuda 楊樹達. 1954. Jiweiju jiawen shuo buci suoji 積微居甲文說、卜辭瑣記, Beijing: Zhongguo kexueyuan. Zhang, Guiguang 張桂光. 1982. “Guwenzi kaoshi size” 古文字考釋四則, Huanan Shiyuan xuebao 華南 師院學報 4: 87–89. Zhao, Cheng 趙誠. 2006. Ershi shiji jiaguwen yanjiu shuyao 二十世紀甲骨文研究述要. Taiyuan: Shuhai chubanshe. Zhu, Fenghan 朱鳳瀚. 2005. “Zuoce Ban yuan tanxi” 作冊般黿探析, Zhongguo lishi wenwu 1: 6–10.

228

11 THE SPIRIT WORLD

JUE GUOTHE SPIRIT WORLD

Jue Guo

Zaiwo said: “I have heard of the names of gui 鬼 and shen 神, but I do not know what they stand for.” Confucius replied: “The ethereal qi is the fullest form of shen. The corporal po is the fullest form of gui. Combining gui and shen is the highest form of teaching.” “The Meaning of Sacrifice,” Records of Rites (ca. third to second centuries bce)

What are spirits in early China? Why is it essential to explore the spirit world to understand early China? The dialogue cited here between Confucius and Zaiwo, his disciple, well illustrates both the gravity and arduousness required to have a clear grasp of these questions. On one hand, when “the great affairs of the state are sacrifice and warfare” 國之大事，在祀與戎 (Zuozhuan, Duke Cheng Year 13) and the outcome of these critical endeavors was believed to be intricately connected to the unseen realm, profound knowledge of the agents presiding over the spirit world justifiably constitutes “the highest form of teaching.” On the other hand, Confucius’ brief but convoluted answer clearly demonstrates the complexity of Zaiwo’s seemingly straightforward inquiry. This chapter hopes to unfold and address such complexities, paying particular attention to methodological issues pertaining to the study of the spirit world in early China.

Definition and method Gui shen is the most frequently identified term for spirits in general in the received texts from early China. It is commonly rendered as “ghost and spirit,” with the former primarily, though not exclusively, referring to the human dead and the latter for non-human spiritual beings. Such rendering may give an impression that the spirit world in early China had well-defined borders and dutifully registered residents, and therefore a census of the spirits should be sufficient to address any inquiry about the spirit world.This is far from being the case. Browsing through the transmitted textual corpora, not only are the various names for gui and shen entities in the texts by no means clearly understood, the boundaries that purportedly separate the spirit enclave are also notoriously difficult to delineate. To complicate the matter further, since the beginning of the twentieth century, modern Chinese archaeology has been excavating sites and unearthing

Jue Guo

manuscript texts from early China. The archaeological materials are not only new sources for further exploring the spirit world, but more importantly, they challenge the established taxonomies of spirits solely derived from the received textual tradition and raise critical questions about the nature of the sources and the ways of using them for historical inquiries. Against such a background, instead of conducting a conventional term-by-term survey (Zhan 1992), which inevitably decontextualizes and treats individual entities in isolation, I will take a source-centered approach to investigate the spirit world in early China, analyzing not only received textual sources but also material culture and recently excavated manuscripts. This approach shows that different types of source materials contain different aspects of the spirit world and treating sources typologically in terms of their nature, production, and preservation can better contextually capture a glimpse of a world of spirits, whose membership and relations were fluid, shifting, and changing. To begin, spirits, broadly defined, despite their different origins, diverse appearances, and vastly varied names, are reported, both historically and ethnographically, to have the following traits across cultures: immaterial, intangible, invisible, but sentient and agentive. Spirits are distinct from living humans in their physical and material form of existence but possess similar cognitive faculty and agentive capacity for taking actions.The spirit world, by extension, is imagined in its structure and ways of functioning as mirroring and in parallel to the living world. Lastly, spirits are believed to interact and be in communication with the human world. These characteristics, although by no means exhaustive, summarize how spirits and the working of their world are generally conceived and conceptualized. Suffice it to say that both the spirits’ materiality and ability to communicate have immediate and important implications for accurately identifying and properly analyzing relevant sources. It is notable that while the lack of a material, physical, or tangible form seems not to have posed difficulties for conceiving of spirits in the realm of the mind, in practice, the demarcation and framing of their presence – materially, physically, corporeally, or linguistically – are necessary and can take any form from an architectural space to a figurine to a name. Recognizing this practical demand of “materializing the invisible” is crucial in distinguishing and unfolding the different types and layers of references to spirits amalgamated in the sources. It may be helpful here to use an idealized process to elucidate the possible types and layers of references that could go into the sources. Imagine a spirit X with a cluster of attributes as a mental construct existing both in the individual and the collective mind of a certain group of people. As an occasion arises, formulated and dictated by what can be called cultural-social practices, spirit X acquires a marked material form that is the material manifestation of part of or all of its attributes. This material form occupies a physical space and interacts communicatively and affectively with the said individual or the group. The interactions as well as the communications are recorded, contemporaneously or retrospectively, in memory or in writing. Mental impressions and the written notes are accumulated, edited, and transmitted through oral or written means. This whole process can take place recursively and even indefinitely with changing variables: the occasion, the marked material form, the participants, the procedure, and the records. This hypothetical process theoretically results in three types of sources that I term as presencing, practicing, and discoursing, respectively.1 While keeping in mind the aforementioned practical requirement of “materializing the invisible,” Hans Ulrich Gumbrecht’s discussion of the “producing presence” is especially useful in devising these three typological lenses. Gumbrecht theorizes “presence” as being “in front of us, in reach of and tangible for our bodies”; and “production” as “the act of ‘bringing forth’ an object in space” (Gumbrecht 2004:17; xiii). The recast meaning of “presence” and “production,” or rather, returning to their Latin roots pre-esse 230

The spirit world

and producere, draws attention to the spatial dimension and the physical proximity of bringing forth something into tangible presence to the human bodily reach. Related to this emphasis on the material, physical-spatial, and bodily engagement in producing such a presence, Gumbrecht also proposes a nonhermeneutic approach to such engagement and their material manifestations, complementing what he calls the meaning-attributing interpretative approach. This consequently drives him to divide cultures into two ideal types: “presence culture” and “meaning culture” (Gumbrecht 1999: 358–359, 2004: 18; 79–85). In light of Gumbrecht’s emphasis on the material and physical-spatial dimension of “producing presence,” abbreviated as “presencing/presenting” in a recent study of the materiality of divine agency in the Ancient Near East,2 I use presencing to refer to a mode of action or process to materialize otherwise intangible and invisible spirits, i.e., presencing spirits in a marked material form in our hypothetical process introduced earlier. The presencing sources then are those marked forms or the remains of them, preserved in what can be broadly defined as material culture. What I call practicing and discoursing approximates the modes of action in Gumbrecht’s “presence culture” and “meaning culture,” respectively. In addition to Gumbrecht’s recognition that these two forms of culture have not “appeared (or will ever materialize) in its pure – in its ideal – form” (Gumbrecht 2004: 79), I further argue that these two modes of action can coexist and are sometimes intertwined in one culture. However, their different focus – participation in “presence culture”/practicing mode and interpretation in “meaning culture”/discoursing mode – results in different types of sources. In our hypothetical process, the interaction and the communication as well as the records of them constitute the practicing sources, which are documentary and instructional. The discoursing sources, on the other hand, are what can be called meta-level discourses that explain spirits and practices involving spirits by attributing cultural and social meaning to them, and therefore are interpretive and reflective. As such, discoursing sources are metaphysical in the sense that they go above and beyond the presencing moment and the practicing details but focus on the attributed significance or meaning of both. It is important to note that presencing, practicing, and discoursing as three modes of action are ideal types and should not be taken as substantive properties of the sources under analysis. They are useful as analytical lenses through which we look into the spirit world in question. Neither should they be taken as mutually exclusive to one another, nor be viewed in a linear progression.

Sources With these considerations in mind, let me turn to early China. It should be noted at the outset that because of the long time span that this chapter covers, from the Neolithic (ca. 9000–2000 bce) to the Han (202 bce–220 ce), the amount and types of available sources pertinent to the spirit world are enormous and inevitably uneven: the presencing sources are the only remaining kind for the first eight millennia or so; the practicing sources became available from the second half of the second millennium bce; the discoursing sources appeared in the second half of the first millennium bce. All three types of sources gradually became abundant from then onward. Similarly, it is important not to assume that there was a progressive development from simple to complicated forms of human-spirit communication and interaction but to consider such uneven distribution as a result of different preference for mode of action and different condition of source preservation. This uneven distribution of available sources thus should be taken into consideration when weighing the observations and conclusions drawn from them to avoid overgeneralization. In what follows, I will examine the selected sources in a chronological order but pay particular attention to the nature of these sources through the threefold typological lens of presencing, practicing, and discoursing. 231

Jue Guo

Presencing sources To begin, as already mentioned, sources about the spirits from the eight millennia or so in Neolithic China are exclusively presencing. In other words, we only have remaining material forms presencing spirits and material traces of past activities involving them.These sources are primarily obtained through archaeology, and their scattered, if not accidental, nature determines that they can only afford us a fragmentary view of the spirit world of the Neolithic period, and the earlier the source, the more so. The most demonstrative evidence of presencing spirits is from the Late Hongshan Culture 紅山in the north and the Liangzhu 良渚 Culture in the south, two complex societies in the fourth millennium bce.3 Located in the present-day western Liaoning 遼寧 and eastern Inner Mongolia 內蒙古, the Liao River valley was home to a series of Neolithic cultures, dated to as early as 7000 bce. The Hongshan Culture (ca. 4500–3000 bce), comprised of the largest number of sites, is considered the first complex society in Northeast China. In addition to the drastically increased number and size of settlements (five times that of its immediate predecessor, the Zhaobaogou 趙寶溝 Culture), domestication of pigs and sheep/goats, and clearer social stratification (Liu 2007: 271; Liu and Chen 2012: 172–178), the Late Hongshan society also had a “spectacular ritual landscape,” characterized by their “monumental public architecture, elaborate burials, and sophisticated jade artifacts” (Peterson and Lu 2013: 56). Our concern here is the so-called monumental public architecture. Archaeologists also refer to them as “temples” and “altars,” which are found adjacent to the elite burials on the mountaintops. A cluster of sixteen such sites, spreading over an area of 50 km2, were discovered in 1981 and dated to the Late Hongshan period (ca. 3650–3150 bce).4 Now known as the Niuheliang 牛河梁 Ritual Complex, this ceremonial center is the largest of its kind in the Hongshan period and has been most thoroughly excavated since its discovery (Guo D.S. 2004: 3). At the center of the Niuheliang complex, occupying the highest point of the mountain ridge, is the most wellknown “Goddess Temple.” The remains of the “Goddess Temple” are comprised of two largescale semi-subterranean architectural structures with walls painted in colorful geometric patterns. The larger of the two structures is a three-chamber construction, measuring 18.4 m long and 6.9 m wide, inside which broken clay pieces of human body parts including multiple heads, ears, shoulders, arms, hands, and breasts are found. The different sizes—from life-size to three times life-size—and styles—seated or standing—of the clay fragments indicate that they belong to approximately six individual sculptures. In the adjacent smaller single-chamber structure, clay fragments of large-size bird claws and wings, bear lower jaws, and what has been dubbed “pigdragon” (a hybrid of a head with pig-like snout and a bird body) as well as large-size pottery vessels are found, one of which was estimated to have a diameter of one meter at the widest). Among these, one well-preserved life-size human head (remaining 22.5 cm high and 16.5 cm wide) with prominent facial features, reddish face paint, and inlaid eyes made of two pieces of round jade has since become known as the “face” of the “Goddess.” Eight meters north of the “temple” there is a stone-walled area of 40,000 m2, within which three man-made platforms as well as round dug pits containing large pottery vessels and animal bones were also discovered (Guo D.S. 2004: 11–22). Another platform in a pyramid shape with a remaining diameter of 100 m, constructed from rammed earth and stones, was found in the southwest part of the Niuheliang complex (Guo D.S. 2004: 80–81). These platforms and pits, archaeologists postulate, may have been the “altars” and “sacrificial pits” used by the elite who were also buried in the Niuheliang complex. The implications of the “temple,”“altar,” as well as the human and animal sculptures have been variously postulated to represent “goddess worship,” “fertility cult,” or “shamanism,” just to name 232

The spirit world

a few. Putting aside the historically changing connotations of these concepts themselves, the material remains – architectural structures and figurative sculptures – do not directly corroborate these highly suggestive interpretations, which are largely anachronistically projected based on later texts or ethnographically inspired due to the lack of direct and contemporaneous evidence other than the material remains.They are, however, unambiguously material manifestations of the enormous endeavor of the people who built and made them literally “larger than life.” It is also important to note the absence of permanent settlements and residential traces in the archaeological records within the 100 km2 surrounding area of the Niuheliang complex. It is not difficult to infer that the construction of the Niuheliang complex would have required a considerable force of human labor, traveling an equally considerable distance from where they permanently lived5 and climbing up to the top of the mountains to construct a building (i.e., the “temple”) with multiple chambers so large, compared to the average Neolithic residential houses, that they can only be considered as a public space. Inside this public space, not only were the walls decorated in colorful patterns, visually imposing sculptures were also displayed. The size of the vessels found at the site suggests a large quantity of food, which point to either a larger number of participants or a perceived need for displaying abundance. The constructed gigantic platforms (i.e., the “altars”) similarly highlight both the material massiveness and the visual displayability. Taken together, it is reasonable to say that the lofty yet made-accessible location, the largerthan-everyday physical space, and three-dimensional sculptures of enlarged scale and emphasized visibility were presencing entities that were otherwise immaterial, invisible, and intangible. Whether these entities were goddesses, ancestors, or forces that can evoke shamanistic mediations is unclear, and the details and nature of the activities involved in such presencing as well as the purpose of food and animals involved in these public spaces are also no longer known to us. What is clear is the marked emphasis on the materiality and physicality in and through presencing, constructing monumental architectural structures with durable materials such as stones and fired earth; making expressive objects by molding clay into three dimensions. What specifically was believed to be presencing is less important than recognizing the displayed nature of what can be called the spirit world of the Late Hongshan people through amplified material and physical visibility in the fourth millennium bce. Looking through the lens of presencing, traces of a spirit world can also be found in the south in the Liangzhu Culture (ca. 3300–2000 bce), although their material and physical expressions were distinctly different. Unlike the Late Hongshan society, which was comprised of hundreds of individual sites that were only loosely “grouped” together around ceremonial centers such as the one at Niuheliang, the Liangzhu society, in addition to the numerous clusters of settlements, had an unambiguous center, what archaeologists call the Ancient Liangzhu City (Liu Bin, 2007: 12). Located around the Lake Tai 太湖 area in the lower Yangtze River region, this was a concentrically constructed city site with a palatial area (30 ha) comprised of a man-made platform of about 10 meters high at the center and several large-scale architecture foundations (up to 3 ha), an inner city (300 ha), a walled outer city (800 ha), and an astonishingly extensive and developed hydraulic system in the surrounding area which controlled the flood and facilitated waterway transportations. The sophisticated engineering and construction of such a city must have required a developed organization of resources and labor in the fourth millennium bce. Like the Late Hongshan society, the Liangzhu society was already hierarchically stratified, and the elite possessed both material and symbolic privileges, exemplified by their richly furnished burials and man-made pyramid-shaped structures (“altars”) found adjacent to the burial grounds, both of which were constructed on mountaintops, northeast and northwest of the fortified city (Liu and Chen 2012: 237–238). 233

Jue Guo

In contrast to the display of life-size or larger three-dimensional clay sculptures in individually identifiable anthropomorphic and zoomorphic forms fixed in the spacious Hongshan “temple,” Liangzhu had a miniature (less than 8 cm high), two-dimensional, and composite form of presencing powerful entities on portable objects made of hard and durable materials.This composite image has two intertwined parts. On the top, was a human shape with a reverse-trapezoid shaped face wearing a dramatic and oversize headdress, round eyes, broad mountain-shaped nose, and an open square mouth with neatly aligned and visible teeth. In an upright position, the human figure appears to extend both arms and hands holding onto a beast beneath, which has two huge round eyes, fierce fangs, and claws. This intriguing image of minute details was meticulously carved, partially in lower relief, onto the hard surface of various jade and stone objects found in large quantities in elite tombs. Given the complexity of the Liangzhu society and the pervasiveness of this image on ceremonial objects commonly identified as symbols of authority—cong-tubes 琮, yue-axes 鉞, and zhang-staff 杖—it has been called the “spirit emblem” (shenhui 神徽) of the Liangzhu world. This image has prompted some scholars to postulate that the owners of these objects—the high elite—were also shamans in charge of mediating between the human and the spirit world (Chang 1989; Liu Bin 2007), a reading highly influenced by the prevalent tropes in evolutionary anthropological and social development theories but not yet convincingly attested archaeologically. As with the Hongshan spirit world, the existing evidence does not allow identifications of individual and specific spirits or access a precise and contemporaneous meaning and function of this image in the lives of the Liangzhou people. Analyzed as presencing spirits, however, their material and visual characteristics are salient, with which we can infer how the Liangzhu spirits might have had been imagined. Because it is a particularly hard and prestigious kind of stone, carving anything onto the surface of jade is a time- and labor-intensive task, not to mention an image of this level of detail and sophistication, which demanded both engraving and lower-relief techniques to create the combined visual effect that elevated the vividness to almost animation. Hard media such as jade and stone may be a material manifestation of the strength perceived in the spirits as well as a means to enhance the physical endurance of the spirits’ presencing. The anthrozoomorphic design of this composite is visually intriguing. Moreover, it is also cognitively invoking both the natural familiarity and the counterintuitive confusion, a property Pascal Boyer calls a “cognitive optimum” for making an enduring impression as well as enabling successful transmission (Boyer 1994: 121). Indeed, such composites are phenomena observed in many regions and cultures throughout time (Wengrow 2014: 74–77). Unlike those mechanically reproduced using cylinder-seals in Mesopotamia in the fourth millennium bce (Wengrow 2014: 65), the Liangzhu image was individually carved and must have required a highly skilled hand and extended hours to place a complicated image so consistently on such a great number of objects. Admittedly, compared to the soft and formable clay, the hard surface of jade and stone makes mechanical seal-impressing difficult, if not impossible, but carving may also have had been considered a necessary action of physically rendering the spirit tangible and thus a crucial part to presencing. The particular material choice, the visual design, and the physical execution all speak of the significance of what this image represented and the need to materialize this significance. Unlike the life-size or larger-than-life size Hongshan sculptures displayed publicly as independent objects, these composite images are miniature in size and an integral part of another object. The meaning, agency, and efficacy of the image and the object are mutually constituted. Two further points are worth noting. First, although the composite nature of the image and its closeup visual effect are intriguing and invoking, minute details in miniature on portable objects were arguably not primarily visual or meant to be viewed by people other than the holder or the 234

The spirit world

owner, as they would be almost invisible to naked eyes even at an arm-length distance. Second, presencing on hand-held objects is telling, because such a material choice enables the holder of these objects to remain physically “in touch” with the now tangible entity. Therefore, the miniature size of both the materialized spirit and the portability of its physical carrier effectively limit the access to the spirit world to the holder and owner, although any claim of social-political authority derived from the object requires a shared knowledge of such particular material form of presencing, at least among the elite groups in the Liangzhu world, if not also by the non-elite at large. Both Hongshan and Liangzhu have been called a “Jade Culture” on account of the quantity as well as quality of jade objects found in their elite assemblages. Material and cultural exchanges, if not trade, clearly had taken place between the two, at least among the elite, despite the long distance between them. However, Hongshan and Liangzhu peoples took different paths to presencing spirit entities, and their different choices, as explained earlier, may reflect the specific properties coming to be associated with their spirit worlds. Differing social environments that resulted in unique ways that spirits were imagined, perceived, and materialized may have also played a role. In particular, social and developmental complexity has often been positively correlated with political and ideological sophistication. Although Hongshan is considered a complex and hierarchical society, it is Liangzhu that has been postulated to be a state-level polity ruled by an elite class that derived its political authority from the spirit world and sustained a theocracy by monopolizing access to it (Liu Li 2007; Liu and Chen 2012: 241). It should be pointed out, however, that both Hongshan and Liangzhu were preliterate societies and left no other traces except their material culture. Although the lens of presencing allows us to discern the existence of a spirit world, characterize the materialized presence of spirits, and assume communications and interactions between the human realm and the spirit realm, we cannot infer further the specific forms of such communication and interaction or the function thereof from presencing sources alone without other corroborative evidence, for instance, the contemporaneous practicing sources.

Practicing sources The practicing sources—here defined as the records of the communication and interaction between human actors and the spirit world—did not appear in China until Late Shang in the second half of the second millennium bce. Shang was known as the second of the “Three Dynasties” (Chang 1980) in traditional texts such as Shiji or The Grand Scribe’s Records, compiled by Sima Qian 司馬遷 (ca. 145–86 bce), more than a millennium after Shang’s demise. The accidental discovery of the “oracle bones” in 1899 led to scientific excavations at Anyang 安 陽, Henan, beginning in the 1920s and continuing until this day. Excavations of large architectural complexes (identified as “palaces” and “ancestral temples”), workshops, royal tombs, and commoners’ cemeteries, for the first time, archaeologically confirmed Yinxu 殷墟, “Ruins of Yin,” as the last Shang capital (Tang 2004: 1–3; Liu and Chen 2012: 355). The finds of more than 100,000 pieces of inscribed animal bones and turtle shells as well as their subsequent decipherment attested to the earliest mature form of Chinese writing (Boltz 2011: 65), and these “oracle-bone inscriptions” formed the first and foremost written sources from and for the Late Shang (ca. 1250–1045 bce).6 The name—“oracle bones”—reveals their primary divinatory nature. Both animal bones and turtle shells were used in a pyromantic procedure that came to be known as bu 卜 or “crackmaking divination” in the later texts. Although pyromancy, in particular, scapulimancy, had been practiced long before the Shang in China—scapulae from goats/sheep, pigs, and deer bearing burn marks having been found at many Neolithic sites along and north of the Yellow River and 235

Jue Guo

dated to as early as 4000 bce—ox scapulae only began to be used in significant numbers in the early Shang and eventually became the dominant kind of animal bones used for Shang scapulimancy, especially at Anyang between the thirteenth and eleventh centuries bce. Compared to the Neolithic scapulae that were directly scorched on the surface, the Shang began to bore holes in the scapulae before applying heat to them, which was believed to make cracking the bones easier and more subject to control (Piao 2011: 15–31). However, what truly separates the Shang pyromancy, especially the Late Shang at Anyang, from the earlier practice are two unique developments.The first was the use of turtle shells in pyromancy. Turtles shells may have already been part of the ritual life of the Neolithic peoples. Complete sets of turtle shells, sometimes containing dice-like pebble stones, have been found in burials as early as 7000 bce at the Jiahu 賈湖 site in the Huai River 淮河 area (Zhang and Cui 2013: 207–209) and later at numerous Dawenkou 大汶口 sites in the Shandong 山東 region in the fourth and third millennia bce (Gao and Shao 1986: 59–63). Some scholars speculate that they may have been used for divination (Zhang and Li 2005).The fact that none of the turtle shells discovered at these Neolithic burial sites bear any sign of having been subjected to fire, but were accompanied by small pebble stones, indicates the mechanism of their usage, if indeed divinatory in nature, was likely numerical. Turtle shells with burn marks first appeared in the Erligang 二里崗 period at the beginning of the second millennium bce and became ubiquitous in the Late Shang (Piao 2011: 32). Turtle shells and animal bones were unambiguously used in divination at Anyang due to the second major development of Late Shang pyromancy, which was the engraving of abbreviated and terse records of divinations directly onto the surface of bones and shells. Inscribing the bones and shells was a milestone, the significance of which few other developments in Chinese history surpass. Not only did this practice preserve the earliest evidence of mature Chinese writing and constitute the first written corpus, which places the Late Shang securely at the beginning of China’s historical period (Keightley 1999: 232), these divinatory inscriptions also furnish the earliest written practicing sources about engaging a world of spirits, which for the first time can be individually identified by their contemporaneous names and their precise role in relation to the living. It should be pointed out that although pyromancy was widely practiced in the Shang, inscribing the bones and shells seemed to be an exclusively elite phenomenon, and nearly all of the inscribed bones and shells were found at Anyang, the political and ritual center of the Late Shang.7 It is certainly not inconceivable that ordinary people living in the Shang shared access to a world of spirits with their elite members. In fact, practicing pyromantic divination and sharing similar burial practices, albeit on a much-reduced scale, suggest they might have; however, without practicing records like those of their elite counterparts, our access to the spirit world of the non-elite in the Shang and the specific ways in which they engaged with that world is limited to their presencing in the material culture. As for the elite, especially the royal elite of the Late Shang, although the fundamental principle of “materializing the invisible” in presencing the spirit world remained the same, the material and physical outlook of doing so changed remarkably, probably as a result of mastering and utilizing writing. Unlike the Hongshan and the Liangzhu spirits, whose presencing is figurative and stable, the presencing of the spirit world of the Late Shang elite was symbolic and transient, brought forth anew through each crack-making divination. As the extant evidence from Anyang shows, ox scapulae and turtle plastrons, occasionally also carapaces, were worked and prepared before rows of holes were bored and chiseled on the backside, which would then be subject to heat until cracks – incidentally in the shape of the graph 卜 – appeared on the front side (Keightley 1978: 12–27). It is the scholarly consensus that this crack-making act was how the living communicated with the spirits. What is not clear is the exact mechanism of such divinatory communications. Two main hypotheses have been proposed, one based on reading the 236

The spirit world

shapes of the cracks and the other on the cracking sounds.8 Neither can be supported or rejected based on available Shang evidence alone, although some fourth-century bce and later texts suggest that the divinatory prognostication was derived from reading the particular forms of the cracks (zhao 兆).9 Regardless of whether the mechanism for communication was semiotic or acoustic, there is no firm evidence that the Shang spirits were presencing figuratively in the divinatory process of crack-making.10 Besides the non-figurative presencing of spirits, it is also unclear where the divinatory communication with the spirits took place. Keightley postulates it to be in the presence of ancestors in their temples (Keightley 1998: 978, 1999: 245),11 but this has not yet been directly and unambiguously corroborated by archaeological evidence. In other words, the available evidence for the communication between the Late Shang elite and their spirit world is not sufficient for a full reconstruction of the entire process; however, the very fact that such actions were documented in writing on the bones and shells does allow us to know precisely the “who” (the spirits and human agents) and the “why” (the occasions and topics), and occasionally the aftereffects of such communications. Since it has been convincingly demonstrated that the inscriptions were engraved onto the surface of the communication medium—scapulae or plastrons—after the crack-making procedure concluded (Keightley 1978: 45; Bagley 2004: 196–197), they are not part of the presencing process but constitute the records of the divinatory practice that engaged the spirit world, and thus practicing in nature. Even though the majority of the Late Shang scapulae and plastrons came to light in fragments and the inscriptions vary greatly chronologically,12 an ideally complete record includes four main parts: preface (and postface), charge, prognostication, and verification may read as follows:13 [preface] Crack-making on the guiwei day, Que divined: [charge] On the next jiashen day, the king will host an entertainment ritual for Shang Jia and Ri. [prognostication] The king prognosticated and said: It will be auspicious to host the entertainment ritual. [verification] [The Shang Jia and Ri] were indeed entertained. Engraved on the top part of a plastron, this brief inscription records the occasion and participants of a pyromantic divination performed on this very shell on a guiwei 癸未 day, the twentieth day in the sexagenary calendrical cycle, in the thirteenth century bce.14 Que, the most frequently recorded diviner at Anyang, posed a question to the shell: on the next day, a jiashen 甲申 day or the twenty-first day, whether it is auspicious for the king to “host an entertainment ritual” (bin 賓), one of the most popular rituals in the Shang oracle-bone inscriptions, for two named spirits.The first is identified by the name Shang Jia 上甲 or “Exalted Jia,” which, according to the reconstructed Shang king list, was the first and highest ancestor in the royal genealogy (Keightley 1978: 185). Jia 甲, here appearing both in the day name, jiashen, and the ancestor’s identification, Shang Jia, is the first of the ten-unit counters, later known as the “stems” (gan 干), which together with another twelve-unit counters, the “branches” (zhi 支), were used to mark time and other sequences throughout pre-modern China (Smith 2010). This overlap between the day and the ancestral identification is not incidental but especially significant in understanding the nature of ancestors in Late Shang, a topic treated in more detail later. The other spirit to be “entertained” was Ri 日, the Sun, a nature power. It was the king himself—as was the case in most of the divinations—who made the judgment, in this case a positive one. Consequently, the entertainment ritual was carried out the next day, in which the two spirits were “indeed entertained,” and the divination was deemed successful. Most inscriptions on the bones and shells conform to this formula, although very few have all four parts (Eno 2009: 51). 237

Jue Guo

It is in such inscriptions, carved onto the bones and shells, that the Late Shang spirits such as Shang Jia and Ri become knowable to us. What we should keep in mind, however, is the fact that the Shang oracle-bone inscriptions are records of actual individual divinations, each of which addressed a specific concern that involved particular spirits, as the preceding example shows. Therefore, it is not surprising that the world of spirits emerging from such records appears to be a collective of disparate entities.This is not to deny that there was general conceptual and structural knowledge about the spirit world as a whole. Indeed, the astonishingly high frequency and the orderly fashion of the Shang pyromantic divinations may be good indicators of such overall understanding of the spirit world (Keightley 1978: 169–170, 1984: 13–14); however, such knowledge would only have been embedded in the divination practice and implicit in the divination records rather than explicitly expressed or theorized, as we will see in discoursing sources almost a millennium later. Although the spirit world of Late Shang cannot be neatly construed as a “pantheon” or explicitly expressing a “theology,”15 scholarly classifications of the otherwise numerous spirits are analytically constructive. For example, David Keightley divides the Late Shang spirits seen in the inscriptions into six groups (Keightley 2004: 5–6): (1) Di 帝, the high god; (2) Nature Powers, like He 河 (the River Power),Yang (the Mountain Power), and Ri (the Sun); (3) Former Lords, like Nao 夒 and Wang Hai 王亥; (4) predynastic ancestors, like Shang Jia and the three Bao 報; (5) the dynastic ancestors, starting with the dynastic founder Da Yi 大乙; (6) the dynastic ancestresses, the consorts of those kings on the main line of descent” Robert Eno perceptively points out that, based on lineage relations, these six categories can be further grouped into what he calls “those whose members stand in uncertain or no relation to the core Shang lineage” (i.e., categories 1, 2, 3) and “those populated by the ancestral members of the core Shang lineage” (i.e., categories 4, 5, 6) (Eno 2009: 54–55). As shorthand, I will call the former the “non-ancestral group” and the latter “ancestors proper.” The two groups not only differ in their group composition but also were represented and treated differently in the practices of divination, ritual, and sacrifice, the three major means through which the Late Shang elite communicated and interacted with their spirit world. Analysis of these differences allows us to infer some salient characteristics of the spirits and of the ways that the Late Shang elite interacted with them as manifested in the divination records, despite the lack of contemporaneous discoursing sources. Let us begin with the ancestors proper. This group is homogenous in the sense that they were all deceased members of the core Shang royal lineage and were all addressed by an ancestral identifier, or conventionally called the “temple name,” composed of a status-marker (lineage and generational position) and a gan-stem ordinal. The ubiquitous presence and deep involvement of the ancestors in the Late Shang elite’s life can be readily seen in the almost daily communications between the living and the ancestors. There were a great variety of topics—from the king’s toothache to the harvest of the year—that called for crack-making divinations and various rituals and sacrifices, especially the gradually standardized and routined “Five-Ritual Cycle.” The deceased kin of the royal lineage did not automatically or immediately ascend into the “Five-Ritual Cycle” to receive the regular cult of the ancestors but only gradually transitioned into it, a process described by Puett as “placing the ancestor” and Keightley as “ancestor making.”16 It consists of three stages, which began with an individual death of a Shang king or an heir-producing royal consort. Presumably they had his or her personal name in life, but when they died, they would no longer be the individual with a personal name; instead, in divinatory 238

The spirit world

and sacrificial contexts, they would be referred to as Father (fu 父) or Mother (mu 母) by the reigning king. The other part of their twofold ancestral identifier was the “day” in the form of a gan-stem, on which he or she received sacrificial offerings. This was also decided by crackmaking divination, such as in the following two records: “Crack-making on bingshen (day 33), [Diviner] Chu divined: ‘In making Xiao Si’s day, let it be a gui.’ Eighth Month” (Heji 23,712); and “Crack-making on renwu (day 19), [Diviner] Da divined: ‘On the next guiwei (day 20), offer to Xiao Si three penned sheep and X-sacrifice one ox’ ” (Heji 23,719).17 Conceivably, Xiao Si would appear thereafter as Mother Gui18 in subsequent divinations, when sacrificial offerings were divined to be offered to her on a gui day. If the first stage was to de-individualize the dead kin by placing him or her onto a map of kinship relations, then the second stage was to replace their personal name with an abstract sign, in this case a numerical marker that corresponds to a calendrical day, which further depersonalizes the once distinct individual. However, entities such as Father Yi or Mother Xin would not be offered the rotation of the “Five Rituals” until they became Grandfather/Ancestor (zu 祖) or Grandmother/Ancestress (bi 妣), that is, at least two generations removed from the reigning king (Keightley 2004: 27). This was the third stage: that the deceased kin, only through the passing of time, was admitted into the regular ancestral cult, where they completed the transition from an individual personality, to a relational entity, and eventually to an ancestral persona, or more accurately, a position-holder in the ancestors proper, whose power derived from being part of a well-structured and organized whole rather than any individual and personal traits or kinship relations. It is through these stages that Lady Hao 婦好, the most renowned consort of King Wu Ding (ca. mid. thirteenth–mid-twelfth century bce), became Mu Xin 母辛 or Mother Xin, and eventually Bi Xin or Ancestress Xin.19 Naturally, the size of the ancestors proper increased as time passed, and by the end of Shang, the last king, known as Di Xin 帝辛 in the later texts, would have had at least six predynastic founders, twenty-eight deceased kings, as well as twenty-one passed royal consorts, spreading over twenty-two generations, at hand in his ancestral “roster,” with all of whom he was expected or even obligated to communicate and interact through regular divination, ritual, and sacrifice (Keightley 2004: 13–14).20 Such a daunting task, as Keightley aptly says, would be the equivalent of a modern person commemorating his or her ancestors beginning with the one who died in the twelfth century (Keightley 2004: 40). One difference in this analogy may be that, rather than remembering the ancestors using the modern conventions, through personal names, individual personalities, particular life experiences, and portraits, when available, the Late Shang ancestors represented in the divination records were anything but individual, personal, particular, or figurative. Instead, the Shang kings and their diviners and ritualists appeared to have calculated and devised a complex schedule through divination that resembles a “ritual calendar,” in which each ancestor was assigned a day which was also part of his or her ancestral identifier. By the end of the Shang, such ritual scheduling would largely overlap with the actual calendrical year, taking about 360 days to complete the entire “Five-Ritual Cycle” (Keightley 2004: 23). It is not entirely clear whether such ritual scheduling was considered as a “calendar” per se,21 but it must have effectively assisted to maintain the order of the ancestral cult. Outside the ancestors proper and the ancestral cult, the non-ancestral group does not appear to be as orderly structured and routinely entreated. It was a heterogeneous group, and its members had or retained self-contained and individual designations such as Di, the River, the Mountain, the sun, and Former Lords like Nao and Wang Hai. The divinatory communications between these non-ancestral spirits and the humans were also ad hoc, and no regular rituals or sacrifices were established for them. They received no, little, or irregular cult.22 Divination records also show that they had a separate domain of influence from those in the ancestors proper. For 239

Jue Guo

instance, the high god, Di—whose origin is still poorly understood (Eno 1990; Puett 2002: 48–49)—and Nature Powers seem to have been communicated with primarily in conjunction with natural events such as rain, snow, hail, wind, and thunder, or for the well-being of the entire community, such as harvest or warfare, but not about the personal affairs of the reigning king such as his own illness or the childbirth of his royal consorts, in which case only ancestral spirits were consulted. Similar divisions of influence can also be observed within the ancestors proper. The most recently deceased Father and Mother spirits as well as the more recently ancestralized members in the ancestral cult were more likely to be involved in the personal affairs of the living king and the royal family members than those who belong to more remote generations of ancestors and ancestresses. It is worth reiterating that, due to the dispersed nature of the oracle-bone inscriptions being the practicing rather than discoursing sources, the exterior contour of what we have been calling “the spirit world” of Late Shang is by no means definite. The lack of references to spirits in abstract and conceptual language that can hint at or point to their self-identified membership in a higher-level community, such as a pantheon, is also a factor for our caution. More importantly, an ancestor-based and ancestralization-oriented reconstruction of the Shang “pantheon” has two potential drawbacks. First, it generalizes the patterns seen in the formation of the ancestors proper—the “strong impulse to classify, impersonalize, and order the dead,” or “Ordnungswille” (Keightley 2004: 29, 1998: 793)—and thus risks overlooking the fact that the non-ancestral spirits were not systematically integrated into the ancestral cult. Their very existence outside the ancestral cult indicates that the attempt at placing a marked and discernable order on the entire spirit world, if ever was meant to be totalizing, was only partially realized because the division between the non-ancestral and ancestral groups remains evident.23 Second, seeing “ancestralization” as an essential feature of the spirit world and using the level of “ancestralness” to characterize the nature of spirits is an anthropocentric perspective that gives the humans a disproportional agency over that of the sprits.“The [Shang] world was filled with manifestations of sacred power, but Shang survival depended upon manipulating, cajoling, and competing with those manifestations, upon a determined effort to interpret, order, and dominate the arena of religious forces” (Keightley 2000: 119). While not denying the Late Shang elite’s meticulous effort to order the unseen spirit world through divination, ritual, and sacrifice, and their seemingly partial success in “making” and “placing” the ancestors, the fact that a significant number of spirit entities remained independent and defied such systematizing efforts is worth noting. It points to the agency of the spirits that played a significant role in creating a tension and necessity, which motivate and maintain the communications between the human realm and the spirit world. In comparison to the later periods in Chinese history, our knowledge about the Shang is extremely limited. Interestingly, the only written source from Late Shang were divination records, with which we are able to see, for the first time in actual practice, what and how spirits were actively involved in almost every significant aspect of the Late Shang elite life.24 The fact that such a wide range of areas of concern were constantly and repeatedly divined suggests that spirits were not only perceived to be within the reach of humans through divination, ritual, and sacrifice, but also recognized as active and powerful agents deeply influential in human affairs. The ordering and systematizing tendency toward abstraction seen in the naming practice in the Late Shang divination records are especially long-lasting elements in the cultural repertoire that later peoples continued to utilize and transform, precisely for the reason that spirits are also resistant and resilient agents against such human efforts. In the centuries following the Zhou conquest of the Shang in 1045 bce, in comparison with the singular source of inscribed bones and shells of the Shang elite’s pyromantic divination practice as well as the silent material remains of past sacrificial activities, practicing sources 240

The spirit world

became increasingly diverse and more informative about the spirit world of the time. What is most worth noting is the appearance of a variety of typologies of the spirit world. In particular, a gradual bifurcation became discernible between what can be called state practice and individual practice pertaining to their respective typologies of the spirits. In its eight hundred years of existence (ca. 1045–256 bce), the Zhou experienced drastic changes and transformations in its socio-political structure and territorial configuration. After the sack of the Zhou capital and the forced relocation to the east in 771 bce, in the subsequent periods known as the Spring and Autumn and Warring States, the Zhou king was only a nominal “Son of Heaven,” and “All-under-Heaven” was divided among powerful regional rulers of the gradually formed territorial states. It is generally agreed that the Zhou initially continued many of the Shang practices—divination, ancestral cult, and sacrificial rituals—while gradually making adjustments and implementing changes. Although the origin and the scale are still debated, bureaucracy germinated not too long after the establishment of the Zhou (Li Feng 2009), developed along the course, and eventually matured in some of the regional states, especially Chu in the south and Qin in the west in the fourth century bce. The centralizing and structuring tendency inherent in bureaucratization and the fragmenting force of the territorial states are the two conditions for the development of different typologies of spirits, especially in the sense of a state-organized pantheon on one hand and the appearance of multiple or even competing regional typologies on the other. The practicing sources, the contemporaneous ones in particular, from Western Zhou (ca. 1045– 771 bce) include transmitted and excavated bronze vessels and bronze inscriptions engraved on them, documenting rituals and events for ancestral veneration and commemoration, as well as sacrificial hymns later collected and transmitted in the anthology Book of Songs (Shijing 詩經) and some royal speeches enlisting ancestral and other spiritual support preserved in the semihistorical Book of Documents (Shujing 書經) (Kern 2009). Similar to the Shang oracle-bone inscriptions, these Western Zhou practicing sources are of elite nature; but unlike the cracked and inscribed bones and shells of the Shang pyromantic divinations, whose practicing contexts can be more convincingly inferred through the material remains of the divination, with the exception of bronze vessels, most of practicing sources in the Western Zhou listed above are transmitted only in textual forms, and their original practicing contexts are no longer retraceable. Furthermore, the spirits—primarily the ancestral ones—also appear in an ad hoc fashion in these Western Zhou sources. It was not until the fourth century bce that a typology of spirits and a Zhou pantheon was reconstructed, in a text called Rites of Zhou (Zhouli 周禮). Zhouli is also known as Zhouguan 周官, “Zhou Offices,” and has six main sections, each corresponding to an idealized branch of the Zhou state bureaucracy—“Tianguan” 天官 (Heavenly Office), “Diguan” 地官 (Earthly Office), “Chunguan” 春官 (Spring Office), “Xiaguan” 夏官 (Summer Office), “Qiuguan” 秋 官 (Autumn Office), and “Dongguan” 冬官 (Winter Office)—and together they symbolize the entire spatial and temporal dimensions of the cosmos.25 Although the Zhouli text itself was of a later date, the reconstruction of the Zhou bureaucracy may have drawn upon earlier sources as well as historical and cultural memories of the past. It offers a typology of the spirit world structured through state bureaucratic means that was conceivable in the fourth century bce. This is a tripartite typology that divides the world of spirits into three realms in three corresponding categories: tianshen 天神 “heavenly spirits,” rengui 人鬼 “human dead,” and dishi 地 示 “earthly spirits.” The “heavenly spirits,” include Heaven (Tian 天), High God (Shangdi 上 帝), the Sun (Ri 日), the Moon (Yue 月), and the Stars (Xingchen 星辰), as well as Overseer of Lineage (Sizhong 司中), Overseer of Lifespan (Siming 司命), Master of Wind (Fengshi 風師), and Master of Rain (Yushi 雨師); the “earthly spirits” include Earth God (She 社), Grain God 241

Jue Guo

(Ji 稷), deities of the Five Offerings (Wusi 五祀), deities of the Five Mounts (Wuyue 五嶽), and forces presiding over mountains, forests, rivers, and marshes, as well as spirits of the Four Directions (Sifang 四方) and Hundred Things (Baiwu 百物); the “human dead” only includes former kings (xianwang 先王). Affairs pertaining to all the spirits in these three categories are under the administration of Zongbo 宗伯, “Bureau of Lineage,” belonging to the Spring Office. The official in charge of the office is Da Zongbo 大宗伯, “Senior Head of the Bureau of Lineage,” and his duties are summarized as being in charge of the state rites pertaining to the heavenly spirits, human dead, and earthly spirits so as to assist the king to establish and protect the state, and offer felicitous sacrifices to serve the spirits of the state. (Zhouli zhushu, 18.270) There are several aspects of this tripartite typology that need some elaboration. First, the tripartite division of the spirits into heavenly, human, and earthly corresponds to a tripartite dividing of the cosmos spatially, in particular, vertically. Second, positioning human spirits in between the heavenly and earthly ones is parallel to an idea that the human realm is “between heaven and earth”（天地之間） , attested in other texts such as Mozi 墨子 and Xunzi 荀子since the fourth century bce. Such positioning is from the humans’ perspective and it eventually developed into an anthropocentric view, which regards humans as the rightful and necessary center of the cosmos, one of the fundamental views that dominated the cosmology in the following two millennia in China. Seeing the outside world from the viewer’s own perspective is biologically inclined, but situating such a perspective centrally and predicating actions upon it is a cultural practice that is usually ideologically and historically perpetuated. Although there is no surviving contemporaneous practicing evidence for this heavenly-human-earthly tripartite typology having been implemented in practice in the Zhou time, its influence and continuity can be observed in the following Qin and Han times, when the imperial states adapted and gradually institutionalized imperial sacrifices according to the same tripartite typology of spirits, forming a state-sanctioned pantheon, although with changing membership of particular spirits in each category.26 Despite the tendency toward a higher level of bureaucratization, centralization, and unification, neither the pre-imperial Zhou nor the Qin and Han empires had the monopoly on the access to the world of spirits and their perceived power or the absolute control over the ways that their subjects saw and interacted with spirits. In contrast to the general lack of contemporaneous and direct practicing evidence of the state as the primary agent interacting with the spirit world, we have some evidence for that of individuals, ranging from the high elite to the lowest stratum of the ranked society, showing not only non-state typologies of their world of spirits but also how these typologies were operated in practice. For the Warring States period (475–221 bce), our practicing sources for individuals come from the southern state of Chu. A combination of elaborate burial practices and favorable climatic conditions results in rich archaeological deposits in the former Chu regions, including delicate organic materials such as bamboo and silk manuscripts. Among such finds, a previously unknown genre of divination and sacrifice records written on bamboo slips is especially informative for looking at individual and personal typologies of the spirits. Caches of such divinatory and sacrificial records have been found at multiple sites and tombs of different status, including enfeoffed lords, high-ranking ministers, royal descendants, and shi 士, the lowest class of the nobility, an indication of a shared practice among the ranked classes. All the finds are dated to the fourth century bce. They were records of divinations and sacrifices performed by professional diviners 242

The spirit world

and ritualists on behalf of the tomb occupants during their lifetime, addressing unfavorable conditions such as socio-political crisis or physical illness that were believed to have been inflicted by spirits upon their patrons.27 Because they were buried not long after they were produced and then unearthed from datable tombs, they furnish the most convincing contemporaneous practicing sources for looking into the spirit world of these individuals, and to a certain extent, their respective social classes, for the time period. This genre of divinatory and sacrificial records is highly formulaic, and the formula is consistent across the finds, indicating that divinatory procedure must have also been rather similar, if not standardized. Based on the records from Baoshan Tomb 2, the only cache that came to light in undisturbed and intact condition, a typical divination of this kind included two stages. The initial or the first stage was to divine the cause of the situation or condition concerned, resulting in identification of particular spirits as the responsible agents and subsequent proposals of specific ritual measures to propitiate or pacify the agitated spiritual forces. The initial identification and sacrifice proposals were divined again in the second stage, in which the prognostication, almost without exception, was positive and thus confirmed the initial result. Taking Baoshan slips 212–215 as an example, this record was a divination performed by a diviner named Gu Ji on behalf of Shao Tuo, the high minister in charge of legal affairs at the king’s court, in 317 bce. The inquiry was about whether there was any fault and blame in his service to the king in the upcoming year. Gu Ji made the first prognostication as follows: The long-term divination is auspicious; [but] there is slight dissatisfaction with the affairs of the king; moreover, he is in distress. According to the causes, make [the following] sacrifices to remedy them. To eliminate previously [identified] baleful influences, make a requital sacrifice to the Grand One (Taiyi 太一) with one circlet of jade pendant; to Earth God (She 社), Overseer of Lifespan (Siming 司命), and Overseer of Disaster (Sihuo 司禍), each with one small jade circlet; to Great Water (Dashui 大水), one circlet of jade pendant; to the Two Sons of Heaven (Ertianzi 二天子), each with one small jade circlet; to Mount Wei (Weishan 危山), one piece of penannular jade. To eliminate the baleful influences [identified] by Ying Hui (i.e., another diviner), make a requital sacrifice to the Earth God of the Palace (Gong dizhu 宮地主) with one black goat. To eliminate the baleful influences [identified] by Shi Pishang (i.e., again another diviner): in the third month of the autumn, make a requital sacrifice to King Zhao 昭 王 (i.e., the royal ancestor from which the Shao(Zhao) lineage received the name, trad. r. 515–489 bce) with a male ox and treat him to offerings of food and drinks; make a requital sacrifice to the Cultured Lord of Pingyu 坪與文君 (i.e., Shao Tuo’s first lineage ancestor and his great great-grandfather), Governor of Wu, Zichun 郚公子春 (i.e., Shao Tuo’s great-grandfather), Commander of War Horses, Ziyin 司馬子音 (i.e., Shao Tuo’s grandfather), and Governor of Cai, Zijia 蔡公子家 (i.e., Shao Tuo’s father), each with a male pig; treat them to offerings of food and drinks; make a requital sacrifice to the mother [of Shao Tuo] with dry meat and treat her to offerings of food and drinks. This lengthy proposal of sacrifices to a variety of spirits that were identified as responsible for Shao Tuo’s distress was divined again by Gu Ji, and the second prognostication was formulaically affirmative. Although all the second prognostications were consistently positive in all the excavated records of this genre so far, not all the proposals were actually carried out, especially when different diviners performed multiple divinations on the same topic, which was a common practice seen in these records. In the case of this particular divination, however, there are indications that they were indeed performed. On the last slip of this record (slip 215), there is a 243

Jue Guo

6- to 7-centimeter-long blank between the end of the second prognostication and a short note that reads: “[Sacrifices to] the Grand [One], Earth God, Overseer of Lifespan, and Overseer of Disaster, Great Water, Two Sons of Heaven, and Mount Wei have all been completed.” Then after an even longer gap, almost toward the end of the slip, a four-character phrase writes “there was good news within a year.” The blanks between these segments hint that they were written at different times; the content of the two segments further suggests that they were added after the proposed sacrifices were carried out successfully and good news came as the result of resolved spiritual inflictions. Unmistakably practicing in nature, these records are sources par excellence for observing how spirits were grouped together and ritually treated accordingly in actual practice. Scholars have taken notice of the practicing nature of these records and their potential value for reconstructing actually operated, instead of idealized, typologies of the spirits of the time. There are two approaches to such reconstruction. One can be called an “external” approach, which classifies spirits seen in these records using etic categories, such as the aforementioned Zhouli tripartite typology: heavenly-human-earthly (Chen Wei 1999) or analytical categories, such as “natural spirits” (ziran shen 自然神) and “ancestral spirits” (zuxian shen 祖先神) (Peng Hao 1991; Li Ling 2000, and Bing Shangbai 1999 and 2009). The other is an “internal” approach, which looks for emic typological terms in contemporaneous sources to reconstruct what is believed to be the indigenous Chu typology of spirits. One fragment of a bamboo slip from Geling Tomb 1 (dated to the first quarter of the fourth century bce) holds the greatest potential to be such a Chu typology of spirits. This slip (A2–40) was damaged at the both ends, but the survived middle part contains a phrase: “, below, inside, and outside, [all] spirits” 內下内外鬼神. Given the nature of these records, it can only be understood as invoking the entire group of spirits in a divinatory or sacrificial context.The “above” and “below” have been seen as the same as the familiar duo of “heavenly” and “earthly” categories in the Zhouli system. The “inside” and “outside,” by comparison, are more complicated. Chen Wei suggests that they are to further distinguish human spirits—one of the three categories of spirits in the Zhouli system—based on kinship.That is, those who are related by blood are “inside,” and those who are not are “outside” (Chen Wei 2007: 106; Yang Hua 2012: 148–149). However, in the context of Warring States Chu practicing sources, in addition to kinship being one way to divide human spirits into “inside” and “outside” categories, it seems that they can also refer to other factors, such as where the death of a person occurred. Those who died outside the family residence, particularly far away from home, were also called “outside death/dead” (waisi 外死) and “outside human spirits” (waigui 外鬼). This fourfold “above-below-inside-outside” typology is also seen in another genre of excavated practicing sources, rishu 日書 or “daybooks,” predominantly coming from tombs dated to the fourth to the second centuries bce.These almanac-like manuals, also previously unknown in the received tradition, contain a great variety of divinatory prescriptions aiming at helping people avoid or avert misfortunes and conduct daily life activities and rites of passage under favorable conditions by paying attention to the hemerological and cosmological nature of time and space.28 In the earliest find of such rishu texts from Tomb M56 at the Jiudian 九店 cemetery, Jingzhou, Hubei, which is located just outside the city wall of the former Chu capital and dated to around 300 bce, slip 26, which mentions the spirits “above and below,” reads: [Those days] are called yang days, on which myriad dealings are smooth and successful: the lord of the state obtaining good harvests and the small peasants [having crops] ripen four times. With [these harvests and crops], offer prayers and sacrifices to those above and below. The spirits enjoy the offerings and then fulfill their (i.e., the lord and small peasants) wishes. 244

The spirit world

In another find of daybooks from Shuihudi 睡虎地 Tomb 11, dated no later than 217 bce, at Yunmeng, Hubei, a former Chu territory under Qin occupation after 278 bce, there are multiple examples of “outside spirits” mentioned as sources of baleful influences and harms to the living. For instance, in what is called the “Twelve-Branch Divination” (shierzhi zhan 十二 支占) section of Daybooks B (乙種), in a highly formulaic fashion, each of the twelve earthly branches (dizhi 地支) is given a list of the prognosticative values for directions, times in the day, and various occasions such as catching thieves, making purchases, and diagnosing illness. In particular, the illness diagnosis for each branch, on which the illness begins, has a description of the development of the illness, the time of recovery or death, the direct cause of the illness, and the inflicting spiritual agents of the illness. Among those identified spiritual agents, many are deceased family members such as “great-grandfather” (gaowangfu 高王父), “grandfather”(wangfu 王父), but there are also “outside human spirits” (waigui), “outside human spirit in the generation of the father” (waigui fushi 外鬼父世), “outside dead in the generation of the mother” (mushi waigui 母世外死), and “outside human spirit in the generation of the brothers” (waigui xiongshi 外鬼兄世).29 Clearly, these “outside human spirit/dead” of father, mother, and brother’s generation are kin; therefore, here “outside” most likely refers to their death occurring outside their normal residence or home place, which was another reason that they were especially susceptible to become trouble-causing spiritual agents. In actual practices, as shown in the Chu divinatory and sacrificial records, which contain more than one hundred different names of spirits, many of which are not seen in the transmitted sources,30 not only were some spirits unable to be unambiguously grouped under one subcategory, but the order in which different subcategories of spirits appeared also by no means always complied to the sequence of “above-below-inside-outside.” For instance, although the order in the aforementioned example (Baoshan slips 212–215) is in accordance with “above-belowinside-outside,” many other examples from the same Baoshan cache and the Geling cache do not. Instead, predynastic or lineage ancestors (i.e., the “inside” subcategory), such as “three Chu forebears” (san Chu xian 三楚先) or King Zhao 昭王, appear before those belonging to the “below” subcategory, such as Master of Earth (dizhu 地主) or the High Hill (gaoqiu 高丘).31 If we consider that a typology is not simply a schematic classification but also implies a hierarchy among the classified, then the discrepancy between a highly abstract typological expression such as the fourfold “above-below-inside-outside” and the seemingly chaotic grouping and nonconforming ordering of spirits and spirit groups in the divinatory and sacrificial records points to a reality that the nature and status of spirits is not static but contingent, and largely dependent on the specific situations and individual causes for which spirits were involved and called upon. In this sense, these Chu practicing sources not only provide knowledge of numerous previously unknown spirits but also make us be more aware of and pay attention to the local and individual practicing context so that we do not gloss over such nuances and can avoid forcing the practicing sources to fit discoursing abstractions. The localization and personalization as well as the contingent appropriation of general and established typology of spirits became even more manifested in the following Qin and Han periods, when the number of spirits also drastically increased. What has been found in a Western Han burial is a good case in point. The Huchang 胡場 Tomb 5 was a modest shaft tomb, dated to 70 bce. It was excavated at a cemetery in Yangzhou, Jiangsu, in 1981. The cemetery belonged to a certain Wang clan, a local powerful family residing in the Kingdom of Guangling 廣陵, ruled by one of Emperor Wu’s sons, Liu Xu 劉胥 (r. 117–54 bce). Based on the personal seals and documents inscribed on wooden boards, the primary tomb occupant was identified as Wang Fengshi 王奉世, who may have worked as a scribal clerk at the Guangling court before his imprisonment. The skeletal examination shows that he had suffered head trauma and died 245

Jue Guo

around age 30.32 His unfortunate, though not uncommon, turn from a low-level official to a punishable convict, especially the shameful death as a prisoner, under normal circumstances, would have deprived him of a proper burial, dictated by the social taboos and prohibitions for providing for and including a disgraced member in the family afterlife landscape. As many Han accounts have shown, those convicted and punished were strongly stigmatized when it came to matters of mourning, funeral, and burial.33 Archaeological finds of the so-called cemeteries of convicts and statute laborers (xingtu mudi 刑徒墓地) near the mausoleum of the First Emperor of Qin (r. 221–210 bce), the Yangling Mausoleum of Emperor Jing of Han (r. 156–141 bce), and at the Eastern Han capital Luoyang (the cemetery dated to ca. 62–172 ce), all point to such common final destiny for the convicts.34 However, not only did Wang Fengshi receive a proper and well-furnished burial suitable to his pre-conviction status of being a minor official, he was also buried in his family cemetery. In addition to the conceivable local influence and privilege that the Wang family may have had in retrieving Wang Fengshi’s body, two inscribed wooden boards that were buried with Wang Fengshi revealed that an important ritual measure intimately involving the regional and local spirits was also taken to rectify and remedy the undesirable nature of Wang Fengshi’s death and ensure his return to the family cemetery. A total of thirteen wooden boards were found in Wang Fengshi’s tomb, among which five were still legible. In addition to one “scribal record” of various official travels and dealings from the year before Wang Fengshi’s imprisonment and one “inventory of sacrificial items” that may have been used at Wang Fengshi’s funeral, the other three boards are our primary concern here. A set of two boards contains what now is commonly known as “Communication to the Underworld” (gaodishu 告地書) or what I call a funerary ‘Relocation Document’ (yi dixia shu 移地 下書), a genre also only coming to light through archaeological finds (Guo J. 2018 forthcoming; 2013: 62–67; 2011: 98–104). Generally speaking, the format of these funerary “Relocation Documents” resembles the Han official transit document (zhuan 传) that was required for people who needed to travel within the empire and conforms to the legal requirement stipulated in the statutes for the living people who relocated or were moved to another place, only in this case, the destination was the underworld. Other specimens of this genre, mostly found in the second-century bce Western Han tombs from the Jiangling region in Hubei, were predominantly focused on specifying the ownership of the property (i.e., grave goods), with which the tomb occupant was buried, or on any privilege, such as exemptions of tax or corvée labors, that the tomb occupant had in the living world and hoped to continue to enjoy in the afterlife. Compared to those specimens, the Huchang document had a different emphasis, as it reads: In the forty-seventh year (i.e., 71 bce), in the twelfth month, in which the first day was bingzi (the 13th day), on xinmao (the 28th day), Deputy to the Chief of Palace Construction [in charge of convicted laborers] of Guangling [Kingdom], Zi, respectfully reported to the Lord of Earth:Wang Fengshi, a man from the Shi Village of Guangling, had a criminal offence and was imprisoned. His incident was concluded and he is registered [for taxation and corvée labor] in [his] former county, township, and village. [Wang Fengshi] will personally bring the dispatching document to report [to you] and move to the Cypress Hill. In the forty-seventh year (?), Prison Clerk Zu wrote [this document]. [Upon receiving this document], act in accordance with the statutes and decrees. After giving the basic census registration information of Wang Fengshi, the document, written in the voice of an official named Zi who applied for this particular relocation document on 246

The spirit world

Wang Fengshi’s behalf, vaguely alluded to his legal predicament but emphatically stressed that his status was officially restored. This is because according to the Han laws, a convict would be stripped of his or her previous status before being sentenced to various punitive labors. This statement therefore was essential in bringing Wang Fengshi back to be buried properly in his family cemetery. In addition, his restored status would also be used by the underworld officials, to whom this “relocation document” was addressed, to register Wang Fengshi with the underworld bureaucracy accordingly, as the very last part of the document requests. In this communication to the underworld, the recipient of this “official” document was addressed as “Lord of Earth,” a generic way to refer to the underworld bureaucracy. Unlike other cases in which there was usually an inventory of the possessions that accompanied the tomb occupant to the underworld, in addition to the highlighted statement of Wang Fengshi’s alleged release and status restoration, Wang Fengshi also had another wooden board buried with him, on which a total of thirty-three names of spirits were inscribed. Notably, this particular board was not “attached” to the “relocation document” as the inventory usually was; instead, it was placed on the top of the inner coffin, the closest to the body of Wang Fengshi, indicating it may have had a different intended function. Although the ink has faded quite severely, the remaining legible part has seven registers, each including two to seven names of various spirits as follows: Register 1 Lord of the Jiang River 江君 Numinous Lord of Shangpu 上蒲神君 Great Lordly King of Gaoyou 高郵君大王 Lord of Man 滿君 Lord of Luxiangfan (?) 盧相氾君 Grandparents Inside and Outside 中外王父母 Numinous Soul 神魂 Register 2 Azure Heaven 蒼天 Heavenly Duke 天公 Register 3 Great Old Man 大翁 Praying [Place] for Chief Supervisors of Zhao 趙長夫所禱 Huai River 淮河 Lord of Yu 堬君 Numinous Earth God of Shi Village 石里神社 Lord Peng of Chengyang 城陽莑君 Register 4 Village Master of Shi Village 石里里主 . . . Prayer [Place] for the Spring Lady . . . of the Palace 宮春姬X 之X禱 King of You 右王 King of Wu 吳王 King of Jing 荊王 Numinous King of Fanyang 氾揚神王 Empress Dowager Chui 大后垂 Register 5 Forbidden Pond inside the Palace 宮中禁池 Numinous Earth God of XX 口口神社 Register 6 247

Jue Guo

Lord Presiding over the Road 當路君 Master of Jing 荊主 Lord of Xi Hill 奚丘君 On the Water 水上 Lordly King of the Palace 宮君王 Earth God of X 口社 Register 7 Earth God of the Bureau of Construction of the Palace 宮司空社 X Town 口邑 Despite the tentative readings and the difficulty to discern the exact logic behind the grouping, with the exception of Azure Heaven and Heavenly Duke, which may be more generic and universal, the rest are unmistakably regional and personal in nature. For example, the Jiang River and the Huai River were the two main local waterways; Gaoyou and Chengyang were two towns in the neighboring regions; and the territory of the Guangling Kingdom used to be that of the State of Chu in the Warring States period and of the Kingdom of Wu in the early Western Han, which explains why the kings of Jing (i.e., Chu) and Wu were included in the list. Similarly, Numinous Lord of Shangpu, Lord of Man, Lord of Yu, King of You, and Lord of Xi Hill are likely local spirits of nearby and surrounding places, waters, and mountains, even though we cannot precisely identify them anymore. But what makes this list remarkably personal and specific to Wang Fengshi were those spirits clearly presiding over his registered native place: Numinous Earth God of Shi Village and Village Master of Shi Village, and those in charge of his most immediate environs around the court of Guangling Kingdom such as Forbidden Pond inside the Palace, Lordly King of the Palace, and Earth God of the Bureau of Construction of the Palace. At the very end of this list, there was a graph sai 塞, which scholars agree as referring to the reciprocation or a requital ritual to fulfill earlier promises to the spirits, by offering sacrifices if the requests or the wishes were realized (saidao 賽禱 or saishen 賽神). Its appearance at the end of such an extensive list of spirits of local significance and personal connection to Wang Fengshi can suggest two possible scenarios. First, these spirits were propitiated to avert Wang Fengshi’s misfortune so as to have him buried without spiritual interferences or afflictions, given that the burial must have had to move his body from the prison ground to his family cemetery and thus involved the spirits of both locales. After this relocation and burial took place, all the enlisted spirits were offered the requital ritual. This board was a record of such interaction with these spirits and was included as part of Wang Fengshi’s relocation package to the underworld. It was placed on the top of the inner coffin, close to the body, so that as Wang Fengshi was lowered into the grave and sent off on his way to the underworld, which these spirits were also part of, this record would be a proof or pass for Wang Fengshi to arrive at his destination safely. In the second scenario, this board could be a reminder to Wang Fengshi that he should perform the requital ritual to these spirits as a thanksgiving when he arrived in the underworld safely. Regardless, it is clear that for both the “relocation document” and this list of spirits, although the genre in both cases was not unique, the specific content and their intended ritual efficacy were for Wang Fengshi and his personal circumstance alone.

Discoursing sources What the practicing sources in general do not tell directly is how the spirit world was abstractly conceptualized and discursively utilized. It was only since the second half of the first millennium 248

The spirit world

bce that written sources of what we call discoursing nature began to become available, marked by the presence of a concept of guishen 鬼神 as a compound term referring to spirits in general35 in historical narratives and philosophical expositions.The “sudden” appearance of an array of diverse discourses indicates that it is likely a result of an accumulation of discoursing spirits. Therefore, the almost millennium-long temporal gap between the appearance of a mature writing system and the written sources of discursive nature may simply reflect the lacuna in the source preservation. In the remainder of this chapter, I will focus on the ways that guishen spirits as a trope were used in political persuasion, moral dictations, personal cultivation, and philosophical reflections and argumentation preserved in the discoursing sources between the fifth century bce and the second century ce. At the same time, it is important to keep in mind that the continued and more diversified practices of divination, ritual, and sacrifice in general as well as the changes in political, social, and economic structures at large must also have been catalytic factors for such increasing and intensified intellectual inquiries about spirits, which in turn had an impact on the actual practices. Speaking of spirits in grouping or generalizing terms beyond individual entities already appeared in the Shang divination records. Shi 示, a nonspecific term for spirit, was used, but only in the contexts referring to clusters of ancestors, such as dashi 大示 “Greater Ancestors,” xiaoshi 小示 “Lesser Ancestors,” yuanshi 元示 “Primary Ancestors,” shangshi 上示 “Exalted Ancestors,” zhongshi 中示 “Middle Ancestors,” or shishi 十示 “Ten Ancestors” (Keightley 2004: 16–20). Later in the Western Zhou bronze inscriptions, phrases such as huangshen zukao 皇神祖 考 “August spirits of the ancestors” (JC 04448, Du Bo xu-container 杜伯盨) were also in use, referring to ancestral spirits in general. Another term shangxia 上下 “Above and Below” was also present in Shang oracle-bone inscriptions36 and Western Zhou bronze inscriptions. Although the precise referent(s) of shang (above) and xia (below) in particular cases are still debated, their use together as a general term for all spirits seems certain (Yang Hua 2012: 143). Generic terms like these in the early practicing sources are indicators of conceptualizing and abstracting spirits on a level beyond the specific and individual entity, which would have been necessary for developing theories about and discourses around spirits. The earliest extant written discourse on the role guishen spirits played in rulership and statecraft is seen in the Zuozhuan 左傳 or Zuo Tradition, a compilation primarily of semi-historical narratives whose multiple textual layers date between the eighth and fourth centuries bce.37 If, as Stephen Durrant suggests, Zuozhuan reflects the transitional Spring and Autumn period when the lineage-based and kinship-centered hereditary aristocratic rule of the Western Zhou began to be challenged by an emerging meritocracy (Durrant, Li, and Schaberg 2016: xxvi–xxvii), similar changes can be discerned in the prescribed attitude and designated duty of a ruler to guishen, as recorded in several speeches of the ministers delivered to their rulers. Serving guishen was already an understood part of being a ruler, a cultural consensus of earlier origins, as shown in the practicing sources introduced earlier, but a clear articulation of such a duty was first seen in the Zuozhuan, such as “a lord presides over the [Temples] for Earth and Grain, attends to sacrifices and cults, supplies for the nobles and commoners, serves guishen, and participates in regional meetings of the lords and at the king’s court” (Lord Zhao Year 7).38 The inscriptional and material records of divination and sacrifice practices in the Shang and Western Zhou periods may give the impression that having proper and bountiful offerings was essential in gaining the spiritual approval and enlisting their blessings. Such an understanding was certainly still present in the Zuozhuan. But a mere display of opulent offerings to the spirits was considered inadequate at this time, as the general discourse had shifted toward seeing guishen as both powerful agents and moral arbiters in judging personal conduct as well as the governing performance of a ruler. In Lord Huan Year 6, a conversation between the ruler of Sui and his able minister, Ji Liang 季良, in the middle of a pending conflict with Sui’s stronger neighbor, 249

Jue Guo

Chu, exemplifies such a shift.When the ruler of Sui was confident that his wholesome, well-fed, and fatty sacrificial animals and brimful of grain offerings would surely secure the blessing of guishen and gain their favorable support in his military endeavor against Chu, Ji Liang voiced his opposition by elaborating on the proper nature of sacrificial offerings in great detail: I have heard that a small [state] can resist a big [state because] the small [state] is proper, but the big [state] is irresponsible.What is called proper is being dedicated to the people and trusted by the spirits. That the ruler thinks of benefiting the people is being dedicated, and that the ritual invocators and scribes use faithful utterance [at the sacrifices] is being trustworthy . . . People are the host for the spirits.This is the reason that a sagacious king first fulfills [the needs of] the people and then devotes his labor to the spirits. Therefore, when [he] presents the sacrificial animals [to the spirits] proclaiming that [the animals are] whole, strong, fatty, and well-fed, he is saying that the resources of the people are all preserved. That is to say that their (i.e., the people’s) livestock are strong, fat, and plentifully reproducing; they are free from suffering itches and insects; they are whole, fat, and sufficiently ample.When [he] presents containers of sacrificial grains [to the spirits] proclaiming that they are pure and bountiful, he is saying that there was no harm in the three seasons [of agricultural work], and his people are harmonious and the harvest is plentiful. When he presents fermented liquor and sweet wine [to the spirits], proclaiming that they are fine, clear, and pure, he is saying that the superior and the inferior [in his state] both have fine virtues and none has a transgressive heart. What is said to be aromatic and fragrant is that there are no slander and vice [in his state]. Therefore, devoting oneself to the work of the three seasons, cultivating and inculcating the five teachings, and endearing and caring for one’s nine kin-relations lead to the pure sacrifices. It is in this way that the people are harmonious and the spirits bestow the blessings. Therefore, taking an action (i.e., against the Chu) results in success. Now the people each have their own heart and the guishen are short of hosting. Although you, the Lord, alone have bountiful [offerings], what blessing will that bring? Ji Liang’s lengthy explanation makes it clear that guishen spirits do not blindly accept and consume the fine offerings but rigorously judge them by their sources.When the ruler and his ritual specialists present the offerings and prayers, they are in fact presenting the material manifestations of the condition of the state and the well-being of the people, which cannot be falsified.The ability of a ruler to secure the welfare of his people and upkeep his state—the two conditions necessary for properly providing for the spirits—is called de 德,“virtue,” and virtue is a yardstick with which the spirits use to measure the fitness of a ruler and of his offerings.Yan Ying 晏嬰, the famous minister of Duke Huan of Qi 齊 桓公 (trad. 685–643 bce), expresses precisely such a view (Lord Zhao,Year 20): “When a virtuous lord does not disregard [his duties] within and outside [his state], when neither the superior nor the inferior [in his state] have any resentment, when his actions do not incur opposition from circumstances, when his ritual invocators and scribes can be faithful [in their prayers] without any shame in their heart, the guishen spirits accept and enjoy his offerings and his state will receive their blessings. Similarly, the lesser known Gong Zhiqi 宮之奇, a minister to a minor state called Yu, remonstrates to his lord, “guishen spirits are not true kin of the living and they only adhere to virtue” (Lord Xi, Year 5). These similarly argued statements from various states that dispersed in the Zuozhuan make it reasonable to say that this view of guishen as a moral agent safeguarding the 250

The spirit world

welfare of the people and the prosperity of a state was a shared discourse as well as an esteemed rhetorical tool of the ministers at the various courts of the Spring and Autumn period, or at least in their Warring States recounting and reimagination. It is worth noting that the existence of guishen and their moral authority was largely taken for granted in the rhetorical and didactic usages of guishen in the Zuozhuan narratives. This is, however, by no means an uncontested view. On the contrary, the existence of spirits and the nature of their existence had been a matter of contemplation and debate since the fifth century bce. The most articulated argument for an unambiguous and necessary existence of spirits is in the Mozi 墨子,39 in particular the “Illuminating gui” 明鬼 chapter. Master Mo first describes a chaotic world after the passing of the sage kings in the following alarming words: When it came to the time after the passing of the sage kings of the Three Dynasties, the world lost its righteousness and the various lords took might as right.Those who are living as rulers and ministers, superiors and the inferiors, are no longer gracious and loyal; father and son, elder and younger brothers are no longer affectionate, filial, brotherly, respectful, upright, and kind. The noble and officials do not attend diligently to governing, and the humble and commoners do not attend earnestly to their work. People are giving themselves to excessiveness, violence, plunder, and chaos.Thieves and bandits with weapons, blades, poison, water, and fire hold up innocent travellers on the highways and the bypaths, robbing them of their carts and horses, clothes and furs, to benefit themselves. Master Mo’s solution to resolve the chaos firmly rest upon the necessity of recognizing the existence of guishen spirits, and more importantly, recognizing their power and authority in judging and passing on reward and punishment for people’s conduct, as it is clearly put in the following: This is all because of the doubt about whether guishen exist or not and the ignorance of the capability of guishen to reward the worthy and punish the bad. Now if everyone in the world can be led to believe that guishen can reward the worthy and punish the bad, how can the world be chaotic? What makes the Mozi exposition distinct from the Zuozhuan narratives is that Master Mo does not simply put forward an authoritative standpoint; instead, an argument was formed and evidence provided. Therefore, the rest of the “Illuminating gui” chapter is to present a variety of evidence, each instance of which is given as an answer to a specific challenge posed by the interlocutor, “those who insist on the nonexistence of spirits.”The range of evidence goes from individual and collective testimonies of the contemporary witnesses of spirits, transmitted and documented “historical” incidents involving spirits, to the recorded words and deeds of the sage kings toward the spirits. As many scholars have pointed out, Mozi promotes a strong utilitarian agenda for restoring a socio-political order that is based on a clear hierarchy, with guishen spirits on the top as the overseer and a control mechanism for regulating human behaviors toward maximizing the overall welfare, especially the material well-being, of the society. The insistence on the existence of guishen and their ability to reward and punish therefore should be seen as a logical necessity for the Mozi discourse. This kind of pragmatic attitude, rather than a metaphysical conviction, toward guishen can be seen in the response of Master Mo to those who deny the existence of spirits and thus consider sacrifice as a waste of material wealth: If the guishen who are entreated exist, then [people’s deceased] fathers and mothers, elder sisters and elder brothers (i.e., the spirits) are served to eat and drink the offerings. 251

Jue Guo

Is this not a great blessing? If the guishen who are entreated do not exist, then it would seem to be [wastefully] spending wealth on the sacrificial fermented liquor and sweet wine as well as brimful of sacrificial grains. As an expense as it is, it is not like throwing [the wealth] into the ditch and gully and discarding them. Those from the family and clan and those outside [the family and clan] in the towns and villages can all come to the feast to eat and drink. Even if the guishen who are entreated do not exist, such [sacrificial offerings] can still unite the joy, gather the multitude, and solicit closeness among the people in the towns and villages. This is a rare occasion in the Mozi. In general, Master Mo strongly opposes waste of material wealth such as in lavish funerals and for ritual ornaments including music. However, here the overall benefit of spending on sacrifices to spirits is considered greater, even if they do not actually exist. The interlocutor—“those who insist on the nonexistence of spirits”—in the Mozi is therefore not merely a rhetorical device. Given how ardently and rigorously that Master Mo tried to argue for the existence of spirits, such doubts about or denial of the existence of the spirits must have been evident and significant enough to give rise to such efforts in the Mozi. Between the two opposite views on the existence of the spirits, there was also the wellknown teaching of Confucius (ca. 551–479 bce): “Revere the spirits but keep them at a distance” (Lunyu 論語6.22). In the Lunyu—the collected sayings and teachings of Confucius compiled over time by generations of his disciples and followers—Confucius, on one hand, was recorded to have spoken little about spirits (Lunyu 7.21) and even criticized a Lu minister named Zang Wenzhong for providing a turtle an extravagant house and foolishly believing in its spiritual foresight (Lunyu 5.18). On the other hand, he recognized the importance of proper sacrifices, with a marked emphasis on the correspondence between the outer performance and inner sincerity. One should not make offerings to spirits to whom one is not supposed to sacrifice (Lunyu 2.24), but when one does sacrifice to proper spirits, one should do so with a sincere attitude and inner reverence, as if the spirits were actually present at the sacrifice (Lunyu 3.12). The saying of “as if ” has prompted debate about Confucius’ attitude toward the existence of spirits; however, it seems clear that for the Confucius in the Lunyu, the heart of the matter for sacrifice is not the existential state of the spirits but taking sacrifices to spirits as one of the opportune occasions to cultivate the proper behavior and genuine emotion. In the Lunyu, Confucius was neither a denier nor an agnostic of spirits but an advocate for shifting the attention that was fixated on, as Mozi argues, the existence of the spirits as a necessary external control mechanism in mediating human affairs and regulating human conduct through reward and punishment, to cultivate proper human disposition and behavior through self-realization and internal motivation. The contrast between the inner cultivation that rendered the existence or presence of spirits nonessential, if not irrelevant, and the outer regulation that ultimately relied on spiritual enforcement was then seen as reflecting the different levels of understanding the spirits and practices such as sacrifice, which in turn place people into different categories, ranging from the most cultured sage to commoners of little cultivation. Such a graded hierarchy of understanding and cultivation began to be seen in the Xunzi 荀 子,40 attributed to a follower of Confucius, Master Xun, who was active in the third-century bce. In the “Discussions of Rites” 禮論 chapter, it concludes: Sacrifices [express] emotions of remembrance and longing, [are] the highest [form] of loyalty, trust, love, and reverence, and [embody] the finest of the ritual observance and cultured bearing. If one were not a sage, they (i.e., sacrifices) cannot be fully understood. Sages clearly understand them; educated gentlemen tranquilly carry them 252

The spirit world

out; officials keep them maintained; and commoners accept them as customs. To the gentlemen they are part of the Way of Man; to the commoners, they are pertaining to the affairs of the spirits. As of the second century bce, a divide between a true understanding of the spirits-related affairs was propagated, indicating higher intellectual complexity and socio-political fitness of the sages, the ideal rulers, vis-à-vis the superficial, if not false, understanding of the same phenomenon among the common people, the ruled. In the “Boundless Discourses” 氾論 chapter of the Huainanzi 淮南子, an encyclopedic collection of essays pertaining to various aspects of rulership compiled at the court of King of Huainan, Liu An 劉安 (179–122 bce), and submitted to the throne in 139 bce (Major, Queen, Meyer, and Roth, 2010), the sages (i.e., ideal rulers) make use of their unique knowledge of spirits and devise prohibitions and regulations to guide and rule the less enlightened masses. For instance, the sages knew that people who encounter the windy qi found in doorways would get sick, but commoners held the belief that “the spirits will step on the heads of those who sleep on a doorsill.” What the masses believed as the spiritual interventions and undertakings are “mere expedients” viewed and devised by the sages because the sages had clearer and higher-level understanding of the true nature and function of the phenomena that the commoners lacked (Csikszentmihalyi 2006: 123).The following passage clearly demonstrates this contrast: Since they (i.e., the common people) hear and observe them (i.e., strange happenings that were believed to be spiritual undertakings) infrequently, their knowledge of these things is shallow. The strange things of the world are what sages alone see; the reversions of benefit and harm are what the knowledgeable alone understand and comprehend. Those that are similar, different, suspicious, and doubtful are what confuse and befuddle the common people of our time. It is because [sages] see things that cannot be made known within the [Four] Seas and hear things that cannot be made clear to the Hundred Surnames (i.e., common people) that [they, the sages] make use of the spirits as well as inauspicious and auspicious omens to establish prohibitions for them (i.e., the common people) and generalize shapes and expand categories, and alter appearances [of the things] for them [to follow].41 These discourses on spirits share one common discursive feature, which is that they all treat guishen or spirits as a self-contained whole and use it as a rhetorical trope. Despite their different stance on the existential state of spirits, it was the utility of spirits in arguing for their respective political, social, moral, and didactic propositions that was theorized accordingly. Since the third century bce, in addition to continuing to highlight the regulatory role of spirits or the knowledge of spirits played in political, social, and personal affairs, the discourses on spirits were extended to also include discussions of the nature of spirits. The most well-known exploration of the metaphysical nature of spirits can be seen in the Lunheng 論衡, or Balanced Discussions, a collection of eighty-five thematic essays of Wang Chong 王充 (ca. 27–90 ce), one of the most original and critical minds of his time. During Wang Chong’s lifetime, the belief that the human dead can turn into willful and harmful gui was widely held, and many practices such as consulting divinatory almanacs, observing everyday taboos pertaining to spirits, and performing propitiatory and exorcist rituals were popular. Wang Chong was a fervent opponent to such beliefs and practices, and he wrote at length debunking their falsehood. Although not entirely devoid of the socio-political undertones that were prevalent in the earlier discourses and by no means strictly consistent, when it comes to guishen, Wang 253

Jue Guo

Chong insisted on building his critique upon the metaphysical origin of guishen and related phenomena such as yaoxiang 妖祥 (inauspicious and auspicious phenomena) from a materialistic and naturalistic perspective. Wang Chong’s argument is premised upon a fundamental conviction that the cosmos or the world, of which humans and other beings are part, is made of the same material qi 氣 through the same process that can be described as “self-becoming” (ziran 自然): The way of heaven is not to interfere [with things].Therefore, spring is not for the sake of planting; summer is not for the sake of growth; autumn is not for the sake of maturing; winter is not for the sake of storing. When the yang qi, of its own accord, comes out, things spontaneously germinate and grow. When the yin qi, of its own accord, arises, things spontaneously mature and get stored.42 From this “self-becoming” or nature’s point of view, Wang Chong argues that “[the phenomena] that inauspicious qi forms gui and gui resembles human form is the way of self-becoming (i.e., nature) and not the doing of some agent.” If so, then as part of the same self-becoming world in which each being acts out of its own accord, humans and things are fundamentally the same; both are wu 物, “material.” If things or non-human beings are naturally born, grow, mature, die, and decay without taking a form and becoming willful and harmful gui,Wang Chong challenges in the “Discourse on Death” 論死, why human beings, of the same material nature and developing course, alone become gui. His argument does not stop merely at this analogy or only on a rhetorical level but goes further to develop his rejection of the popular conceptions of guishen that was construed as the conscious and agentive postmortem form of the dead, by looking specifically at the physiological mechanism of human life and its material change at and after death: What keeps humans alive is the essential qi. At death the essential qi is extinguished. What carries the essential qi is the blood in the veins and arteries. When a person dies, the blood in the veins and arteries is exhausted. [When the blood in the veins and arteries] is exhausted, then the essential qi is extinguished. [When the essential qi is] extinguished, then the form and body decays. [The form and body] decays and becomes dirt and earth. As such, Wang Chong describes this process from life to death of humans primarily as a material transformation from a living body in the form of a human, animated and vitalized by the essential qi in the blood circulating in the body, to a decaying body at and after death when the blood that carries the vitalizing essential qi ceased to circulate, and eventually dissolving into dirt and earth, completely absent of blood and essential qi, and no longer resembling a human form. Such a transformation, he concludes, leaves no room in the material sense for the dead to become an agentive gui. However, Wang Chong’s theory of guishen is different from those that simply deny their existence, in that although Wang Chong strongly argues that there is no metaphysical basis or empirical evidence for the belief of guishen as agentive entities, he does not simply and lightly dismiss guishen as natural phenomena, which should be understood and explained in material and naturalistic terms. In other words,Wang Chong argues that the way guishen were represented in popular beliefs and practices were simply misconceptions and misnomers based on certain natural phenomena: Decay [results in] dissolution and disappearance, diffusing, murky, and invisible. This is why calling it guishen . . . Guishen is the name for what is diffusing, murky, and invisible. 254

The spirit world

When human beings die, their essential shen rise up to the heaven and their bodily bones return to the earth. This is why we call them gui. Gui means “to return.” Shen is what is diffusing, murky, and shapeless. Some say guishen are the names for yinyang. The yin qi welcomes things to return, therefore it is called gui. The yang qi leads things to be born, therefore it is called shen. Shen means “to extend.” [Qi] extends and returns without stopping, from the end again going back to the beginning. Humans take shen (i.e., extending, yang) qi to be born, and when they die, they again return to shen qi. Yinyang is called guishen; when humans die, they are also called guishen. Qi giving birth to humans is like water becoming ice. Water congeals, becoming ice. Qi congeals, becoming humans. Ice thaws, becoming water. Humans die, returning to shen. Its name being shen is like when ice thaws, its name is changed to water. People see that the names are different, and then say that it possesses consciousness and is able to take form and harm people. This is to argue for a position without evidence. In the spirit of getting to the bottom of the matter,Wang Chong not only points out what he considers as the misconceptions of guishen, he also provides explanations for how such misconceptions originate and why they persist. For instance, he attributes the claim of seeing gui and suffering their attacks to a distorted psychological state when people fall ill. Although Wang Chong built his arguments on a materialistic and empiricist foundation that was not widely shared in his time, his social critique of the popular beliefs and practices pertaining to guishen spirits expressed moralistic and pragmatic concerns similar to those of his more prominent and powerful intellectual peers, the Confucian scholar-officials. In the “Response Essays” 對作, he explicitly explained that in writing chapters such as “Discourse on Death,” “Discussions of gui,” and “The Falsehood of Death,” he hoped to “make people be frugal in funeral and burial” because if he can “illuminate that the dead are not conscious and cannot become gui,” then those who read his writings, once they understood his arguments, would refrain from lavish burials and become economical in their expenditures. In “On Exorcism” 解 除, he similarly said: People do not cultivate their conduct but [only] make the sacrificial offerings bountiful. They do not revere their superior but fear their spirits. They blame spiritual inflictions for unfortunate death and disasters, saying that the inflicting spirits have not been caught. Once having caught the inflicting spirits and carefully prepared the offerings, if the disasters still are frequent and not stopping, they blame the sacrifices for it, saying that the sacrifices are not [offered] with reverence. In all discussions of exorcism, exorcism is not beneficial; in the discussions of sacrifices, sacrifices cannot make up [for the disasters]; in the discussions of shamans and invocators, shamans and invocators are powerless. The fact that [the key] eventually lies in people, not in spirits, and in virtue, not in offerings, is clear! Despite Wang Chong’s shared concerns with the ruling elite and official scholars, precisely because he was not in any politically or socially powerful position, his extensive writings and his argumentative style on the topic of guishen bear particular discoursing significance.

Conclusion Three kinds of sources—presenting, practicing, and discoursing—pertaining to the world of spirits in early China are surveyed here separately. However, it is paramount to always keep in mind that, 255

Jue Guo

as a result of the modes of communicating and interacting with a perceived world of spirits, not only are the three kinds of sources not mutually exclusive, they necessarily inform one another in the realm of ideas and constitute one another in practice. These three modes should be considered as coexisting, even if not always explicitly. The seemingly linear progression of their “appearance” should be attributed to source preservation rather than a developmental evolution. Given the diversity within the world of the spirits in early China and the variation of its portrayals in excavated and transmitted texts, it is worth noting that there are similarities in the ways spirits are presenced, their presence is communicated with, and then both processes are conceptualized and discussed.Yet perhaps more interesting is the way that the balance of these three types of sources that treat the spirits shifts over time. In other words, while the spirit world may have been a timeless and stable aspect of early China, the concerns and techniques of those who acted upon and wrote about shifted along with changes in the culture and institutions in early China.

Notes 1 I italicize these three categories throughout the chapter to emphasize their typological function. 2 See Pongratz-Leisten and Sonik’s usage of “presencing or presenting,” their abbreviated version of Gumbrecht’s theory of the “production of presence” (Pongratz-Leisten and Sonik 2016: 10). 3 The archaeological reports and studies about these two cultures are primarily published in Chinese. For a general introduction and summaries of main finds and issues in English, see Christian E. Peterson and Lu Xueming, “Understanding Hongshan Period Social Dynamics” (Underhill 2013: 55–80) and Qin Ling, “The Liangzhu Culture” (Underhill 2013: 555–573); also Liu and Chen 2012: 172–183; 236–242. 4 For the archaeological site report, see Liaoningsheng wenwu kaogu yanjiusuo, 1986. 5 By the time of the Late Hongshan period, people were fully sedentary. Although they may not have entirely subsisted on farming, agriculture was already certainly practiced and villages had appeared (Liu and Chen 2012: 174–178). 6 It should be noted that oracle-bone inscriptions are not the only written sources from the Late Shang. Toward the very end of the Late Shang, cast bronze inscriptions also appeared; however, not only are the majority of Late Shang bronze inscriptions extremely short, the number of the inscribed bronzes is also not comparable to that of the oracle-bone inscriptions (Keightley 1978: 134; Eno 2009: 41, n.1). 7 Noninscribed scapulae were widely found at Anyang and other Shang sites, including villages such as the one at Guandimiao 關帝廟, Henan (Henansheng wenwukao yanjiusuo 2008: 45). Inscribed bones and shells were also found outside Anyang, such as at Daxingzhuang 大辛莊, Shandong, but in much smaller numbers. (Eno 2009: 50–51; 98) 8 Keightley writes “the Shang diviners believed that the sound, shape, and speed of the stress cracks that they formed in the bones and shells by the application of heat were a message sent by the ancestors” (Keightley 1998: 797–798). This well encapsulates the uncertainty in understanding of the exact mechanism of such pyromantic divination. 9 In Zhouli 周禮 or Rites of Zhou, a text dated to the fourth to the third century bce, the Grand Diviner (Dabu 大卜) in the idealized Zhou bureaucracy was in charge of “methods of three cracks” (sanzhao zhifa 三兆之法), that is “jade cracks” (yuzhao 玉兆), “tile cracks”(wazhao 瓦兆), and “plains cracks” (yuanzhao 原兆), whose “bodies of warped cracks all having one hundred and twenty [forms]” (其經 兆之體，皆百有二十). 10 Sarah Allan has suggested that the shape of the turtle shells may have symbolized the Heaven and the Earth and thus were attributed with special powers (Allan 1991: 104–111). Keightley acknowledges such possibilities but points out that the analogy between the shape of the turtle shells and the cosmic model of Heaven and Earth was only known in later texts and cannot be demonstrated in the Shang oracle-bone inscriptions (Keightley 1999: n. 21; 245). 11 Keightley includes a crack-making process in the “Preamble” to his Sources of Shang History that is worth reading as long as one is keeping in mind that it is an “imaginative reconstruction,” especially the details such as the location and specific ritual procedures (Keightley 1978: 1–2). 12 For a general discussion of the relative chronology of the inscriptions, see Keightley (1978: 91–133). 13 Heji 001248 正 and its translation is based on that of Puett (2002: 50), with modification. 14 This plastron was dated to the Yinxu phase I.

256

The spirit world 15 As Eno rightly points out, it is misleading to construe the Shang pantheon in parallel to the Egyptian or Greek ones that “ ‘include a relatively fixed dramatis super-personae, predictably deployed in myth and art as well as worshipped in cult,” precisely for the reason that the inscriptions themselves do not contain or indicate such a “panoply” (Eno 2009: 54, n.32). 16 Reconstructing the structure and the working mechanism of the Shang “ancestral cult” has been the center of scholarly attention. Moving away from the earlier shamanistic interpretations that primarily drew upon later textual sources (Chang 1983), Puett (2002) and Keightley (2004), among others, rightly emphasize the need of relying on the contemporaneous oracle-bone inscriptions. However, due to the lack of contemporaneous discoursing sources, the existing reconstructions are largely influenced by anthropological and social theories. 17 Puett 2002: 45, translations modified. 18 In the Shang oracle-bone inscriptions, the deceased female royal members tend to have their day of receiving cult designated on a xin or gui day, while the other gan-stems from jia to geng are usually reserved for the males. 19 Fu Hao tomb (M5) was found in 1976, and it was the only Shang royal tomb that remained intact until scientifically excavated. The many bronzes made by her or her husband on her behalf bear her royal identification, Fu Hao, when she was alive, and some bronzes made by her sons for her death and sacrifices bear her ancestral identifier Mu Xin (Zhongguo shehui kexueyuan kaogu yanjiusuo 1980). 20 Di Xin was portrayed by his conquerors, the Zhou, as a licentious individual and unworthy tyrant. One of the crimes, of which he was accused, was his failure to carry out regular sacrifices to his ancestors (Shiji 3.135). 21 There were records in which the sacrifice day and cycle was used to mark time in non-divinatory or sacrificial situations (Keightley 2004: 23–24). 22 Di received no cult at all; Nature Powers received some sacrifices; Former Lords occasionally received cult on certain gan days (Keightley 2004: 8). 23 Puett and Keightley are also aware of the fact that only those in the “ancestors proper” or the “true kin” were integrated into the “Five-Ritual Cycle” by receiving a gan-name and regular cult, as Puett writes, “the Shang ancestral cult represented an attempt (my emphasis) to forge nature spirits and the ghosts of deceased humans into a single, unified system,” and Keightley writes, “Shang liturgists would have (my emphasis) ancestralized a Power by incorporating it into the ritual system.” See Puett 2002: 53 and Keightley 2004: 8. 24 Keightley summarizes seventeen major areas of concerns that were divined in Period I: (1) sacrifices; (2) military campaigns; (3) hunting expeditions; (4) excursions; (5) the ten-day week (xun 旬); (6) the night or the day; (7) the weather; (8) agriculture; (9) sickness; (10) childbirth; (11) distress or trouble; (12) dreams; (13) settlement building; (14) orders; (15) tribute payments; (16) divine assistance or approval; (17) requests addressed to ancestral or nature powers (Keightley 1978: 33–35). 25 The last section, “Dongguan,” was already lost in the Western Han. In its place, a text called Kaogongji was compiled and added to the extant Zhouli. For a recent work and translation of Kaogongji, see Jun Wenren 2012. 26 In the fifth year of the Yuanshi Reign of Emperor Ping of Western Han (r.9 bce–6 ce), regent Wang Mang (ca. 45 bce–23 ce) proposed to the young emperor to establish an imperial sacrifice system and the pantheon according to the Zhouli. This system is commonly known as the “Yuanshi Rites,” which were implemented by Wang Mang when he became the emperor of his short-lived Xin dynasty (9–23 C.E.), but also was the basis for the imperial sacrifice system and pantheon in the restored Han dynasty, i.e., the Eastern Han (23–220 ce). For the evolution of the Qin and Han imperial sacrifices, see Tian 2015. 27 For a detailed discussion of the formulaic form and different components of those from the Baoshan Tomb 2, the only completely preserved corpus of this genre of the records, see Guo Jue 2008: 154–169. 28 For a general introduction of the genre of “daybooks,” see Harper and Kalinowski 2017. 29 Liu Lexian 1994: 368–370. 30 Yan Changgui estimates the number of spirits seen in excavated Chu sources to be around one hundred, but due to the fragmentary condition of the majority of the divination and sacrificial records as well as the phenomenon that some spirits had more than one name, especially those of ancestral nature, the actual number may be even larger (Yan 2009: 77) 31 See Geling slip B4–26 and Baoshan slips 240–241. More examples can also be found in Yang 2012: 161–162. 32 Yangzhou bowuguan and Hanjiang xian tushuguan, 1981.

257

Jue Guo 33 The Western Han historian and a survivor of a castration punishment, Sima Qian 司馬遷 (ca. 135 –87 bce), was said to have deeply lamented his unfortunate punishment and the severe consequence of such a fate, “I suffered this distress because of words. I am not only laughed at by the fellow villagers and friends, but also humiliated and dishonored my forebears. With what kind of self-respect can I go to visit my parents’ grave again?” 僕以口語遇遭此禍，重為鄉黨戮笑，汙辱先人，亦何面目復上 父母之丘墓乎? (Hanshu, 62.2725–2736). Compared to Sima Qian’s self-awareness and self-regret of violating the social norm, the Eastern Han critic Wang Chong 王充 (ca. 27–100 ce) adamantly attacked taboos and practices that prevent convicts, especially those who have suffered corporeal punishment, from participating in mourning and funeral services to their parents as irrational and ignorant. In the case of Li Gu 李固 (93–147 ce), an official who fell into disgrace during the reign of Emperor Huan 桓 (r. 146–167 ce), he was recorded to have told his family to place him in a simple coffin and bury him in an infertile field and to have forbidden them to return his burial to the family graveyard and “insult and contaminate the grave site of his forebears” (Hou-Han shu, 63.2087). 34 For those found near the First Emperor’s mausoleum, see Shihuangling Qinyongkeng kaogu fajuedui, 1982; For those found near Yangling, see Qin Zhongxin 1972; for the Eastern Han Luoyang cemetery of the convicts, see Zhongguo shehui kexueyuan kaogu yanjiusuo 2007. 35 Gui in the Shang oracle-bone inscriptions was mostly used to refer to a group of non-Shang people from guifang 鬼方. Shen, sometimes in the graph form of 申, appeared in the Western Zhou bronze inscriptions and was used as a generic term for spirits, especially the ancestral ones.The only occurrence of gui and shen together is from the inscription on the Bo Dong gui-tureen 伯冬簋, dated to the Middle Western Zhou (ca. tenth–ninth centuries bce): “pacifying shen and gui spirits” 妥(綏)神褱(鬼) (JC 04115), in which 鬼 appeared in a variant graph 褱. It was not until the fifth century bce that guishen was unambiguously together as a compound to refer to spirits in general. 36 Olga M. Gorodetskaya (Guo Jingyun 郭靜雲) points out that in the Shang oracle-bone inscriptions of the early and middle periods, it was xia shang 下上, and only at the very end of the Shang did shang xia appear (Guo J.Y. 2007). 37 For a detailed and accessible introduction to the text as well as a complete translation of the Zuozhuan in English, see Stephen Durrant, Wai-yee Li and David Schaberg 2016. 38 Translations are my own unless otherwise noted. 39 For an accessible translation of Mozi, see Watson (2003). Translations of the Mozi passages are mine unless otherwise noted. 40 Also see Knoblock (1988). The translations of Xunzi passages are mine unless otherwise noted. 41 Major, Queen, Meyer, and Roth, 2010: 523–524. 42 All the texts in Lunheng are from Huang Hui’s Lunheng jiaoshi (1990), and all translations are mine unless otherwise noted.

Works cited Allan, S. (1991) The Shape of the Turtle: Myth, Art, and Cosmos in Early China. Albany: SUNY. Bagley, W. R. (2004) “Anyang Writing and the Origin of the Chinese writing System,” in S. D. Houston (ed.) The First Writing: Script Invention as History and Process. Cambridge: Cambridge University Press, 190-249. Bing, Shangbai 邴尚白. (2009) Geling Chujian yanjiu 葛陵楚簡研究, Taibei: Taida chubanshe. Bing, Shangbai. (1999) “Chuguo bushi jidao jian yanjiu” 楚國卜筮祭禱簡研究. M.A.Thesis,Taiwan Ji’nan Guoji daxue. Boltz, G.W. (2011) “Literacy and the Emergence of Writing in China”, in Li Feng and D.P. Branner (ed.) Writing and Literacy in Early China: Studies from the Columbia Early China Seminar, Seattle and London: University of Washington Press. Boyer, P. (1994) The Naturalness of Religious Ideas: A Cognitive Theory of Religion, Berkeley, Los Angeles and London: University of California Press. Chang, K. C. (1980) Shang Civilization, New Haven and London:Yale University Press. Chang, K. C. (1983) Art, Myth, and Ritual: The Path to Political Authority in Ancient China, Cambridge: Harvard University Press. Chang, K. C. (1989) “An Essay on cong”, Orientations 20.6: 37–43. Chen, Wei 陳偉 (1999) “Hubei Jingmen Baoshan bushi jian suojian shenqi xitong yu xiangji zhidu” 湖北 荊門包山卜筮簡所見神祇系統與饗祭制度, Kaogu 考古4: 51–9.

258

The spirit world Chen, Wei (2007) “Churen daosi jilu zhong de rengui xitong ji xiangguan wenti” 楚人禱祀記錄中的人 鬼系統及相關問題, Gudai wenzi yu gudaishi 古代文字與古代史, 363–389. Csikszentmihalyi, A. M. (2006) Readings in Han Chinese Thought, Indianapolis and Cambridge: Hackett. Durrant, S., Li,Wai-Yee, and Schaberg, D. (trans). (2016) Zuo Tradition: Commentary on the Spring and Autumn Annals, Seattle and London: University of Washington Press. Eno, R. (1990) “Was There a High God Ti [Di] in Shang Religion,” Early China 15: 1–26. Eno, R. (2009) “Shang State Religion and the Pantheon of the Oracle Texts,” in J. Lagerwey and M. Kalinowski (ed.) Early Chinese Religion, Part One: Shang through Han (1250 bc-220 ad), Leiden and Boston: Brill. Gao, Guangren 高廣仁 and Shao, Wangping 邵望平. (1986) “Zhongguo shiqian shidai de guiling yu quansheng中國史前時代的龜靈與犬牲,” in Zhongguo kaoguxue yanjiu bianweihui (ed.) Zhongguo kaoguxue yanjiu: Xia Nai xiansheng kaogu wushinian jinian lunwenji 中國考古學研究 – 夏鼐先生考古 五十年紀念論文集, Beijing: Wenwu chubanshe. Gumbrecht, U.H. (1999) “Epiphany of Form: On the Beauty of Team Sports,” New Literary History, 30.2: 351–372. Gumbrecht, U.H. (2004) Production of Presence:What Meaning Cannot Convey, Stanford: Stanford University Press. Guo, Dashun 郭大順 (ed.) (2004) Niuheliang yizhi 牛河梁遺址. Beijing: Xueyuan chubanshe. Guo, Jingyun 郭靜雲 (Gorodetskaya, M. Olga). (2007) “Jiaguwen ‘shangxiaruo’qidao zhanci yu tiandi xiangjiao guannian”甲骨文 “上下若” 祈禱占辭與天地相交觀念, Zhouyi yanjiu 周易研究 1: 7–13. Guo, Jue (2008) “Reconstructing Fourth-Century b.c.e. Chu Religious Practices: Divination, Sacrifice, and Healing in the Newly Excavated Baoshan Manuscripts,” unpublished Ph.D dissertation, University of Wisconsin-Madison. Guo, Jue (2011) “Concepts of Death and the Afterlife Reflected in Newly Discovered Tomb Objects and Texts from Han China,” in A. L. Olberding and P. J. Ivanhoe (eds.) Mortality in Traditional Chinese Thought, Albany: SUNY. Guo, Jue 郭珏 (2013) “Qin Han chutu wenxian zhong de ‘zhisi’ yu ‘shisi’: yige jiyu ‘xingcheng kuangjia’ de shi fenxi ji fangfalun shang de sikao” 秦漢出土文獻中的“知死”與“事死” – 一個基於“形成框架”的 試分析及方法論上的思考, Jianbo 簡帛8: 49–67. Guo, Jue. (2018, forthcoming) “The Making of the Dead: Negotiating and Appropriating State Authority in the Case of Funerary “Relocation Document” of the Second to First Centuries B.C.E. China,” in Bamboo and Silk,Vol.2. Leiden and Boston: Brill. Harper, D. and Kalinowski, M. (2017) Books of Fate and Population Culture in Early China: The Daybook Manuscripts of the Warring States, Qin, and Han. Leiden and Boston: Brill. Henansheng wenwu kaogu yanjiusuo 河南省文物考古研究所. (2008) “Henan Xingyangshi Guandimiao yizhi Shangdai wanqi yicun fajue jianbao” 河南滎陽市關帝廟遺址商代晚期遺存發掘簡報, Kaogu 考古7: 32–64. Huang, Hui 黃暉. (1990) Lunheng jiaoshi 論衡校釋, Beijing: Zhonghua shuju. Jun, Wenren. (2012) Ancient Chinese Encyclopedia of Technology: Translation and Annotation of Kaogong ji, The Artificers’ Record, London and New York: Routledge. Keightley, N.D. (1978) Sources of Shang History:The Oracle-Bone Inscriptions of Bronze Age China, Berkeley, Los Angeles and London: University of California Press. Keightley, N.D. (1984) “Late Shang Divination: The Magico-Religious Legacy,” in H. Rosemont, Jr. (ed.) Explorations in Early Chinese Cosmology, Journal of the American Academy of Religion Studies 50.2: 11–34. Keightley, N.D. (1998) “Shamanism, Death, and the Ancestors: Religious Mediation in Neolithic and Shang China (ca. 5000–1000 B.C),” Asiatische Studien 52.3: 763–828. Keightley, N.D. (1999) “The Shang: China’s First Historical Dynasty,” in M. Loewe and E. L. Shaughnessy (eds.) The Cambridge History of Ancient China: From the Origins of Civilization to 221 B.C, Cambridge: Cambridge University Press. Keightley, N.D. (2000) Ancestral Landscape:Time, Space, and Community in Late Shang China, Ca. 1200–1045 B.C., China Research Monograph 53, Berkeley: Institute of East Asian Studies. Keightley, N.D. (2004) “The Making of the Ancestors: Late Shang Religion and Its Legacy,” in J. Lagerway (ed.) Religion and Chinese Society, volume 1: Ancient and Medieval China. Hong Kong: The Chinese University Press. Kern, M. (2009) “Bronze inscriptions, the Shangshu, and the Shijing: The Evolution of the Ancestral Sacrifice during the Western Zhou,” in J. Lagerwey and M. Kalinowski (ed.) Early Chinese Religion, Part One: Shang Through Han (1250 bc to 220 ad), Leiden and Boston: Brill.

259

Jue Guo Knoblock, J. (1988) Xunzi: A Translation and Study of the Complete Works, volume III: Books 17–32. Stanford: Stanford University Press. Li, Feng. (2009) Bureaucracy and the State in Early China: Governing the Western Zhou, Cambridge: Cambridge University Press. Li, Ling 李零. (2000) Zhongguo fangshu kao (xiudingben), Beijing: Dongfang chubanshe. Liu, Bin 劉斌. (2007) Shenwu de shijie: Liangzhu wenhua zonglun 神巫的世界：良渚文化綜論. Hangzhou: Zhejiang sheying chubanshe. Liu, Lexian 劉樂賢. (1994) Shuihudi Qinjian rishu yanjiu 睡虎地秦簡《日書》研究. Taibei: Wenjin chubanshe. Liu, Li. (2007) “Early Figurations in China,” in C. Renfrew and I. Morley (ed.) Image and Imagination: A Global Prehistory of Figurative Representation, Cambridge: McDonald Institute for Archaeological Research. Liu, Li and Chen, Xingcan. (2012) The Archaeology of China: From the Late Paleolithic to the Early Bronze Age, Cambridge: Cambridge University Press. Major, S. J., Queen, A. S., Meyer, S. A., and Roth, D. H., trans. (2010) The Huainanzi: A Guide to the Theory and Practice of Government in Early Han China, New York: Columbia University Press. Peng, Hao 彭浩. (1991). “Baoshan erhaomu bushi he jidao zhujian de chubu yanjiu” 包山二號墓卜筮和 祭禱竹簡的初步研究, in Baoshan Chumu 包山楚墓. Beijing: Wenwu chubanshe. 555–563. Peterson, E. C., and Lu, Xueming. (2013) “Understanding Hongshan Period Social Dynamics,” in A. Underhill (ed.) A Companion to Chinese Archaeology, Oxford: Wiley-Blackwell, 55–80. Piao, Zaifu 樸载福. (2011) Xian Qin bufa yanjiu 先秦卜法研究. Shanghai: Shanghai guji chubanshe. Puett, J.M. (2002) To Become a God: Cosmology, Sacrifice, and Self-Divinization in Early China, Cambridge, MA and London: Harvard University Asia Center for the Harvard-Yenching Institute. Qin, Zhongxin 秦中行. (1972) “Han Yangling fujin qiantumu de faxian” 漢陽陵附近鉗徒墓的發現, Wenwu 文物 7: 51–53. Shihuangling Qinyongkeng kaogu fajuedui 始皇陵秦俑坑考古發掘隊. (1982) “Qin shihuang ling xice Zhaobeihu cun Qin xingtu mu” 秦始皇陵西側趙背戶村秦刑徒墓, Wenwu 3: 1–11. Smith, A. (2010) “The Chinese Sexagenary Cycle and the Ritual Origins of the Calendar,” in J. M. Steele (ed.) Calendars and Years II: Astronomy and Time in the Ancient and Medieval World, Oxford: Oxbow Books. Tang, Jigen 唐際根 (2004) “The Social Organization of Late Shang China – A Mortuary Perspective,” unpublished Ph.D. Dissertation, the University of London. Tian, Tian 田天. (2015) Qin Han guojia jisi shigao 秦漢國家祭祀史稿. Beijing: Sanlian shudian. Underhill, A. (ed.) (2013) A Companion to Chinese Archaeology, Oxford: Wiley-Blackwell. Watson, B. (trans) (2003) Mozi: Basic Writings. New York: Columbia University Press. Wengrow, D. (2014) The Origins of Monsters: Image and Cognition in the First Age of Mechanical Reproduction, Princeton: Princeton University Press. Yan, Changgui 晏昌貴. (2009) Wugui yu yinsi: Chujian suojian fangshu zongjiao kao 巫鬼與淫祀：楚簡所 見方術宗教考, Wuhan: Wuhan daxue chubanshe. Yang, Hua 楊華 (2012). “Chujian zhong de ‘shangxia’yu ‘neiwai’: jianlun Churen jili zhong de shenling fenlei wenti” 楚簡中的“上下”與“內外”：兼論楚人祭禮中的神靈分類問題, in Guli xinyan 古禮新 研, Beijing: Shangwu chubanshe, 136–166. Yangzhou bowuguan 揚州博物館 and Hanjiang xian tushuguan 邗江縣圖書館. (1981) “Jiangsu Hanjiang Huchang wuhao Hanmu” 江蘇邗江胡場五號漢墓, Wenwu 文物 11: 12–23. Zhang, Deshui 張得水 and Li, Lina 李麗娜. (2005) “Zhongguo shiqian de gubu, guibu, he yubu” 中國 史前的骨卜、龜卜和玉卜in Zhongguo yuwenhua yuxue luncong 中國玉文化玉學論叢, vol. 3, Beijing: Zijincheng chubanshe. Zhang, Juzhong 張居中 and Cui, Qilong 崔啟龍. (2013) “The Jiahu Site in the Huai River Area,” in A. Underhill (ed.) A Companion to Chinese Archaeology, Oxford: Wiley-Blackwell. Zhongguo shehui kexueyuan kaogu yanjiusuo 中國社會科學院考古研究所 (1980). Yinxu Fu Hao mu 殷 墟婦好墓, Beijing: Wenwu chubanshe. Zhongguo shehui kexueyuan kaogu yanjiusuo 中國社會科學院考古研究所 (2007). Donghan xingtu mudi 東漢刑徒墓地, Beijing: Wenwu chubanshe.

260

12 RELIGIOUS THOUGHT

ORI TAVORRELIGIOUS THOUGHT

Ori Tavor

Defining early Chinese religion is a difficult task. As opposed to the institutional organized religious traditions of Daoism and Buddhism, which took form in the early stages of the first millennium ce, early Chinese religion is a particularly amorphous entity, as it lacks many of the features modern scholars view as fundamental – a canonical set of sacred scriptures, organized clergy, or a fixed pantheon. In fact, the very label of “early Chinese religion” does not refer to a specific empirical singularity. It is heuristic device, a term coined by later scholars to help make sense of the ideas, beliefs, and practices that circulated in China between the Shang and Han Dynasties. Despite all of that, recent years have witnessed a surge in book-length monographs devoted to ancestral worship (Brashier 2011), funerary practices and visions of the netherworld (Cook 2006; Wu 2010; Lai 2015), and self-cultivation and individual pursuits of immorality (Harper 1998; Poo 1998), as well as an imposing two-part edited volume featuring thematic essays by leading scholars (Lagerwey and Kalinowski 2009). Drawing on the expanding corpus of newly excavated texts, tombs, and artifacts, these studies offer us exponentially more nuanced account of the world of practice in early China, including the divinatory and sacrificial rituals described in the previous chapter of this volume (Chapter 11, by Jue Guo). This chapter will draw on a combination of transmitted and excavated sources to address a different issue, the emergence of new ways of thinking about ritual and justifying religious innovation. The gradual political waning of the centralized Zhou regime was accompanied by a complementary decline in the authority of the ritual system associated with it. New rituals, designed by pioneering religious innovators, began to emerge, challenging old ways of interacting with the divine realm. Alarmed by these challenges, elite thinkers who saw themselves as guardians of the old ritual system [li, 禮] of the Zhou were forced to create new ways of theorizing religion and explaining ritual efficacy. The following pages will provide an outline of this process by of depicting it as a growing rivalry between two modes of religiosity: a practical theology associated with a mechanical approach to ritual utilized by religious innovators to justify the invention of new practices and an alternative mode of religiosity advocated by the old guard, which reconceptualized ritual using a moral and cosmological framework and stressed the need for a complete sense of religious piety and devotion to a fixed body of ritual practices.

261

Ori Tavor

Religion in the Shang and Western Zhou In order to fully understand the emergence of new philosophical ideas about religion and ritual in the Warring States period we must first turn our attention to the religious landscape of the Shang and Western Zhou eras. Unfortunately, compared to the relatively rich collection of primary sources from the fifth century bce onwards, evidence on earlier religious practices is quite scant. Excavations in sites associated with the Bronze Age cultures of Erlitou [二里頭, first half of the second millennium bce] and Erligang [二里岡, mid-second millennium bce] have uncovered multiple bronze vessels and other ceremonial objects that point to the existence of standardized ritual practices mainly concerned with the proper disposal of the dead (Thorp 2006: 102–104). In the Anyang site [安陽], the location of the capital of the late Shang dynasty [商, ca. 1200 bce], archaeologists have unearthed large caches of turtle plastrons and bovine scapulae that were used in divination rituals. Known as oracle bones, the inscriptions carved on these artifacts disclose the existence of a complex ritual system accompanied by a specialized vocabulary and strict schedules centered on ritual sacrifices to a wide variety of supernatural beings, from nature gods and local deities to the ancestral spirits of the royal clan. While the precise mechanism that governed oracle-bone divination is unclear, most scholars agree that the Shang believed that the ancestors, as well as other deities and spirits, had the ability to exert their influence on the human realm. Divination was thus used to communicate with the spirits in order to ascertain the correct ritual procedure to solicit their blessing and avoid their wrath (Keightley 2000: 101; Itō 1996: 24). The practice of ancestral worship did not disappear after the fall of the Shang in 1046 bce and continued to occupy a central role in the religious system of the subsequent Zhou dynasty. Inscriptions on bronze vessels attest to the importance of the ancestral cult as the prominent religious institution of the elite during the Western Zhou period (Kern 2009: 143). As ritual objects that were most likely placed in lineage temples to be used during sacrifices, the texts inscribed on the bronze vessels are commonly interpreted as proclamations made by the living to their deceased ancestors, communicating their achievements and asking for their approval and support (Falkenhausen 1993: 146–157; Rawson 1999: 387). While the inscriptions themselves are quite terse, other Western Zhou sources contain brief descriptions of ancestral rituals and help us shed some light on the mechanics of the ritual performed on their behalf. The Book of Odes [Shijing, 詩經] contains several odes and hymns that offer a description and analysis of sacrifices to supernatural deities. In the “Birth of the People” [shengmin, 生民, Mao #247], for instance, we find a narrative that provides us with an explanation for the origin of the sacrificial rituals of the Zhou. In this mythological account, the invention of sacrifice is assigned to Lord Millet [Houji, 后稷], the progenitor of the Zhou people, who was born out of a miraculous encounter between his human mother, Jiang Yuan [姜嫄], and his divine father, Lord-on-High [Shangdi, 上帝]. In addition to teaching the people how to properly cultivate the land, Lord Millet also shows them how to perform the proper annual sacrifices that would ensure a successful crop. Following a detailed description of these sacrifices, the ode concludes with the following statement: 卬盛于豆 We heap the wooden trenchers full; 于豆于登 wooden trenchers, earthware platters. 其香始升 As the scent begins to rise 上帝居歆 Lord-on-High is pleased. 胡臭亶時 “What smell is this, so strong and good?” 后稷肇祀 Lord Millet initiated the sacrifices, 262

Religious thought

庶無罪悔 and without error or fault, 以迄于今 we continue to perform them today (Cheng 1985: 526; translation adapted from Waley 1996: 247) A similar description of the mechanism of sacrifice can be found in “Thorny Caltrop” [Chuci, 楚茨, Mao #209], which features an account of a ceremonial exchange between humans and spirits involving the offering of wine and food to satisfy the appetite of the latter and ensure they bestow their blessings on their descendants: 苾芬孝祀 Fragrant is the pious sacrifice 神嗜飲食 The spirits enjoy the wine and food 卜爾百福 The oracle predicts for you a hundred blessings 如幾如式 According to the proper quantities, according to the proper rules 既齊既稷 If you have brought sacrificial grain, you have brought millet 既匡既敕 You have brought baskets, you have arranged them 永錫爾極 We will forever give you the utmost blessings 時萬時憶 Ten-thousandfold, myriadfold (Cheng 1985: 427; Quoted in Falkenhausen 1993: 149) Both odes describe sacrifice as an act of communication between human and supernatural agents, whether it be a high deity like Lord-on-High or the ancestral spirits of their own lineage. In addition, they reveal an underlying assumption: the correct performance of ritual can ensure the desired outcome.Thus, despite the limited nature of Shang and Zhou sources, most modern scholars support the claim that their system of religious practice followed a mechanical du-et-des understanding of the relationship between the human and the divine realms. According to this mode of thought, once the correct ritual procedure is performed by humans, the spirits will have no choice but to respond favorably and bestow their blessings (Keightley 1978; Falkenhausen 1993; Goldin 2015; Pines 2002; Puett 2002). Given the mechanical nature of this interaction, the key for asserting control over the spirit world thus lies in identifying the deity responsible for the situation and ascertaining the exact ritual procedure needed to solve the problem they create, as attested in a large number of Shang oracle-bone inscriptions: 丁巳卜, 尹貞, 王賓父丁彡, 亡尤。 Dingsi day cracking, Yin divining: When the king hosts [the ancestral spirit of] Fu Ding and performs the rong sacrifice,1 will there be no disapproval. (Guo 1965: 441, quoted in Itō 1996: 24) In order to control the results of what may be a potentially volatile exchange with the supernatural, an absolute adherence to the proper ritual forms was required. By addressing the deceased by their proper name and entering them into the sacrificial schedule the worshipers were able to take an unpredictable and potentially dangerous ghost and make it into a proper ancestor, thus mollifying his or her will. For these reasons, one of the main outcomes of this pragmatic theology based on a du-et-des relationship between the human and divine worlds was the growing systematization of the ritual system (Keightley 2004; Puett 2002: 41; Sterckx 2007: 30). Archaeological evidence suggests that such a process of standardization took place in the tenth and ninth centuries bce. An analysis of excavated bronze vessels from that period suggests an increasing stress on homogeneity 263

Ori Tavor

in shape and design, accompanied by a growing uniformity in the form, content, and execution of the inscriptions. Dubbed as the Western Zhou “ritual revolution,” such a process of standardization could not have been possible without a centralized system of religious and political control. Driven by a desire to consolidate their authority, the Zhou regime began to regulate the production of religious artifacts (Rawson 1999: 419–438). This was accompanied by changes to the religious procedures themselves, namely a shift from collective rituals practiced by the entire community to more standardized liturgical rituals led by ritual professionals (Shaughnessy 1997: 165–195).

The decline of the Zhou and the rise of religious innovation The rise of ritual professionalization was a direct result of the increasing complexity of the Zhou ritual system. As the descriptions in the Odes suggest, sacrifices to spirits and ancestors were highly intricate choreographed religious spectacles that employed musicians, dancers, libationers, invocators, personators, and other ritual experts. Performed in the Zhou capital, as well as other centers of political power, these rituals played a central role in establishing the Zhou ruler’s religious authority and the regime’s claim over the land. During the Spring and Autumn period [770–481 bce], however, while the nominal sovereignty of the Zhou kings was still generally accepted, the de facto control over the territories of the Zhou state moved to the hands of local rulers. As these local rulers were striving for independence, they began rejecting the idea that royal ancestral spirits and royal performances of state sacrifices were essential for the prosperity of their own lands. Textual sources from the late Spring and Autumn and the early Warring States [453–211 bce] periods confirm the gradual decline in the authoritative status of the Zhou ritual system and the rise of vigorous effort at religious innovation. The Analects [Lunyu, 論語] reflects a growing preoccupation with the idea of religious innovation, a departure from the old normative ritual system of the Zhou. In passage 3.17 we find the following exchange: 子貢欲去告朔之餼羊。子曰：「賜也，爾愛其羊，我愛其禮。」 Zigong wished to do away with the offering of a sheep in the New Moon Sacrifice. The Master said: “You cherish the sheep; I cherish the ritual.” (Yang 2007: 29). The fact that Zigong, Confucius’ disciple, feels comfortable in suggesting a change to the Zhou New Moon Sacrifice is quite telling, as it reveals its declining status. Even more important, though, is Confucius’ categorical denial of any such attempts and his adamant devotion to the Zhou ritual system of li. Confucius, after all, made his living as a ritual expert. Unfortunately, it seems that he was living in a time when the demand for his particular sets of knowledge and skill were no longer in high demand. The following two passages, which discuss the notion of ingratiation, or toadyism [諂], reflect Confucius’ frustration with the current state of affairs: 子曰：「事君盡禮，人以為諂也。」 The Master said: “Serving one’s ruler in complete observance of ritual practice is seen by others as ingratiation.” (3.18) 子曰：「非其鬼而祭之，諂也. . .」 The Master said: “To offer sacrifice to an [ancestral] spirit that does not belong to one’s own lineage is [an act of] ingratiation.” (2.24) (Yang 2007: 22, 30) 264

Religious thought

While the problematic nature of the Analects, its authorship, and its dating make it hard to reconstruct a comprehensive religious world view, the passages quoted here indicate that religious innovation was quite prevalent at that time. The rapidly fading memory of the golden age of Zhou rule paved the way for attempts to modify the old ritual system associated with it. The growing influence of local rulers at the expense of the royal house of the Zhou led some ritual experts to design new sacrifices in order to ingratiate themselves with the new powers that be, who did not necessarily favor the old system of li.The Zuo Commentary of the Spring and Autumn Annals [Chunqiu Zuozhuan, 春秋左传], another text from the late Spring and Autumn or early Warring States period, is strewn with accounts of such conflicts between religious innovators and supporters of the Zhou ritual system. The next narrative is a good example of this tension: during the performance of ancestral rites, the presiding master of ritual [zongbo, 宗伯] in the state of Lu, Xia Fuji, decided to rearrange the tablets in the temple, placing the tablet of the most recent ruler, Lord Xi, above that of his half-brother and predecessor Lord Min, an action that was in direct violation [ni 逆] to the sacrificial system. Asked to explain his actions, Xia Fuji offered the following explanation, followed by a rebuttal by the noble person [junzi, 君子], who is the exponent of the Zhou ritual system: 吾見新鬼大，故鬼小，先大後小，順也，躋聖賢，明也，明順，禮也。君子 以為失禮，禮無不順，祀，國之大事也，而逆之，可謂禮乎? “I saw that the new ghost is larger and the old ghost is smaller.To put the larger first and the smaller last is to follow the right order.To elevate sages and worthies is wise.To be wise and follow the right order is in accordance with ritual propriety.” The noble man considered this a deviation from ritual propriety: “In the performance of ritual there is nothing that does not follow the right order. Sacrifices are among the great affairs of the domain. Can it be called ritual propriety to violate the right sacrificial order?” (Yang 1990: 524; Durrant, Li and Schaberg 2016: 473–475) This account is an illustrative example of an ongoing dispute between a new guard of religious innovators, who had no problem suggesting alterations to the ritual system to suit their own agenda, and the old guard, who saw themselves as the preservers of Zhou culture.2 A similar point is conveyed in another anecdote from the Zuozhuan: 鄭大旱，使屠擊，祝款，竪柎，有事於桑山，斬其木不雨，子產曰，有事於 山，蓺山林也，而斬其木，其罪大矣，奪之官邑。 There was a major drought in [the state of] Zheng. [The king] sent Tu Ji, invocator Kuan, and an attendant named Fu to perform a sacrifice on Mulberry Mountain. They cut down the trees [for the sacrifice], but it did not rain. Zichan said: “[the goal of] performing a sacrifice on the mountain is to nourish its forests. These [men] have cut down the trees and thus their crime is immense.” He proceeded to take away their official positions and fiefdoms. (Yang 1990: 1382; Durrant, Li and Schaberg 2016: 1539) Zichan [子產, also known as Gongsun Qiao 公孫僑, d. 522 bce] is mentioned throughout the text as a critic of popular religious ideas, especially the practice of interpreting astral phenomena as portents from divine forces (Goldin 1999: 39–45). Less is known about the identity of Tu Ji, Kuan, and Fu, except that they appear to be ritual specialists who held office in the state of Zheng and that they outranked Zichan in matters of religious affairs. When his state suffered a 265

Ori Tavor

drought, the king’s instinctive reaction was to send the three to perform a sacrifice to on top of the sacred Mulberry Mountain. The procedure they performed seemed to deviate from the established traditional sacrifice in that it involved the removal of trees, presumably to be served as offerings to the spirits deemed to be responsible for the drought. When this performance failed to achieve the desired results, the three were castigated by Zichan. His criticism, however, was directed toward their decision to assuage the spirits of the mountain by departing from the established ritual procedure for this type of case. Much like the Confucius of the Analects, Zichan deemed religious innovation as a threat to the ritual system of the Zhou and the ideology it encapsulated. Defending the Zhou ritual system thus required the construction of a new theoretical framework to replace the old du-et-des model, which was now used by religious innovators to justify their actions. This new model offered a revised explanation of the relationship between the human and divine realms, emphasizing the moral aspects of this interaction. It also involved the construction of a new discourse on ritual, reconceptualizing li as a set of ethical, sociopolitical, and religious guidelines that govern the behavior of the individual and the state.3 According to this theory, piety to the li was not only the mark of a refined cultivated human being but also the only viable way to interact with the divine realm.

Morality and religious thought in the Warring States period The growing emphasis on morality as a key component of religious behavior can be traced to the mid-to-late Western Zhou period. Shang oracle-bone inscriptions and early Western Zhou bronze inscriptions and hymns are utterly devoid of any reference to morality in interacting with divine forces, instead emphasizing the mechanical du-et-des nature of this relationship. The key for obtaining blessing and avoiding punishment lies in knowing the correct ritual procedure. Later sources, however, reveal the emergence of an alternative religious model centered around Heaven [tian, 天], an anthropomorphized deity that bestows its mandate [ming, 命] on the Zhou ruler, known as the Son of Heaven [tianzi, 天子], to govern on its behalf in a moral and virtuous [de, 德] fashion. The notion of an all-seeing Heaven that assigns rewards and punishments based not on the correct offering of sacrifices but on the merit of one’s moral behavior became a central component of political and religious thought from that point onward (Pines 2002: 58; Poo 1998: 30). One of the best examples for this new mode of religiosity can be found in the Mozi [墨子]. In the “Will of Heaven” [Tianzhi, 天志] chapter, we find the following admonition to a potential ruler: 然則天亦何欲何惡？天欲義而惡不義。然則率天下之百姓以從事於義，則我 乃為天之所欲也。我為天之所欲，天亦為我所欲。 然則我何欲何惡？我欲福 祿而惡禍祟。 若我不為天之所欲，而為天之所不欲，然則我率天下之百姓， 以從事於禍祟中也。 Now, what is it that Heaven desires and what does it loathe? Heaven desires what is right and loathes what is not right. Thus, if I lead the people of the world to act in accordance with what is right, then I will be doing what Heaven desires. And if I do what Heaven desires, then Heaven will do what I desire. Now, what is it that I desire and what do I loathe? I desire good fortune and emoluments, and loathe calamities and disasters. If I do not do what Heaven desires but rather what it loathes, then I will be steering all the people under Heaven to act in ways that lead them into disaster and calamity. (Sun 2001: 193) 266

Religious thought

The religious model of the Mozi can be described as a blend of the old du-et-des mode of religiosity and the notion of a moral universe governed by an anthropomorphized Heaven. On one hand, it features a relatively straightforward relationship between humans and Heaven – if we follow the basic principles of moral behavior, Heaven will be pleased and repay us with good fortune and material benefit, whereas immoral behavior will result in misfortune and adversity. On the other hand, this mechanical system of give and take is not based on the offering of sacrifices. The Mozi contains little mention of religious ritual. In fact, one of the most well-known features of the text is its condemnation of li, especially the prolonged funerary rites and the elaborate musical performances that accompanied the rituals of the Zhou elite, which the author perceives as a frivolous waste of resources and manpower that can be otherwise used to promote the benefit of the state. In that sense, the ideas articulated in the Mozi stand in opposition to the attitude expressed by such figures as Confucius and Zichan, who sought to protect and maintain the integrity of the traditional ritual system of the Zhou at all costs. This view, in addition to the text’s adamant efforts to prove the existence of ghosts and spirits, has led some scholars to identify Mozi as an anti-elite reformer and argue that the religious model expressed in the text belongs to the realm of archaic popular religion (Graham 1989: 47). When read in the context of the religious innovation debate, however, it becomes clear that the relationship between the human and the divine articulated in the “Explaining Ghosts” [Minggui, 明鬼] chapter has more in common with the new moral theology than the old du-et-des model (Sterckx 2013). Much like Heaven, the ghosts and spirits of the Mozi are omniscient anthropomorphized deities that have a keen interest in the human world: 雖有深谿博林，幽澗毋人之所，施行不可以不董，見有鬼神視之 . . . 嘗若鬼神 之能賞賢如罰暴也，蓋本施之國家，施之萬民，實所以治國家利萬民之道也。 Even in the deepest valleys or vast forests, in those hidden places where no one lives, you must always act properly. For the ghosts and spirits will see what you do. . . . Once the notion that ghosts and spirits can reward the worthy and punish the wicked be firmly established and executed among the various states and the common people, it could surely be used to bring order to the state and great benefit to the people. (Sun 2001: 234, 243) As this passage suggests, ghosts and spirits play a crucial role in the philosophical system of the Mozi. Much like Heaven, they function as an awe-inspiring deterrent for immoral behavior, an external device designed to ensure sociopolitical order. Given their significance, it is hardly surprising that the author spends most of the “Explaining Ghosts” chapter attempting to convince his readers of their existence.These sustained efforts, however, also indicate a waning belief in the power of supernatural beings, their perspicuity, and their ability to influence the human world. The recent discovery of excavated manuscripts that were not preserved in the transmitted canon reinforces the importance of this topic in Warring States religious and philosophical discourse. The Shanghai Museum text, The Perspicuity of Ghosts and Spirits [Guishen zhi Ming, 鬼神之明], for example, revolves entirely around this issue.4 Similarly to the Mozi, the text adamantly argues for the existence of ghosts and spirits and treats them as ultimate arbiters of reward and punishment. In dealing with their omniscient nature, however, it promotes a more skeptical claim in arguing that there are areas in which ghosts and spirits are perspicuous and areas in which they are not perspicuous [夫鬼神有所明, 又有所不明] (Ma 2005: 310; Brindley 2009: 216). The Great Drought of Lu [Lubang Dahan, 魯邦大旱], another excavated text from the Shanghai Museum corpus, offers more evidence for the centrality of the ghosts and spirits debate in the Warring States period while also shedding more light on the tension between religious 267

Ori Tavor

innovators who continued to operate under the du-et-des mode of religiosity and supporters of the new moral theology. When a great drought occurred in the state of Lu, Duke Ai [魯哀公, r. 494–468 bce] summoned Confucius and pleaded for his advice. Confucius, in return, explained that the drought was caused by the duke’s failure to practice moral government.When asked for a concrete solution to the problem, Confucius provided the following statement: 庶民知說之事鬼也,不知刑與德。汝毋愛圭壁幣帛於山川,正刑與德。 The common people only know of the shui rainmaking sacrifice5 [directed towards] the spirits but know nothing of [ruling through] law and moral government.Thus, you must be generous in offering jades and silks to the [Spirits of the] Mountains and Rivers and also implement laws and moral government. (Ma 2002: 205–206) Confucius’ recommendation to pursue both courses resonates with his famous assertion in the Analects regarding the need to venerate ghosts and spirits but to keep them at a distance (Yang 2007: 61–62). The text, however, does not end with that. Upon his return, Confucius reports the case to his disciple Zigong [子貢, 520–446 bce] and asks for his opinion. Zigong’s response is quite surprising: 若夫政刑與德,以事上天,此是哉！若夫毋薆圭璧币帛于山川, 毋乃不可。夫 山,石以為膚,木以為民,如天不雨,石將焦,木將死,其欲雨,或甚于我,何必恃乎名 乎？夫川, 水以為膚, 魚以為民, 如天不雨, 水將沽, 魚將死, 其欲雨, 或甚于我, 何 必恃乎名乎？ Ruling through law and moral government, thereby serving Heaven above, this is correct! Lavishly offering jades and silks for the [Spirits of the] Mountains and Rivers, this I cannot endorse. As for mountains, stones are their skin and trees are their people. If the sky does not send down rain, the stones will roast and the trees will die. Their desire for rain is certainly deeper than ours – how can they rely solely on our words [of evocation]? As for rivers, water is their skin and fish are their people. If the sky does not send down rain, the water will dry up and the fish will die. Their desire for water is certainly greater than ours – how can they rely solely on our words [of evocation]? (Ma 2002: 207–209) This case raises a few important points. First, it further reinforces the tension between the ritual experts, who sought to perform the rainmaking sacrifices in order to appease the Mountain and River Spirits, and their opponents, who believed that serving Heaven can only be achieved through moral government. Secondly, it demonstrates that while the viability of this old model was certainly questioned by educated elites, its popularity among the common people remained intact. From a practical point of view, offering sacrifices to the Mountains and Rivers based on the du-et-des model of interacting with the divine was still seen as the most immediate and commonsensical solution to the state of drought. The following Zuozhuan passage conveys a similar sentiment: 秋，七月，有神降于莘，惠王問諸內史過曰，是何故也。對曰：國之將興， 明神降之，監其德也；將亡，神又降之，觀其惡也 . . . 王曰：若之何？ 對曰： 以其物享焉，其至之日，亦其物也，王從之，內史過往，聞虢請命，反曰， 虢必亡矣，虐而聽於神，神居莘，六月，虢公使祝應，宗區，史嚚，享焉，

268

Religious thought

神賜之土田，史嚚曰，虢其亡乎，吾聞之國將興，聽於民，將亡，聽於神， 神聰明正直而壹者也，依人而行，虢多涼德，其何土之能得。 In autumn, in the seventh month, there was a spirit that descended at Shen. King Hui asked the court scribe Guo, “What is the reason for this?” He responded: “When a domain is about to prosper, radiant spirits descend there and inspect its virtue. When it is about to fall, spirits also descend there to observe its offenses . . . .” The king then asked: “What should I do about it?” He replied: “Use the appropriate objects in making offering to the spirit.These ought to be the objects corresponding to the day of its arrival.” The king followed this. The court scribe Guo traveled to Shen to present the offerings. There he heard that the duke of Guo had been requesting favors [from the spirit]. Upon his return, he said: “The state of Guo is sure to perish. Its ruler is cruel and heeds spirits.” The spirit dwelled in Shen for six months. The duke of Guo then instructed Invocator Ying, Ancestral Attendant Qu, and Scribe Yin to make offerings to it. The Spirit promised to give him the state of Guo. Scribe Yin said, “Surely the state of Guo will perish! I have heard that when a state is about to prosper, its ruler heeds the people; but when it is about to perish, he heeds the spirits. Spirits are keen of ear and eye, upright, straightforward in responding to human behavior.The state of Guo in many cases has shown little enough virtue – how can it possibly expand its territory?” (Yang 1990: 251–253; Durrant, Li, and Schaberg 2016: 223, with slight modifications) As in the previous examples, this story points to an ideological conflict between two opposing sides. On one side, we have the duke of Guo, the representative of the old du-et-des mode of thought, who believes that if he provides the proper offerings to the descending spirit it will repay him by expanding his territory. On the opposite side, court scribe Guo and scribe Yin, much like Zigong in the Great Drought of Lu, epitomize a new mode of religious thought based on a belief in a moral universe. Spirits, they argue, cannot be swayed by lavish offering. Their actions are a result of the ruler’s behavior – ethical conduct and proper government will be repaid by blessings, while a lack of virtue will result in disaster. Moreover, by arguing that the king should offer the spirit the proper offerings according to its day of descent, it can be said that the author is trying to establish a strong link between moral behavior and piety to a standardized system of ritual. The Drought of the Great King of Jian [Jian Dawang Bohan, 柬大王泊旱], another Shanghai Museum excavated text, makes an analogous argument. When a harsh drought fell upon his kingdom, the ruler of Jian, a small kingdom inside the southern state of Chu [楚], ordered one of his diviners to figure out which deity is responsible for the drought so that they may offer a sacrifice to it in the proper place and stop the drought. The king insisted on participating in the divination process while standing in the blazing sun, and this caused him to fall ill.6 Taking his illness as another indicator for the dissatisfaction of the deities, the king becomes increasingly distressed and attempts to persuade his diviners to look for an alternative site for the sacrifice. His idea of performing sacrifices to the Mountain and River Spirits that resided outside the kingdom of Jian, however, attracts much criticism in the royal court. In order to solve this dispute, the rival sides seek the advice of the Chief Minister. After hearing both side of the argument, he responds: 君入而語僕之言於君王： 君王之騷從 今日以瘥 . . . 君王元君，不以其身變釐 尹之常故；釐尹爲楚邦之鬼神主，不敢以君王之身變亂鬼神之常故。夫上帝 鬼神高明甚，將必知之。君王之病將從今日以已。

269

Ori Tavor

“Please go back and convey these words to the king. Tell him that from today he will start to recover from his illness. . . .The king is a good ruler. He did not change the fixed rules of divination for his own sake. You, diviner, control the [sacrifices] to the ghosts and spirits in the state of Chu. You also did not dare to change the fixed rules only for the sake of your ruler thereby creating disorder among the ghosts and spirits. Shang Di, the ghosts, and the spirits are highly discerning. They will surely recognize this. Thus, from this day, the king will start to recover from his illness.” (Ji,Yuan and Chen 2007: 75) As in the Great Drought of Lu, this passage suggests that the most natural reaction to a state of drought at the time was to perform a rainmaking sacrifice directed at natural deities. In addition, it also gives us of more information about the structure of these rituals, the identity of the ritual specialists who performed them, and the religious framework that was used to explain them. According to this passage, the sacrificial procedure begins with a divination designed to ascertain the identity of the responsible deity and locate the appropriate location for the sacrifice. Ritual is perceived as a repertoire of techniques placed at the disposal of the ritual specialist in order to create a sacred space in which interaction with the divine is possible. The ultimate success of the sacrifice depends on the ritualist’s ability to use his repertoire to manipulate the deities into reciprocating. This type of trial-and-error du-et-des model associated with Shang and Western Zhou religiosity was thus still quite pervasive during the Warring States (Sterckx 2007: 32–37). As the Drought of the Great King of Jian suggests, however, the practical model advocated by natural experts and ritual specialists was criticized by the new aspiring elite thinkers, who offered their own model of ritual efficacy focused on piety to a fixed ethical system of practice. Like Zichan and Zigong before him, the Chief Minister stresses the overall devotion to the system as a whole over the performance of a specific ritual. The state of Chu, he argues, has fixed rules about sacrifice. Changing them for the sake of the king’s selfish wish for divine blessings will not only harm him politically but will also create chaos in the divine realm. Devotion to this holy fixed system of rituals, on the other hand, will not escape the eyes of the Lord-on-High and other divine powers. These deities will repay such religious piety by healing the king and, by extension, his state.

Xunzi’s Moral theology The examples from the Zuozhuan and the Shanghai Museum manuscripts reveal the emergence of a moral mode of religiosity that links the efficacy of sacrifice to a sustained adherence to a strict ethical, political, and ritual system. Nonetheless, they also demonstrate that this mode of thinking about the relationship between the human and the divine was by no means homogenous at that time. On one side of the spectrum, we have Zigong, who vehemently denies the possibility of an interaction between humans and ghosts and spirits. Supernatural beings, he argues, do not need our offerings. The whole notion of swaying them through prayer and sacrifice is therefore completely useless. On the other side, we find figures such as Court Scribe Guo, Scribe Yin, and the Grand Minister, who still believe in a universe in which ghosts and spirits can be convinced to bestow their blessings and avoid causing harm. The only viable way to sway them, however, is by remaining devoted to the li ritual system and the ethico-religious principles it represents. The text associated with the late Warring States thinker Xunzi [荀子, ca. 310–218 bce], which contains the most comprehensive and influential criticism against the du-et-des mode of religiosity, leans toward the skeptical side of the spectrum, similar to the views articulated by Zigong in the Great Drought of Lu. 270

Religious thought

The “Discourse on Heaven” [Tianlun, 天論] chapter has often been described as the fullest and most systematic version of the philosophical skepticism and critical attitude toward the popular belief in ghosts and spirits or in a sentient and compassionate Heaven that is actively engaged in human affairs. One of Xunzi’s main goals in this chapter is to clearly distinguish between the natural and the human world. Each realm, he claims, has its own rules and mechanisms: Heaven’s Way [Tiandao, 天道] and the Human Way [Rendao, 人道] are discrete realms, and there is no way to communicate between the two spheres, let alone manipulate this communication to our advantage. According to this world view, calamities do not arise due to malicious supernatural powers or a disgruntled Heaven: 星隊木鳴，國人皆恐。曰：是何也？曰：無何也！是天地之變，陰陽之化， 物之罕至者也。怪之，可也；而畏之，非也。夫日月之有食，風雨之不時， 怪星之黨見，是無世而不常有之。 When stars fall and trees cry, all the people in the state are afraid. They ask: why is this happening? I answer: for no particular reason.Those things occasionally occur due to the transformation of Heaven and Earth and the transformation of yin and yang. We may be surprised by them but we should not fear them. Solar and lunar eclipses, unseasonable rains and winds, and dubious sightings of strange stars – these things have been quite common throughout the ages. (Wang 1988: 313) This argument can be likened to other anti-portent arguments found in the Zuozhuan. Omens, argues that author, are inevitable. We cannot avoid them by appealing to supernatural beings. The only way to do so is first to understand the pattern and movement of Heaven and then to use this acquired knowledge to our advantage.This can be gained through observing the course of Heaven, Earth, and the Four Seasons empirically, recording its configuration, sequence, and movements. These type of assertions have led many modern scholars to hail Xunzi as a staunch critic of religion and the forebear of rationalist thought in China (Feng 2007: 232). Reading this passage against the backdrop of the Shanghai Museum texts, however, offers us an opportunity to contextualize his theory of ritual within the larger Warring States religious discourse as a mature articulation of an emerging new theology designed to reassert the authority of the fixed body of religious practices known as li. Set against the du-et-des model of their rivals, this elite mode of religiosity seeks to create an indissoluble link between ritual as a system of ethical and sociopolitical guidelines and its divine cosmic origin. Xunzi’s critique of the popular theory of ritual efficacy is presented in the following passage from the “Discourse on Heaven” chapter: 雩而雨，何也？曰：無佗也，猶不雩而雨也。日月食而救之，天旱而雩，卜 筮然後決大事，非以為得求也，以文之也。故君子以為文，而百姓以為神。 以為文則吉，以為神則凶也。 If a rainmaking sacrifice is held, and then it rains, what of it? I say, there is no reason. It would still rain even if we do not hold the sacrifice. When the sun and moon are eclipsed, a sun-saving rite is performed; when Heaven sends a drought, a rainmaking sacrifice is performed; before deciding upon serious matters, tortoise shell and milfoil divinations are performed. These [rituals] are not held in order to get a result, but in order to establish a pattern.Thus, the gentleman takes [ritual] as a matter of establishing a pattern while the common people take it as a matter of [sacrificing to the] spirits.To take [ritual] as creating a pattern is auspicious. To take it as [sacrifice to the] spirits is ill-fated. (Wang 1988: 316) 271

Ori Tavor

It is important to note that Xunzi does not object to the performance of these rituals but to the religious mentality that underlies them. Instead of the old mechanical mode of ritual interaction in which rituals are performed for the sake of the spirits, Xunzi’s model strives to portray ritual participation as an activity that establishes a pattern. Writing for a new elite audience of educated scholar-aspirants, Xunzi wishes to establish a new mode of religiosity based on an absolute sense of devotion to the system of li and the ethico-religious values it represents. The efficacy of ritual, he argues, is not based on its ability to mollify supernatural beings but on its ability to promote two interrelated religious goals: individual ethical, physical, and spiritual transformation and the maintenance of harmony between the human and the divine. In addition to the shifts in the realm of religious thought and the emergence of new theoretical models of ritual and sacrifice, the late Warring States period also witnessed the rise of new religious practices, such as meditation, sexual regimens, and calisthenics, designed to achieve a variety of personal goals, from the prolongation of life to the attainment of divine-like powers (Harper 1998; Roth 1999; Despeux 2004). The dissemination of these practices and the selfdivinization claims they embodied was aided by rising literacy rates and the development of an active manuscript culture (Tavor 2016). This process, however, represented a clear threat to ritual specialists, whose status and authority were based on the notion that their services were the only viable way to mollify the spirits and solicit their blessings (Puett 2002: 155–116). As a ritual specialist himself, Xunzi responded to this challenge by portraying ritual as an efficacious technique of self-cultivation that can induce a physical and cognitive transformation and achieve the same bounties promised by proponents of self-divinization practices but at the same time also promote social, as well as cosmic, harmony (Tavor 2013). This idea is clearly articulated in the following passage from the “Discourse on Music” [Yuelun, 樂論] chapter: 君子以鐘鼓道志，以琴瑟樂心；動以干戚，飾以羽旄，從以磬管。故其清明 象天，其廣大象地，其俯仰周旋有似於四時。故樂行而志清，禮脩而行成， 耳目聰明，血氣和平，移風易俗，天下皆寧，美善相樂。 The gentleman utilizes the bells and drums in order to create correspondence between his consciousness and the Way and the zithers and lutes to gladden his mind. He moves wielding the shield and battle-axe. Adorned with oxtails and plumes, he follows the rhythm of the chime stones and pitch pipes. In his purity and brilliance he models himself after Heaven, in his greatness and vastness he models himself after Earth, and in his posturing and movements he models himself after the Four Seasons. Thus, when music is performed, his will becomes pure, and when ritual is cultivated his conduct is perfected. His hearing becomes acute and his vision clear, the flowing of his blood and qi harmonious and uniform, his practices altered and his customs changed. All under Heaven is made tranquil and everybody joins together in the joy of beauty and goodness. (Wang 1988: 381–382) This description is important for several reasons. First, it clearly shows that for Xunzi, ritual is an embodied religious activity that can produce a bio-spiritual transformation on a communal level. Secondly, it also articulates a new notion of the interaction between the human and the divine that presents ritual not as a tool for mollifying the spirits but as a vehicle for humans to participate in the workings of the cosmos. Ritual performances, argues Xunzi, allow us to use the components of the cosmos as models for our mental attitude and bodily movements. By following the ritual script and playing the part of a deity, natural force, or a cultural hero, humans are thus able to enter into a relationship with the divine. Moreover, as opposed to the 272

Religious thought

mechanical du-et-des mode, in this relationship humans are equal parts in a triad with Heaven and Earth [參於天地矣] (Wang 1988: 443) instead of mere blessing-seekers. All of this is due to the cosmic origin of ritual: 水行者表深，表不明則陷。治民者表道，表不明則亂。禮者，表也；非禮， 昏世也；昏世，大亂也。故道無不明，外內異表，隱顯有常，民陷乃去。 Those who cross waterways mark them where it is deep. If the markers are not clear, the people will drown. Those who govern people mark the Way. If the markers are not clear, disorder will arise. Ritual is the marker. Opposing ritual means throwing the world into darkness. Casting darkness upon the world will bring great disorder. Thus, when the Way has nothing which is not clear, when different markers are set to distinguish between the inner and the outer and when darkness and light are constant, then the things which cause people to drown would be eradicated. (Wang 1988: 318–319) Rituals, argues Xunzi, are not arbitrary.The system of li functions as a set of markers left by sages, a script that can be used as a guiding light for the rest of humanity to follow. Moreover, since rituals are based on the fixed patterns of the Way, one must adhere to the ritual system of li without attempting to alter it. Xunzi’s attitude concerning the Way can thus be best understood as one of religious reverence or devotion. By creating an indissoluble link between the structure of the universe and the system of li, Xunzi offers an explicit theological justification for a new mode of elite religiosity focused on a commitment to a body of ethico-religious behavioral guidelines. According to this moral theology, rituals are not performed in order to seek an anticipated result from a supernatural deity. Instead, the performance of rituals of the Way is a pattern-establishing activity that denotes the religious devotion and the moral stature of the practitioner.

Early imperial religion and the cosmic nature of ritual While Xunzi did not live to see the unification of China under the rule of the Qin dynasty [221–206 bce], his philosophical theories played a significant role in shaping the newly emerging imperial ideology, especially during the early decades of the Western Han [206 bce – 9 ce] (Goldin 2007). The new sociopolitical circumstances brought forth a new wave of religious innovation. In their desire to augment their power and announce their supremacy, autocratic emperors such as the First Emperor of the Qin [Qin Shi Huangdi, 秦始皇帝] and Emperor Wu of the Han [漢武帝] embarked on ritualistic tours of inspection, offering cult to a variety of deities and announcing their ascent to both the human and divine realms. In addition, they relied on the services of religious experts to design new rituals, such as the Feng [封] and Shan [禪] sacrifices on Mount Tai [泰山], to the Yellow Emperor [Huangdi, 黃帝] at Yong [雍], and to the deity the Grand One [Taiyi, 太一 [at Ganquan [甘泉] (Kern 2000; Lewis 1999; Bujard 2009). These rituals, which were centered on the figure of the emperor as a semi-divine figure and his own personal quest for immortality (Puett 2002: 258), attracted much criticism from educated literati who saw themselves as guardians of the old ritual system of the Zhou. Alarmed by these grandiose attempts at religious innovation, they sought to offer their own model of imperial religion that incorporated many elements of Warring States elite religiosity, including the emphasis on piety to the overall system of li and the moral theology that stood at its base. One of the best examples of their efforts is the Records of Rites [Liji, 禮記]. Edited during the Western Han, based on some earlier material, some of it dating back to the Warring States period, the compilation of the Liji was part of an organized project led by a 273

Ori Tavor

group of literati designed to produce an ideal authoritative and standardized ritual framework for the new imperial religion (Riegel 1993; Nylan 2001). In addition to prescriptive descriptions of various rituals and sacrifices, the Liji also contains multiple passages that can be read as attempts to theorize ritual and explain the origin, nature, and function of the system of li. Following the line previously articulated in the writings of Xunzi, it depicts rituals as human artifacts created by the sages based on cosmic patterns (Puett 2009: 697). The “Meaning of Sacrifice” [Jiyi, 祭義] chapter, for instance, begins with the following assertion: 祭不欲數，數則煩，煩則不敬。祭不欲疏，疏則怠，怠則忘。是故君子合諸 天道：春禘秋嘗。霜露既降，君子履之，必有凄愴之心，非其寒之謂也。 春，雨露既濡，君子履之，必有怵惕之心，如將見之。樂以fl來，哀以送往， 故禘有樂而嘗無樂。 Sacrifices should not be frequently repeated. Such frequency is indicative of importunateness; and importunateness is inconsistent with reverence. Nor should they be at distant intervals. Such infrequency is indicative of indifference; and indifference leads to forgetting them altogether. Therefore, the gentleman, in accordance with the ways of Heaven, offers the di sacrifice in the spring and chang sacrifice in autumn. When he treads on the dew which has descended as hoar-frost he cannot help a feeling of sadness, which arises in his mind, and cannot be ascribed to the cold. In spring, when he treads on the ground, wet with the rains and dews that have fallen heavily, he cannot avoid being moved by a feeling as if he were seeing his departed friends.We meet the approach of our friends with joy, and see them off with sadness, and hence the di spring sacrifice in spring includes musical performances, but not at the chang sacrifice in autumn. (Sun 1989: 1207–1208; translation adapted from Legge 1885: 210) This passage illustrates one of the main tenets of the Liji in particular and the new imperial religion in general – sacrifice is only effective when its performed at the right time according to the seasonal ritual schedule. In the Monthly Ordinances [Yueling, 月令] texts, a new genre that emerged in the third and second centuries bce, we find detailed monthly schedules for the performance of state rituals. Materials pertaining to this ritualistic timetable can be found in the “Seasonal Patterns” [Shize, 時則] chapter of the Huainanzi and the “Monthly Ordinances” chapter of the Liji, but the most detailed version is depicted in the first part of the Spring and Autumn Annals of Master Lü, also known as the “Twelve Chronicles” [shierji, 十二記]. Divided into twelve sections by month, each chapter provides ritual instructions for the ruler, including detailed descriptions of the changes in the ruler’s clothes, regalia, diet, and policies, all in correspondence to the monthly cycles. Failure to follow the schedule can thus result in disaster, as detailed in the opening chapter of the Spring and Autumn Annals of Master Lü, 1.1/7: 孟春行夏令，則風雨不時，草木早槁，國乃有恐。行秋令，則民大疫，疾風 暴雨數至，藜莠蓬蒿並興。行冬令，則水潦為敗，霜雪大摯，首種不入。 If summer ordinances are carried out in the early spring, winds and rain will not follow their proper timing, plants and trees will wither prematurely, and terror will sweep across the state. If autumn ordinances [are carried out in the early spring], the common people will suffer great plagues, strong winds and torrential rains will frequently occur, and pestilent weeds will flourish. If winter ordinances [are carried out in the early spring], floods and monsoons will bring calamity, frost and snow will wreak ferocious havoc, and the first-sown crops will fail to grow. (Chen 1984: 2) 274

Religious thought

Much like the Xunzi, this passage can be read as a representation of a new mode of religiosity that focuses on the human instead of the divine.Whereas in the du-et-des model natural disasters are seen as an outcome of a disgruntled divine being, according to the new theory, calamities are caused by the inability of humans to abide by the correct ritual schedule created by the sages based on cosmic patterns. Religious innovation, attempts to appeal to ghosts and spirits through sacrifices that are not a part of the system, is thus posited as the cause for misfortune, rather than the remedy. Another good example for this attitude can be found in the “Evolution of Ritual” [Liyun, 禮運] chapter of the Liji, an essay that contains one of the most comprehensive theories of the origin and function of ritual in early China (Ing 2012; Puett 2010). It begins with a dialogue between Confucius and one of his disciples, Yan Yan, after they attend the seasonal zha [ 蜡] sacrifice, which marks the end of the lunar year. Confucius uses this occasion to lament the waning status of the system of li and explain its extreme importance in ensuring sociopolitical, as well as cosmic, order. In order to prove this point, he offers his disciple a detailed account of the evolution of ritual, going back to ancient times: 夫禮之初, 始諸飲食. 其燔黍捭豚, 污尊而抔飲, 蕢桴而土鼓, 猶若可以致其敬於 鬼神. 及其死也, 升屋而號, 告曰: 皋! 某複! 然後飯腥而苴孰. As for the origin of ritual, it started with various acts of offering drink and food [to the spirits]. These included broiling millet, slicing and roasting pork, digging small holes in the ground and using it as goblets, and using the kui plant to fashion drumming sticks in order to pound on the earthen drums. [The ancients believed that through these offerings] they can extend their reverence to the ghosts and spirits. When somebody died, they climbed on the roof of their houses and shouted: “Oh! Come back!” and then put uncooked rice in the mouth of the deceased and presented him bundles of cooked food. (Sun 1989: 586–587) The passage identifies the offering of food and drink to the ghosts and spirits, which is conducted according to the du-et-des mode of religiosity, as an important phase in the development of ritual. It is important to note, however, that the author is careful to situate this scene in the remote past, before the development of complex societies. In ancient times, people lived in caves and straw huts, gathered fruits and nuts, ate the raw meat of animals, and dressed themselves in simple clothes made out of feathers and animal pelts. This situation, however, changed with the appearance of the sage-rulers, whose cultural innovation helped the common people improve their situation. They taught them how to use fire, produce tools and utensils, process hemp, and erect permanent dwellings. In addition, they also constructed a system of ritual based on cosmic patterns: 夫禮，必本於大一，分而為天地，轉而為陰陽，變而為四時，列而為鬼神。 The ritual system is surely based on the Grand One, which separated to become Heaven and Earth, rotated to become yin and yang, transformed to become the four seasons, and it arrayed to become the ghosts and spirits. (Sun 1989: 616; Puett 2010: 366) Xunzi’s basic claim that li were markers left by sages based on their knowledge of the patterns of the Way is expanded in this passage into a more comprehensive claim about the cosmic origin of ritual. By identifying the evolution of the Zhou system of li with the evolution of universe, the “Liyun” author is positing it as the only viable tool for achieving correspondence 275

Ori Tavor

between the human and the divine, thereby confirming the continued prosperity of human society: 聖王所以順，山者不使居川，不使渚者居中原，而弗敝也。用水火金木，飲 食必時。合男女，頒爵位，必當年德 . . . 故無水旱昆蟲之災，民無凶饑妖孽之 疾 . . . 故天降膏露，地出醴泉，山出器車 . . . 則是無故，先王能修禮以達義, 體 信以達順，故此順之實也。 The sage-kings used [the ritual system] to create concordance. They did not send the mountain dwellers to live next to the river; neither did they send the island dwellers to live in the central plains. They used water, fire, metal, wood and various foods and drinks according to the right timing. They approved marriages and assigned titles and official positions according to age and moral virtue. . . .Thus, there were no calamities of floods, droughts, and pestilence. The common people did not suffer from the afflictions of starvation and scourge. Heaven sent down its luscious dew, Earth sprouted its sweet springs, the mountains provide resources for making tools and chariots. . . . All of this was a result of the former-kings’ ability to promote rightness through cultivating ritual and to promote concordance through embodying truthfulness. These are indeed the fruits of concordance. (Sun 1989: 622)

Conclusion This chapter surveyed the development of religious thought from the Shang to the Han. The changing sociopolitical reality in the mid-to-late Western Zhou and the decline of the ritual system associated with it opened the door for religious innovation. The sources surveyed here suggest that the old ritual system of the Zhou was no longer seen as entirely sacred and authoritative but instead open to change.This challenged the religious authority of its guardians, which, in response, began to develop a theoretical framework designed to instill new meaning into religious practice and conserve the authority of the ritual system of li. In order to reassert its legitimacy, they had to come up with new ways to justify it and explain the source of its efficacy. Their solution was to present ritual as the earthly human counterpart of cosmic order, the only viable way to maintain harmony between the human and divine realms. In the early imperial period, this mode of religious thought culminated in the creation of a textual ritual canon. In that sense, works like the Liji can be seen as the clear outcome of Xunzi’s previous endeavors, as they are prescriptive texts that outline a comprehensive and standardized ritual system and include a built-in theology. The rise of moral religiosity, however, did not signal the disappearance of the du-et-des mentality.The sources analyzed in this chapter were written by highly educated elite scholar-officials and were intended for an elite audience. As “ideological explanations by the intellectual elites that are aimed at explicating particular ideological positions,” they do not represent the popular perception of the supernatural (Poo 2003: 295). While the architects of the new imperial religion of the Han dynasty utilized the moral theological framework to construct an official state cult in which the emperor’s performance of grand sacrifices is seen as instrumental to the maintenance of social, political, and cosmic harmony, the a-moral practical theology associated with the Warring States ritual experts did not die out. Recent studies of Han religion clearly demonstrate that personal religious practices aimed at obtaining individual practical benefit, which did not belong to the official system of li, continued to flourish on all levels of society, from the common people to the emperor. Much like the Shang and Western Zhou rituals, they were 276

Religious thought

not accompanied by an explicit theology, but the world view they depict seems to be devoid of moral concerns, and their conception of human-divine interaction is primarily practical in nature (Poo 1998; Harper 2004). The tension between these two modes of religiosity that emerged in early China became one of the key features of Chinese religious discourse and was instrumental in the subsequent formation of organized religious traditions such as Daoism and Buddhism. After the fall of the Han, for example, this tension manifested itself in the rhetoric used by the followers of a new religious movement, Celestial Masters Daoism [Tianshi Dao, 天師道], who attempted to undermine the popularity of their religious rivals – local cults. One of the main strategies used by the Celestial Masters in asserting the superior efficacy of their rituals was to claim that their system of practice was based on a moral theology sent down to earth by the deified Way. By establishing a link between the bureaucratic organization of the celestial realm and their earthly ethicoreligious codes, the Celestial Masters were able to identify their system as orthodox [zheng, 正] and the rituals of rival local cults as heterodox [xie, 邪] (Lai 1998: 11–13). Understanding the development of religious thought and its competing modes of religiosity in early China is thus instrumental, as it offers us new ways of understanding the religious traditions that are still practiced in the Chinese cultural sphere today.

Notes 1 For more on this sacrifice, which is sometimes rendered as 酉彡, see Liu 2004: 108–116. 2 For similar accounts, see Lord Wen 15.5 (5), Lord Cheng 2.4 (5), and Lord Xiang 2.3 (3) in Durrant, Li, and Schaberg 2016: 547, 725, 895. 3 For an alternative reading of this process, which depicts the creation of the system of li as a distillation of normative sociopolitical and ethical principles from the collapsing Zhou religious framework, see Pines 2000. 4 The Shanghai Museum corpus includes more than 1,200 inscribed bamboo strips purchased in 1994 by the Shanghai Museum on the Hong Kong antiquities market. Written in archaic script that was identified as originating in the state of Chu [楚], these manuscripts, which include alternative versions of texts that were preserved in the received tradition, as well as many previously unknown texts, were assigned the approximate date of 300 bce. See Ma 2001: 1–4. 5 Reading 敓 as 說. This interpretation is suggested by both Liu Lexian and Li Xueqin (Liu 2003: 60–61 and Li 2004: 98). 6 For more on self-exposure and rainmaking rituals in Ancient China, see Cohen 1978 and Schafer 1951.

Works cited Brashier, K.E. Ancestral Memory in Early China. Cambridge: Harvard University Press, 2011. Brindley, Erica. “’The Perspicuity of Ghosts and Spirits’ and the Problem of Intellectual Affiliations in Early China.” Journal of the American Oriental Society 129.2 (2009): 215–236. Bujard, Marianne. “State and Local Cults in Han Religion.” In: Early Chinese Religion, Part One: Shang through Han (1250 BC to 220 AD). Eds. John Lagerwey and Marc Kalinowski. Leiden: Brill, 2009, 777–811. Chen, Qiyou 陳奇猷. Lüshi Chunqiu Jiaoshi 呂氏春秋校釋. Shanghai 上海: Xuelin Chubanse 學林出版 社, 1984. Cheng, Junying 程俊英. Shijing Yizhu 詩經譯注. Shanghai 上海: Guji Chubanshe 古籍出版社, 1985. Cohen, Alvin. “Coercing the Rain Deities in Ancient China.” History of Religions 17.3/4 (1978): 244–265. Cook, Constance. Death in Ancient China:The Tale of One Man’s Journey. Leiden: Brill, 2006. Despeux, Catherine. “La Gymnastique Daoyin dans la Chine Ancienne.” Etudes chinoises 23 (2004): 45–86. Durrant, Stephen, Wai-yee Li, and David Schaberg. Zuo Tradition/Zuozhuan: Commentary on the ‘Spring and Autumn Annals’. Seattle: University of Washington Press, 2016. Falkenhausen, Lothar von. “Issues in Western Zhou Studies: A Review Article.” Early China 18 (1993): 139–226.

277

Ori Tavor Feng, Youlan 馮友蘭. A short History of Chinese Philosophy (bilingual edition) 中國哲學簡史 (英漢對照). Tianjin 天津: Tianjin Shehui Kexueyuan Chubanshe 天津社會科學院出版社, 2007. Goldin, Paul R. “The Consciousness of the Dead as a Philosophical Problem in Ancient China.” In: The Good Life and Conceptions of Life in Early China and Graeco-Roman Antiquity. Ed. R.A.H. King. Berlin: De Gruyter, 2015, 59–92. ———. Rituals of the Way:The Philosophy of Xunzi. Chicago: Open Court, 1999. ———. “Xunzi and Early Han Philosophy.” Harvard Journal of Asiatic Studies 67.1 (2007): 135–166. Graham, A.C. Disputers of the Tao. Chicago: Open Court, 1989. Guo, Moruo 郭沬若. Yinqi Cuibian 殷契粹編. Beijing 北京: Kexue Chubanshe 科學出版社, 1965. Harper, Donald. “Contracts with the Spirit World in Han Common Religion: The Xuning Prayer and Sacrifice Documents of A.D. 79.” Cahiers d’Extrême-Asie 14.1 (2004): 227–267. ———. Early Chinese Medical Literature:The Mawangdui Medical Manuscripts. London and New York: Kegan Paul International, 1998. Ing, Michael David Kaulana. The Dysfunction of Ritual in Early Confucianism. New York: Oxford University Press, 2012. Itō, Michiharu. “Part One: Religion and Society”. In: Studies in Early Chinese Civilization: Religion, Society, Language, and Palaeography. Vol. 1: Text; Vol. 2: Tables and Notes. Eds. Itō Michiharu and Ken'ichi Takashima. Osaka: Kansai Gaidai University Press, 1996, 1–178. Ji, Xusheng 季旭昇, Yuan Guohua 袁國華, and Chen Siting 陳思婷. Shanghai Bowuguan Cang Zhanguo Chuzhushu (si) Duben 上海博物館藏戰國楚竹書(四)讀本. Taibei 台北: Wanjuan Loutushu Youxian Gongsi 萬卷樓圖書有限公司, 2007. Keightley, David. The Ancestral Landscape: Time, Space, and Community in Late Shang China, ca. 1200–1045 B.C. Berkeley: Institute of East Asian Studies, University of California, Center for Chinese Studies, 2000. ———. “The Making of the Ancestors: Late Shang Religion and Its Legacy.” In: Religion and Chinese society. Ed. John Lagerwey. Hong Kong: Chinese University of Hong Kong Press, 2004, 3–63. ———. “The Religious Commitment: Shang Theology and the Genesis of Chinese Political Culture.” History of Religions 17.3/4 (1978): 211–225. Kern, Martin. “Bronze inscriptions, the Shangshu, and the Shijing: The Evolution of the Ancestral Sacrifice during the Western Zhou.” In: Early Chinese Religion, Part One: Shang through Han (1250 bc to 220 ad). Eds. John Lagerwey and Marc Kalinowski. Leiden: Brill, 2009, 143–200. ———. The Stele Inscriptions of Ch’in Shih-huang: Text and Ritual in Early Chinese Imperial Representation. New Haven: American Oriental Society, 2000. Lai, Chi-tim.“The Opposition of Celestial-Master Taoism to Popular Cults During the Six Dynasties.” Asia Major 11.1 (1998): 1–20. Lai, Guolong. Excavating the Afterlife: The Archaeology of Early Chinese Religion. Seattle: University of Washington Press, 2015. Legge, James. Sacred Books of China: The Texts of Confucianism, Part IV: The Li Ki, XI-XLVI. Oxford: The Clarendon Press, 1885. Lewis, Mark E. “The Feng and Shan Sacrifices of Emperor Wu of the Han.” In: State and Court Ritual in China, Ed. Joseph P. McDermott. New York: Cambridge University Press, 1999, 50–80. Li, Xueqin李學勤. “Shangbo Chujian Lubang Dahan Jieyi 上博楚簡《魯邦大旱》解義.” In: Shangbo Guancang Zhanguo Chuzhushu Yanjiu Xubian 上博館藏戰國楚竹書研究續編. Eds. Zhu Yuanqing 朱 淵清 and Liao Mingchun 廖名春. Shanghai上海: Shanghai Shudian Chubanshe 上海書店出版社, 2004, 97–101. Liu, Lexian 劉樂賢. “Shangbo Jian Lubang Dahan Jianlun 上博簡魯邦大旱簡論.” Wenwu 文物2003.5: 60–64. Liu, Yuan 劉源. Shang Zhou Jizu Li Yanjiu 商周祭祖禮研究. Beijing 北京: Shangwu Yinshuguan 商務印 書館, 2004. Ma, Chengyuan 馬承源. Shanghai Bowuguan Cang Zhanguo Chuzhushu,Vol.1 上海博物館藏戰國楚竹 書(一). Shanghai上海: Shanghai Guji Chubanshe 上海古籍出版社, 2001. ______. Shanghai Bowuguan Cang Zhanguo Chuzhushu,Vol.2 上海博物館藏戰國楚竹書(二). Shanghai上 海: Shanghai Guji Chubanshe 上海古籍出版社, 2002. ———. Shanghai Bowuguan Cang Zhanguo Chuzhushu,Vol. 5 上海博物館藏戰國楚竹書(五). Shanghai上 海: Shanghai Guji Chubanshe 上海古籍出版社, 2005. Nylan, Michael. The Five “Confucian” Classics. New Haven:Yale University Press, 2001.

278

Religious thought Pines, Yuri. “Disputers of the Li: Breakthroughs in the Concept of Ritual in Pre-Imperial China.” Asia Major 13.1 (2000): 1–41. ———. Foundations of Confucian Thought: Intellectual Life in the Chunqiu Period, 722–453 b.c.e. Honolulu: University of Hawai’i Press, 2002. Poo, Mu-chou. “Imperial Order and Local Variation: The Culture of Ghost in Early Imperial China.” Acta Orientalia 56.2/4 (2003): 295–308. ———. In Search of Personal Welfare: A View of Ancient Chinese Religion. Albany: State University of New York Press, 1998. Puett, Michael. To Become a God: Cosmology, Sacrifice, and Self-Divinization in Early China. Cambridge: Harvard University Press, 2002. ———. “Combining the Ghosts and Spirits, Centering the Realm: Mortuary Ritual and Political Organization in the Ritual Compendia of Early China.” In: Early Chinese Religion, Part One: Shang through Han (1250 BC to 220 AD). Eds. John Lagerwey and Marc Kalinowski. Leiden: Brill, 2009, 695–720. ———. “Ritualization as Domestication: Ritual Theory from Classical China.” In: Ritual Dynamics and the Science of Ritual,Vol. 1. Ed. Axel Michaels. Wiesbaden: Harrassowitz Verlag, 2010, 365–376. Rawson, Jessica. “Western Zhou Archaeology.” In: The Cambridge History of Ancient China: From the Origins of Civilization to 221 B.C. Ed. Michael Loewe and Edward L. Shaughnessy. Cambridge, New York: Cambridge University Press, 1999, 352–459. Riegel, Jeffery. “Li Chi.” In: Early Chinese Texts: A Bibliographical Guide. Ed. Michael Loewe. Berkeley: Society for the Study of Early China and the Institute of East Asia Studies, University of California, Berkeley, 1993, 293–297. Roth, Harold. Original Tao: Inward Training (Nei-yeh) and the Foundations of Taoist Mysticism. New York: Columbia University Press, 1999. Schafer, Edward. “Ritual Exposure in Ancient China.” Harvard Journal of Asiatic Studies 14.1/2 (1951): 130–184. Shaughnessy, Edward L. Before Confucius: Studies in the Creation of the Chinese Classics. Albany: State University of New York Press, 1997. Sterckx, Roel. “Mozi 31: Explaining Ghosts, Again.” In: The Mozi as an Evolving Text: Different Voices in Early Chinese Thought. Eds. Carine Defoort and Nicolas Standaert. Leiden: Brill, 2013, 95–141. ———. “Searching for Spirit: Shen and Sacrifice in Warring States and Han Philosophy and Ritual.” Extrême-Orient, Extrême-Occident 29 (2007): 23–54. Sun, Xidan 孫希旦. Liji Jijie 禮記集解. Beijing 北京: Zhonghua Shuju 中華書局, 1989. Sun,Yirang 孫詒讓. Mozi Xiangu 墨子閒詁. Beijing 北京: Zhonghua Shuju 中華書局, 2001. Tavor, Ori. “Authoring Virile Bodies: Self-Cultivation and Textual Production in Early China.” Studies in Chinese Religions 2.1 (2016): 45–65. ———. “Xunzi’s Theory of Ritual Revisited: Reading Ritual as Corporal Technology.” Dao: Journal of Comparative Philosophy 12.3 (2013): 313–330. Thorp, Robert L. China in the Early Bronze Age: Shang Civilization. Philadelphia: University of Pennsylvania Press, 2006. Waley, Arthur. The Book of Songs. New York: Grove Press, 1996. Wang, Xianqian 王先謙. Xunzi Jijie 荀子集解. Beijing 北京: Zhonghua Shuju 中華書局, 1988. Wu, Hung. The Art of Yellow Springs: Understanding Chinese Tombs. Honolulu: University of Hawai’i Press, 2010. Yang, Bojun 楊伯峻. Chunqiu Zuozhuan Zhu 春秋左传注. Beijing 北京: Zhonghua Shuju 中華書局, 1990. ———. Lunyu Yizhu 論語譯注. Beijing 北京: Zhonghua Shuju 中華書局, 2007.

279

13 POLITICAL THOUGHT

YURI PINESPOLITICAL THOUGHT

Yuri Pines*

The three centuries that preceded the establishment of the Chinese empire in 221 bce were an age of exceptional intellectual flourishing. No other period in the history of Chinese thought can rival these centuries in creativity, boldness, ideological diversity, and long-term impact.Values, perceptions, and ideals shaped amid intense intellectual debates before the imperial unification contributed decisively to the formation of the political, social, and ethical orientations that we identify today with traditional Chinese culture. More broadly, the ideas of rival thinkers formed an ideological framework within which the Chinese empire functioned from its inception until its very last decades. These ideas stand at the focus of the present chapter. The centuries under discussion are often dubbed the age of the “Hundred Schools of Thought.”The school designations were developed primarily by the Han (206/202 bce–220 ce) literati (Smith 2003; Csikszentmihalyi and Nylan 2003) as a classificatory device for the variety of pre-imperial texts. This classification, even if belated, may be heuristically convenient insofar as it groups the texts according to their distinct ideological emphases, distinct vocabulary, and distinct argumentative practices. For instance, followers of Confucius (551–479 bce) and Mozi 墨子 (ca. 460–390 bce) were prone to prioritize morality over pure political considerations, in distinction from those thinkers who are – quite confusingly (Goldin 2011a) – dubbed Legalists (fa jia 法家). Confucians (Ru 儒) and Legalists also differed markedly with regard to the nature of elite belonging (see later). This said, it is fairly misleading to imagine “schools” as coherent ideological camps, as was often done through the twentieth century and beyond.Their ubiquity notwithstanding, the school designations cannot serve as a proper analytical unit. Rather than addressing the intellectual dynamics of the pre-imperial age through the prism of competing schools, it is more useful to identify a broad perspective of a common discourse in which most contemporary thinkers and statesmen took active part. This perspective will allow us both to highlight common ideas and perceptions of rival thinkers and to outline with greater precision fields of disagreement and foci of acute debates.

Background: why politics? Among the major world intellectual traditions that took shape during what Karl Jaspers (1965) dubbed “the Axial Age” (Achsenzeit, eighth–third centuries bce), Chinese thought appears as the most politically oriented. One will have a hard time finding either a philosophical or historical 280

Political thought

text from the pre-imperial or early imperial period that does not discuss such issues as the nature of rulership, ruler-minister relations, an intellectual’s political involvement, ways of controlling the people, and the like. It would not be an exaggeration to say that questions of power, authority, and proper methods of maintaining sociopolitical relations – what can be broadly defined as political thought – dominate the texts that survived the vicissitudes of history. The preponderance of political discussions in pre-imperial texts is not surprising, given their historical context. The outburst of intellectual creativity from the age of Confucius 孔子 (551– 479 bce) and his disciples onward took place against the backdrop of a severe systemic crisis.The end of the Spring and Autumn period (Chunqiu, 770–453 bce) was marked by the progressive disintegration of political structures in the Zhou realm. First, the Zhou dynasty’s (ca. 1046–255 bce) kings, the proud “Sons of Heaven” (tianzi 天子), lost their power to their nominal subordinates, the regional lords (zhuhou 諸侯); then the latter were eclipsed by powerful ministerial lineages within their domains; and soon enough heads of these lineages were challenged by their rebellious kin or even by their stewards. In the wake of this devolution of the ruler’s power, the Zhou world became entangled in a web of debilitating struggles among rival polities, between powerful nobles and the lords, and among aristocratic lineages within each polity. By the fourth century bce, a degree of recentralization in individual polities was achieved, but interstate warfare further intensified, giving, in retrospect, the new era an ominous name: the age of the Warring States (Zhanguo 戰國, 453–221 bce). How to “stabilize All-under-Heaven” became the central concern addressed by competing thinkers. Crises and bloodshed aside, the Warring States period was also an age rife with opportunities for intellectually active individuals. It was an exceptionally dynamic period, marked by novel departures and profound changes in all walks of life. Politically, the loose aristocratic entities of the Spring and Autumn period were replaced by centralized and bureaucratized territorial states (Lewis 1999a: 597–616). Economically, the introduction of iron tools (Wagner 1993) revolutionized agriculture, allowing higher yields, prompting the development of wastelands, and bringing about demographic growth, as well as accelerating urbanization and a commercialization of the economy. Militarily, new technologies, such as the crossbow, as well as new forms of military organization brought about the replacement of aristocratic, chariotled armies by mass infantry armies staffed by peasant conscripts, resulting in a radical increase in warfare’s scale and complexity (Lewis 1990: 53–96, 1999a: 620–632; Yates 1999: 25–30). And socially, the hereditary aristocracy that dominated the Zhou world during much of the Bronze Age (ca. 1500–400 bce) was eclipsed by a much broader stratum of shi 士 (sometimes translated as “men-of-service”), the men who owed their positions primarily to their abilities rather than their pedigree (see later). These profound changes required new approaches to a variety of administrative, economic, military, social, and ethical issues: old truths had to be reconsidered or reinterpreted. For intellectuals eager to tackle a variety of new questions, this was the golden age.1 The magnitude of challenges and opportunities was by itself conducive to the flowering of political thought during the Warring States period. In addition, this flowering benefited from the relatively relaxed intellectual atmosphere. In the fragmented world of that age, no government could impose effective political orthodoxy; nor was there any institution on a par with religious establishments elsewhere able to impose – or even just to define – orthodoxy in the intellectual realm. This resulted in most remarkable intellectual pluralism. Even upon a cursory reading of the texts by different thinkers, the sheer variety of approaches impresses. Some ascertained the divinity of Heaven and deities, while others rejected it; some advocated the political involvement of the intellectuals, while others ridiculed it; blatant militarists debated with staunch pacifists; supporters of state activism rivaled advocates of laissez-faire policy.The evident 281

Yuri Pines

absence of politically “forbidden zones” during that age contrasts sharply with a much chillier atmosphere under the unified empire. Yet remarkable as it is, intellectual pluralism of the Warring States period should not obscure the common set of problems faced by competing thinkers and, more importantly, the similarities in their proposed solutions.The common goal of these thinkers was attaining peace, stability, and orderly rule for the entire subcelestial realm. There was also a broad agreement about the basic parameters of the due-to-be stabilized world.The joint commitment of the majority of thinkers to political unity of All-under-Heaven, to the monarchic principle of rule, to the concept of meritocracy, to the dictum of intellectual’s political involvement, and to the need to maintain a decent livelihood for the commoners formed a common intellectual framework within which the future imperial Chinese political culture evolved. But how should these goals be attained, and how exactly should the future unified realm be maintained? These questions generated bitter polemics which made the intellectual life of that age so fascinatingly rich.

Stability is in unity Perhaps the single most pronounced point of consensus among the thinkers of the Warring States period is their unanimous insistence on political unity as the only way out of the turmoil of perennial war. At the age of profound political fragmentation, the age when the competing Warring States attempted to strengthen their domestic cohesiveness and separate their subjects from those of the foreign states (Shelach and Pines 2006: 219–222), intellectually active members of the shi stratum acted contrarily to this trend. These intellectuals, who frequently crossed the borders in search of new appointments, developed a “universal” outlook: they concerned themselves not with a fate of an individual state but with the entire subcelestial realm. And, given the woeful failure of the efforts to attain a viable inter-state order (Pines 2000a: 282–297), it became clear to all: peace is possible only under the aegis of a single ruler who will impose his will on All-under-Heaven. Thinkers proposed different rationales for unification. For Confucius, for instance, it meant primarily restoring the functioning mode of the Western Zhou (ca. 1046–771 bce) polity, in which “ritual, music, and punitive expeditions” are administered by the Son of Heaven alone and not by regional lords (Lunyu 16.2). Mozi, in distinction, created a different model, which he placed in the remote past: “when the people just arose.” Back then there was a beast-like war of all against all, which ended only when “the worthiest and the most able [man] in Allunder-Heaven” was established as Son of Heaven, creating thereafter a perfectly centralized and uniformly ruled universal state (Mozi III.11:109–110 [“Shang tong shang”]). Mozi’s audience might have well understood that this narrative invokes the past to serve the present: the political myth aimed to demonstrate that unification is the only way out of current disorder and devastating mutual strife. Other thinkers, such as the author(s) of an exceptionally influential fourth-century bce text, the Laozi 老子, dispensed with the past altogether. Their justification for political unification was metaphysical: just as the universe is ruled by the uniform and allpenetrating force of the Way (Dao), so should the society be unified under a single Monarch whose position will match that of Heaven, Earth, and the Way (Laozi 25). Yet the most compelling rationale for unification was provided by one of Confucius’s most eminent followers, Mengzi 孟子 (aka Mencius, ca. 380–304 bce). When asked by a regional ruler “how to stabilize All-under-Heaven,” Mengzi plainly replied: “Stability is in unity” (Mengzi 1.6). Mengzi’s reply reflects the common belief of competing thinkers. The texts from the second half of the Warring States period seem no longer to be preoccupied with justifications for the future unification, since the need to unify the entire subcelestial realm became the 282

Political thought

unquestionable common desideratum. Henceforth, debates revolved primarily not about why the world should be unified but about how the unity should be achieved. Many thinkers hoped that this could be done through non-violent means. Mengzi, for instance, ridiculed those who wanted to subjugate All-under-Heaven militarily as daydreamers who “look for fish by climbing a tree” (Mengzi 1.7); elsewhere he stated that only he who has “no proclivity to kill, will be able to unify” the world (Mengzi 1.6). However, laudable as it was, Mengzi’s and like-minded thinkers’ vision of peaceful unification under a morally upright sovereign was impractical. Mengzi himself lamented that the True Monarch – the ultimate unifier – comes once in five hundred years, and his coming is long overdue (Mengzi 4.13). Other thinkers preferred not to wait for a savior but to hasten unification practically. The most notorious – and most successful of these – Shang Yang 商鞅 (d. 338 bce), plainly stated that the True Monarch is the one who commits himself to resolute war, in which he will subjugate his rivals and bring about the long-desired peace and tranquility (Shangjunshu 7.2). The difference in means between Mengzi and Shang Yang could not be greater, but the bottom line remained all the same: “stability is in unity.” Thinkers of the Warring States period disagreed not only about the proper ways to attain unity but also about the nature of future unification. Would it be restoration of a loose Western Zhou-type of polity under the ritual supremacy of the Son of Heaven, as implied by Confucius in his aforementioned saying? Or would it be a more tightly organized and centralized polity, as implied by Mozi? And what should be the limits of the due-to-be-unified All-under-Heaven? Should it include the Zhou oikouménē alone, namely the areas of shared elite culture (written language, mortuary rites, ritual gradations), or should it encompass the alien periphery as well? A conservative vision, represented for instance in the Warring States period document “The tribute of Yu” (“Yu gong” 禹貢), now a chapter in the canonical Classic of Documents, was that of limited unification. The text explains how the legendary demiurge Yu, having subdued the flood, arranged the world into Nine Provinces (jiu zhou 九州). The Nine Provinces (the precise location and names of which vary from one text to another) are fundamentally congruent with the territories of China proper, i.e., with the Zhou realm. This terrestrial organization implies that the entire known world is a complete and closed system, organized in a 3-by-3 grid, which cannot be meaningfully altered (Dorofeeva-Lichtman 2009). The immutability of this scheme becomes even clearer from a parallel “field-allocation” (fen ye 分野) astrological system, which divides the sky into nine partitions associated with each of the provinces below. As noticed by Paul R. Goldin, this association meant that “no tenth region [to the Nine Provinces] could ever have been added. There would simply have been no tenth part of the sky to identify with it” (Goldin 2015: 44). The Nine Provinces scheme (the origins of which may well precede the Warring States period) is purely Sino-centric, as it glosses over the areas associated with alien tribes. An alternative Sino-centric vision, which is also present in “The Tribute of Yu” as well in several other texts, is more attentive to the aliens’ presence. It divides the world into five (elsewhere nine) concentric zones: the internal ones are ruled directly by the Son of Heaven and his regional lords, while the external ones are inhabited by the barbarian tribesmen and banished Chinese criminals (Shangshu 3: 202–206). Here the alien periphery is incorporated in the realm under the control of the Son of Heaven, but this incorporation is primarily symbolic. Both the Nine Provinces and the Five Zones models represent therefore a particularistic and partly or fully exclusivist vision of the future unification.Yet this approach was challenged by other texts which emphasized the true universality of the ancient paragons’ deeds and implied that the coming unification as well should include both the Central States and the alien periphery (e.g., Mozi IV.15:160 [“Jian ai zhong”]). The inclusive vision is clearly pronounced, for instance in the Gongyang commentary 公羊傳 on the canonical Springs and Autumns Annals (Chunqiu 春秋) 283

Yuri Pines

(on which see Gentz 2015). Since “the True Monarch wants to unify All under Heaven,” the divisions between the “internal” (Chinese) and the “external” (alien lands) are just temporary (Gongyang zhuan, 417 [Cheng 15]). The future unity should be truly universal. The emphasis on absolute universality of unification, which eventually became the cornerstone of the imperial propaganda under the unifying Qin dynasty (221–207 bce; Kern 2000: 151–153), reflected a peculiar optimism of the Warring States period thinkers, who considered the alien tribesmen as only temporarily “barbarian”; in due time they can be educated and incorporated into a civilized community (Pines 2005b). It was only after the Qin-Han unification and the encounter with the steppe nomads that Chinese thinkers had to profoundly re-evaluate their views of the Other and recognize the limits of imperial expansion (Di Cosmo 2002; Goldin 2011b).Yet in the pre-imperial period the aliens remained, overall, insignificant in discussions of unification or otherwise (pace Bünger 1987: 321–322).What mattered for thinkers and statesmen of that age is how to subjugate rival Sinitic states, which were the major source of disorder and war under Heaven. The primary goal of unification was attaining peace in the Central Plains; other issues were secondary. It is worth mentioning that aside from explicit calls for unity, the philosophical discourse of the Warring States period facilitated future imperial unification in a variety of other ways. For instance, the political mythology of that age backdated the notion of unity to the remote past, implying thereby that political fragmentation is an aberration and not an acceptable state of affairs (Pines 2008a, 2010). Ritual compendia postulated the existence of a universal sociopolitical pyramid headed by the Son of Heaven as the singularly appropriate arrangement, de-legitimating thereby the current situation of competing loci of authority (Pines 2009: 28–30). The very language of political discourse, with its repeated postulates of the superiority of universality to particularity (Lewis 2006), was conducive to the goal of unification. Yet perhaps the most interesting aspect of pro-unification discourse is not in what was said but in what the thinkers did not say. That not a single individual or text is known ever to have endorsed a goal of a regional state’s independence is most remarkable. Even in the texts unearthed from the supposedly culturally distinctive state of Chu we find a clear commitment to the “universal” perspective which postulates the superiority of “All-under-Heaven” over its component parts (Pines, 2018). Thus, denied ideological legitimacy, separate polities became intrinsically unsustainable in the long term. It was this common quest for unity that the eventual unifier, the Qin dynasty, utilized to bolster its legitimacy (Pines 2012: 19–22). And while the Qin experiment of excessive centralization failed (Shelach 2014), the idea of political unity as a singularly acceptable way of maintaining political life remained the most recognizable feature of Chinese political culture well until the end of Imperial China and even beyond (Pines 2012: 11–43).

The Monarch’s power The principle of monarchic rule can be considered the second major pillar of the Warring States period political thought. It is closely related to the principle of political unity; like the latter, it emerged as the thinkers’ preferred solution to the aggravating sociopolitical crisis. The process of political disintegration of the preceding Spring -and Autumn period was intrinsically linked to the devolution of the ruler’s power: first from the Son of Heaven to regional lords and then from regional lords to heads of powerful ministerial lineages in each polity. Political reforms of the Warring States period were aimed primarily at stemming this disintegration by restoring the ruler’s power and creating what Mark E. Lewis aptly names the “ruler-centered state” (Lewis 1999a: 597). The ideology of monarchism evolved parallel to the practical strengthening of the ruler’s authority; it both reflected this strengthening and greatly contributed to it. 284

Political thought

Thinkers of the Warring States period provided multiple justifications for the ruler’s exaltedness. Ritual masters, starting from Confucius himself, emphasized the ritual superiority of the monarch, whose exclusive sumptuary privileges should sharply distinguish him from his underlings. Other thinkers emphasized the ruler’s importance as a moral exemplar to his subjects. In an idealized Mozi’s model, the ruler was supposed to be “the worthiest and the most able” person (Mozi III.11: 109–110 [“Shang tong shang”]). Mengzi plainly claimed: “When the ruler is benevolent – everybody is benevolent; when the ruler is righteous – everybody is righteous; when the ruler is correct – everybody is correct” (Mengzi 7.20 and 8.5). Laozi, as mentioned earlier, elevated the Monarch to the position of the counterpart of Heaven, Earth, and the Way; many later texts developed this equation further, providing metaphysical stipulations for the ruler’s supremacy (Pines 2009: 36–44). Yet all these approaches pale in their importance in comparison to the most practical consideration: the ruler is simply essential for the proper functioning of the political apparatus and of society in general.The society and the state will be torn apart by conflicting private interests (si 私) unless there is a single person who represents the common good (gong 公, meaning both “common” and “the lord”; cf. Goldin 2013: 3–4). The ruler’s unifying presence is the only means to ensure social and political health. The emphasis on the ruler’s sociopolitical importance permeates the texts of the Warring States period. Xunzi 荀子 (d. after 238 bce), arguably the single most sophisticated political thinker of that age, explains that only due to the ruler’s presence would the human beings be able to form a collective, which is essential for attaining supremacy over nature (Xunzi V.9: 165 [“Wang zhi”]). Shen Dao 慎到 (fl. late fourth century bce) explains that without “the single esteemed [person], there is no way to carry out the principles [of orderly government, li 理]” (Shenzi, 16 [“Wei de”]). The authors of the major pre-imperial compendium, Lüshi chunqiu 呂 氏春秋 (composed ca. 240 bce), summarize: The True Monarch upholds oneness and becomes the rectifier of the myriad things. The army needs the general: thereby it is unified. The state needs the ruler: thereby it is unified. All under Heaven needs the Son of Heaven: thereby it is unified.The Son of Heaven upholds oneness, thereby unifying it [the realm]. Oneness brings orderly rule; doubleness brings chaos. (Lüshi chunqiu 17.8: 1132) This brief statement, which embeds references to the Laozi (par. 39) and to Shen Dao, is a convenient summary of the ideology of monarchism of the Warring States period. The unity of the Way (viz. the “oneness”) should be logically matched by administrative unity of decision making. Any dispersal of authority means inevitable struggle and turmoil. Just as the army cannot act without a clearly defined chain of command with a supreme commander at its top, so, too, the state requires a unified command as the only way to survive in the violent competition with its neighbors. Moreover, since political unification is the only reasonable solution to ongoing warfare, it should logically culminate in the unification of power in the hands of a single person. Any alternative to this strict monarchism will have devastating effects on the entire realm. The almost unanimous consensus in favor of a strictly monarchic form of rule achieved during the most pluralistic and innovative period in the history of Chinese political thought explains the hegemonic power of the ideology of monarchism throughout the subsequent imperial millennia (Pines 2012: 44–75). Yet it should be immediately noted that thinkers of the Warring States period were by no means sycophants of the current rulers. To the contrary, they were bitterly critical of the rulers’ inadequacies and were ready to criticize contemporary sovereigns as benighted mediocrities. Many thinkers considered themselves intellectually and 285

Yuri Pines

morally superior to the rulers, as the teachers and not just subjects of the monarchs (see later). Yet none of the rulers’ critics – not even the authors of iconoclastic Zhuangzi 莊子, who derided all the rulers, past and present, as bloodthirsty tyrants (Zhuangzi, 778–779 [“Dao Zhi”]; 252–256 [“Qu qie”]; see also Yu Youqian 1982) – posed any alternative to the monarchic form of rule. Nor did any of them propose institutional limitations to monarchic power. An individual ruler could – and should – be criticized; he was encouraged to consult his meritorious aides and heed their opinion; yet his was the final say on any matter of importance. A remonstrator and a dissenter could possess a high moral ground, but they had no administrative means of annulling the ruler’s order. “Exclusive decision-making” (du duan 獨斷) remained the singular prerogative of the sovereign. If the principle of monarchism is inviolable, then how should one deal with an inept sovereign? This was not an idle question. Thinkers of the Warring States period were fully aware of the pitfalls of the system of individual rule. The major problem, pace common views of modern opponents of authoritarianism, was not necessarily the emergence of monstrous despots: first, because their appearance was infrequent, and, second, because ever since the early Zhou age it was legitimate to overthrow a tyrant (Pines 2008b). The real problem was not with monsters but with mediocrities. It was tacitly understood that the prevailing dynastic principle of rule was prone to produce less than brilliant individuals. In the age of the Warring States, in particular, as a meritocratic system of entry into officialdom replaced the pedigree-based aristocratic order (see later), the rulers remained the only executives who owed their position exclusively to their rights of birth and not to their abilities. In the middle Warring States period attempts were made to circumvent this problem by proposing non-hereditary power transfer, specifically by encouraging the ruler to abdicate in favor of a meritorious minister (Pines 2005a; cf. Allan 2015 and Allan 2016). These attempts failed, however, and the hereditary succession remained the rule. Practically, this meant that officials who served the ruler were frequently intellectually and morally superior to their master. The contradiction between political and intellectual authority, between the ruler and his aides, was the major pitfall of the monarchic form of rule. It generated persistent tensions, which are readily palpable even in the writings of the staunchest supporter of the ruler’s undisputed authority, Han Fei 韓非 (d. 233 bce) (Graziani 2015). How to resolve these tensions became the most challenging task. The solution can be divided into an overt and ideal one, and a more covert and practical one. Ideally, thinkers hoped for the coming of the True Monarch, a semidivinized sage (Puett 2002) who would stand at the apex of the moral and intellectual and not just political pyramid. Like the paragon monarchs of the past, the future True Monarch will unify All-under-Heaven under his rule; he will impose perfect social order, imbue his subjects with pure morality, and attain universal compliance. Xunzi summarizes: The [True] Son of Heaven is the most respectable in terms of his power and position and has no rivals under Heaven. . . . His morality is pure; his knowledge and kindness are extremely clear. He faces southwards and makes All-under-Heaven obedient. Among all the people, there is none who does not politely hold his hands following him, thereby being compliantly transformed. There are no recluses under Heaven, nobody’s goodness is neglected. He who unites with him is right, he who differs from him is wrong. (Xunzi XII.18: 331 [“Zheng lun”]) The expectation that a sage ruler will, by the sheer power of his intellect and morality, order All-under-Heaven was a lofty ideal, but it was not easily realizable. Mengzi acknowledged that 286

Political thought

the True Monarch is an exceptional personality, one who arrives “once in five hundred years” (Mengzi 4.13). In the meanwhile, insofar as the True Monarch does not materialize, the problem of inept sovereigns should be solved otherwise. The solution, outlined with greatest clarity by Xunzi, was that the sovereign should retain absolute ritual power, maintain his right of the final say, but not intervene in everyday policy making. In Xunzi’s view, the ruler should relegate his everyday tasks to a chancellor and confine himself to annual checks of the latter’s performance (Xunzi, “Wang ba”VII.11: 223–224; more in Pines 2009: 90–97). Not all thinkers accepted Xunzi’s specific solution: Han Fei for instance insisted that a chancellor may be a malevolent plotter, and relegating power to him will pave the way to usurpation. Yet to protect himself against usurpation, Han Fei recommended the ruler to preserve stillness and non-action, to remain secretive and dispel with any manifestations of his personal inclinations to the point of complete self-nullification. The argumentation differs markedly from that of Xunzi, but the bottom line remains the same: the monarch should minimize his intervention in everyday affairs (Pines 2013a). This bottom line is shared by the texts from all sides of the political spectrum of the Warring States period. Rationalizations for the nullification of the ruler’s personality and for his adherence to non-action differ: either the need to comply with the cosmic Way, or the advantages of following impersonal human law; either moral imperatives, or the need to preserve power against scheming ministers.Yet differences aside, the practical result of all these recommendations is an inactive ruler who refrains from everyday administrative tasks and who relegates de facto power to his ministers. The tension between the thinkers’ commitment to empowerment of the monarchic institution and their fear of malfunctioning monarchs was never adequately resolved. From the Qin’s First Emperor (emp. 221–210 bce) on, China’s monarchs routinely proclaimed themselves sages, implying thereby that they deserve utmost intellectual and not just political authority (Liu Zehua 2014, 2015). Some of them (most notably the First Emperor himself) took these claims quite literally, overpowering their courtiers and intervening actively in the realm’s life. Many others acquiesced – either willingly or grudgingly – to the conditions outlined by Xunzi, namely retaining symbolic supremeness but relegating everyday tasks to meritorious officials. Neither solution worked neatly. Not a few thinkers throughout the imperial period expressed their frustration with the persistent inadequacy of the throne’s occupants. Yet none of these thinkers – even such radical critics of the dynastic rule as Deng Mu 鄧牧 (1247–1306 ce) and Huang Zongxi 黃宗羲 (1610–1695 ce) – had ever departed from the ideological framework established during the Warring States period.2 The belief that common good can be attained only under the rule of a morally impeccable and selfless monarch remained unshakeable throughout the imperial millennia. Only the painful encounter with the West in the nineteenth century caused Chinese thinkers to start searching for alternatives to the monarchic system of rule (Pines 2017c; cf. Zarrow 2012).

Social structure: meritocracy, mobility, and hierarchy The Warring States period witnessed the demise of the aristocratic form of rule. An epochal event by itself, this decline of hereditary aristocracy became doubly important because of its intellectual implications. The proliferation of meritocratic discourse, the idea that an individual (more precisely a male individual) can transcend his social origins and advance the social ladder, and reconceptualization of the nature of social hierarchy – all these developments of the Warring States period exercised a lasting impact on Chinese political culture. The preceding Spring and Autumn period was the golden age of China’s hereditary aristocracy. Members of this stratum monopolized political, social, and military power in every 287

Yuri Pines

polity (Zhu Fenghan 1990: 525–593). Elaborate ritual system developed in the latter half of the Western Zhou period assigned every noble duties and privileges according to his birthright (Falkenhausen 2006: 29–73). That system aimed to solidify the hereditary social order and perpetuate it indefinitely. Moreover, back then the aristocrats enjoyed intellectual authority as well. Insofar as we can judge from the major historical work that deals with this age, Zuo zhuan 左傳, it was nobles and only nobles who defined rights and wrongs in the political and ethical realm (Pines 2002; cf. Schaberg 2001; Li Wai-yee 2007). By the sixth century bce the aristocrats’ power reached its apogee. It became all but impossible for outsiders to enter high echelons of power, and even the rulers could no longer effectively challenge leading nobles. In not a few states (e.g., Lu 魯, Jin 晉, and Zheng 鄭) the coalition of a few aristocratic lineages established a de facto ruling oligarchy which completely eclipsed local lords. Yet it was precisely then that the crisis of aristocracy began. As aristocratic lineages destroyed each other in internecine struggles, the lords (some of whom were former ministers who usurped power in their home states) found it expedient to appoint members of the lower nobility – the shi – to positions of power. Parallel to this, the ritual system, which was eroded due to persistent transgressions of ritual norms by powerful nobles, lost its stabilizing impact as well, especially as members of the sub-elite, the shi, began appropriating sumptuary rights of their superiors, blurring the dividing line between high and petty nobility (Falkenhausen 2003 and 2006: 370–399). Even earlier, back in the second half of the Spring and Autumn period, the belief in the importance of pedigree had been shaken already, as increasing number of noble scions proved to be intemperate, inept, or otherwise inadequate, causing their lineage’s downfall. It was then that some of the nobles started revaluing the prestigious designation “noble men” (junzi 君子). Henceforth, it was not automatically applied to every person of noble birth, but only to those who excelled intellectually and morally. Nobles who misbehaved were treated as “petty men” (xiao ren 小人).This ethical interpretation of status belonging had its limits: an inept noble could be recognized as a petty man, but a shi would never be called a noble man, at least not in Zuo zhuan (Pines 2002: 165–171).Yet the seeds of change were already sown. It is against this backdrop that the shi revolution began.The first known thinker who belonged to the shi stratum, Confucius, lived at the heyday of the aristocratic age and dared not openly challenge the pedigree-based social order.Yet his major contribution to the demise of this order was reconceptualization of the term junzi as an overwhelmingly ethical term: a designation of a moral and self-cultivated person, including – primarily – the shi (cf. Gassman 2007; Brindley 2009; Pines 2017b). By shifting an emphasis from pedigree to self-cultivation as the way to attain the “noble man” status, Confucius undermined the very foundations of the aristocratic order. A century later, Mozi was much more resolute in rejecting this order altogether. Mozi put forward the slogan of “elevating the worthy” (shang xian 尚賢): the country should be governed by the most able people, whatever their social origins are. According to Mozi, when the former sage kings implemented this policy, neither the officials were perpetually esteemed, nor the people forever base. . . . At that time, even among those ministers who enjoyed rich emoluments and respected position, none was irreverent and reckless, and each behaved accordingly; even among peasants and artisans, each was encouraged to enhance his aspirations. (Mozi II.8: 67–68 [“Shang xian shang”]) Mozi is unequivocal: even among the low strata of peasants and artisans, some people may contribute to the state’s well-being; accordingly, there should be no limitations at all on social mobility, and one’s position should reflect exclusively one’s worthiness and righteousness. What 288

Political thought

is amazing is that this radical assault on centuries-old hereditary rights apparently went unopposed, without traceable attempts to defend the pedigree-based social order. Soon enough, the idea that one’s career should be determined by abilities alone and not by birthrights became a consensus. Even thinkers who came from high aristocratic families, such as Shang Yang (a scion of the Wei 衛 ruling house) and Han Fei (a scion of the Han 韓 ruling house), did not protect the pedigree-based order but rather dismissed it entirely. Shang Yang was particularly renowned as a promoter of radical social restructuring in theory and practice. The new system of ranks of merit designed by him radically reshaped the Qin society, introducing an unprecedented degree of social mobility (Pines et al. 2014: 24–26). Similar, even if less radical, transformations occurred throughout the rest of the Warring States. That heredity should not determine the elite belonging was acceptable to most if not all thinkers; but how this belonging should be determined was a matter of bitter controversy. For Confucius and his followers the answer was clear: the noble men’s status should be determined primarily by his education and moral qualities. A noble man is he who is able to either refine his innate goodness (Mengzi) or overcome his innate greediness and selfishness (Xunzi). But how to distinguish between real and fraudulent “noble men”? How to ensure that moralizing discourse would not be manipulated by unscrupulous career-seekers? This issue preoccupied the Confucians throughout the period under discussion and much beyond. Some optimistically expected that a superior would immediately recognize the noble man’s true worth (Henry 1987). Others in distinction proposed to diagnose a man’s character through a series of observances and tests that would explore his sincerity, his will, external expressions of his feelings, his countenance, his hidden motivations, and the matching between his words and deeds (Richter 2005 and 2017). Answers varied, but fundamentally the Confucians’ desire was that the noble man’s true value would be determined by his peers, i.e., ideally by Confucian Masters themselves. Thinkers whom we now call Legalists could not disagree more. For them, the very idea that elite members will determine themselves who deserves the elite status meant weakening the ruler’s authority and empowering self-serving talkative intellectuals. The Legalists dismissed the core Confucian belief in the possibility that a cultivated “noble man” transforms himself into a moral political actor. Rather, the political system should be based on the premise that everybody seeks his own interest only. The properly functioning state should not try to better the subjects but rather to channel their quest for riches and social prestige toward desirable social and political ends (cf. Pines 2017a: 59–99 for Shang Yang; Harris 2016: 11–62 for Shen Dao; Goldin 2013 for Han Fei). Accordingly, it is up to the rulers to staff the elite with those deemed useful, be they valiant fighters (Shangjunshu 17.2) or skilled civilian and military officials who started with a low-level bureaucratic or military job and were promoted according to their performance (Han Feizi XX.50: 1137 [“Xian xue”]).The nature of “merits” as understood by Confucians and Legalists was bitterly contested, but the principle of meritocracy as superior to hereditary rights was accepted overwhelmingly, notwithstanding a few dissenting voices (e.g. Mengzi 2.7; see more in Pines 2013b). Justification of social mobility in the Warring States period thought was potentially detrimental to the idea of social hierarchy which was so powerfully embedded in the Zhou ritual system. The very openness of the shi stratum to ambitious newcomers from below was conducive to reduction of social barriers. By the late Warring States period shi proudly adopted self-designation as “plain-clothed” (buyi 布衣), emphasizing thereby that they came up from the bottom of society. The anecdotes of that age tell of have-nots who by diligent learning succeeded to overcome negative circumstances and make an illustrious career (Pines 2009: 141–145). These anecdotes should be read cum grano salis, but the degree of social mobility in the third century bce was indeed remarkable, as testified for instance from Qin documents (Pines et al. 2014: 289

Yuri Pines

24–26). How to reconcile this mobility and the breakdown of hereditary order with the need to preserve social hierarchy was a formidable task. Of many thinkers who proposed solutions to this challenge, Xunzi appears as most significant. Xunzi resurrected the by then semi-abandoned notion of ritual (li 禮) as a universal social regulator (Pines 2000b).Yet in distinction from earlier texts, such as Zuo zhuan, ritual in Xunzi is predicated not on the maintenance of hereditary distinctions but on regulating a society in which everybody can advance. Xunzi clarifies: Although a man is the descendant of kings, lords, shi and nobles, if he does not observe the norms of ritual and propriety, he must be relegated to the status of the commoner; although he is a descendant of a commoner, if he accumulates learning of the texts, rectifies his behavior, and is able to observe the norms of ritual and propriety – then he must be elevated to the rank of high ministers, shi and nobles. (Xunzi V.9: 148–149 [“Wang zhi”]) The statement is unequivocal: rather than utilizing ritual as an impediment to social mobility, Xunzi employs ritual behavior as a substitute to pedigree, turning it into the primary criterion of attaining appropriate social status. The thinker explains elsewhere why the pedigree cannot serve as an adequate determinant of one’s position: it is because everybody – from the most revered paragons to the despicable “petty men” – possesses the same inborn qualities. Only through learning and self-cultivation can one transform himself into a “noble man” (Xunzi II.4:61 [“Rong ru”]).Yet once one made a choice, he must accept its consequences: he should work hard to overcome his innate greediness and selfishness. If successful, the “noble man” deserves a position at the top of society and manifold social, economic, and political privileges. In distinction, the “petty men” deserve only protection of their basic economic interests but should forever remain at the society’s bottom. To a certain extent Xunzi’s combination of rigid social hierarchy with considerable meritocratic mobility outlined China’s social desideratum (even if not necessarily social practice) for millennia to come.

An intellectual and the state The rise of the shi was not just a political and social phenomenon: it was accompanied by profound reconceptualization of the nature of intellectual authority. In the Spring and Autumn period, the rulers’ courts served as a locus of intellectual activity. From the time of Confucius, Mozi, and their disciples, this locus shifted from the rulers to the shi Masters (zi 子).The Masters’ intellectual authority derived not from their position of power, which they often lacked, but rather from their superior understanding of the Way (Dao), i.e., of the guiding moral, sociopolitical, or cosmic principles essential to the well-being of the state and a single person. “Possessing the Way” was the prerogative of outstanding intellectuals, not of the rulers or their courtiers. To put it differently, the shi combined the roles of political and intellectual elite.This double role had greatly bolstered their self-confidence. The Lüshi chunqiu, composed on the eve of imperial unification by a group of shi eager to promote their stratum, reflects this: Shi are the men who, when acting in accord with [proper] patterns, do not escape the difficulties; when facing the troubles, forget the profits; they cast aside life to follow righteousness and consider death as returning home. If there are such men, the ruler of a state will not be able to befriend them, the Son of Heaven will not be able to make them servants. At best, stabilization of All-under-Heaven, or, second to it, stabilization 290

Political thought

of a single state must come from these men. Hence a ruler who wants to attain great achievements and fame cannot but devote himself to searching for these men. A worthy sovereign works hard looking for [proper] men and rests maintaining affairs. (Lüshi chunqiu 12.2: 622–623) This passage is plain and unsophisticated. First, it hails the high morality of the shi, who prefer righteousness to gains and even to life. Second, it hails their loftiness: the mere ruler of a state would be unable to befriend them, and the Son of Heaven would fail to turn them into servants. Then the authors go to the most important part of their message: they advise the ruler to acquire the services of these lofty shi as the one necessary precondition for overall success. With these servants, the ruler will rest – presumably because the worthy aides will maintain affairs in his stead. The passage’s message contains therefore as an overt contradiction. The shi simultaneously proclaim their aloofness of the rulers, whom they treat as inferiors, but also seek employment by these very rulers. These contradictory messages reflect the complex social and political standing of shi in general and of their Masters in particular. On the one hand, they remained forever economically and socially dependent on the throne; besides, for many of them government service was the only way for realization of their lofty ideals. On the other hand, proud in their perceived intellectual and moral superiority over the rulers, these shi intellectuals refused to acquiesce to the position of the ruler’s servitors. This generated persistent tension between the shi desire to serve the rulers and their hope to maintain personal dignity and self-respect in the monarchic political order that they themselves helped to design. This tension between the desire to serve and the insistence on doing it on one’s own terms permeates the writings of Confucius and his disciples. For them service was a means of moral self-realization: the noble mission through which their deepest aspirations could be fulfilled. Confucius promised: “[O]ne who would employ me will attain results within a year, and [the tasks] will be completed within three years,” and he was willing to serve even politically dubious figures insofar as this could allow revival of the “Zhou in the east” (Lunyu 13.10, 17.5, 17.9). Mengzi claimed that through rectifying the ruler’s heart he would be able to impose universal morality (Mengzi 7.20). He further argued that government service is an essential occupation of a shi, just like tilling is for the peasants; a shi who remains without an office for three months should be consoled (Mengzi 6.2). This dedication to service notwithstanding, neither Confucius nor Mengzi succeeded in holding an office for long; their life was a series of appointments and subsequent resignations, or unsuccessful appointments. The major reason for their career failure, if we judge from their disciples’ accounts, was the Masters’ insistence on preserving their integrity and their unwillingness to compromise lofty principles in exchange for successful careers. Confucius’ loyalty to a ruler was secondary to his commitment to the Way; hence he stated that the Great Minister is the one who “serves the ruler according to the Way, and when it is impossible, stops [serving]” (Lunyu 11.24). Elsewhere Confucius clarifies: Riches and honors are what every man desires; but if they cannot be attained in accordance with the Way, do not accept them. Poverty and base status are what every man detests. But if they cannot be avoided in accordance with the Way, do not avoid them. (Lunyu 4.5) Confucius recognizes legitimate aspirations of his fellow shi for riches and honor; these by themselves are not shameful.Yet fulfilling these aspirations requires commitment to the supremeness 291

Yuri Pines

of one’s moral Way; whenever this Way is compromised, one should avoid service, even if this means poverty and social debasement. Drawing the line between legitimate and illegitimate employment – a topic discussed at length by Mengzi as well (e.g., Mengzi 12.14) – meant distinguishing oneself from despicable petty man, who served only for the sake of self-enrichment. Mengzi equates these latter yes-men to concubines who lack personal integrity. His ideal is different: it is a Great Man whom “wealth and high status cannot tempt, poverty and low status cannot move, awesomeness and military might cannot subdue” (Mengzi 6.3). Confucius’ and Mengzi’s frustration with the rulers who invariably failed to live up to the Masters’ expectations was shared by other lofty shi, some of whom took it to the extreme.These purists considered any service as a filthy task and preferred to shun the courts altogether (Vervoorn 1990). A few thinkers, such as contributors to the Laozi and Zhuangzi, criticized the mere idea that political service should be a means of self-realization. In the eyes of these critics the service was first, futile; second, dangerous; and third, potentially immoral: after all, lofty proclamations aside, too many shi (Confucians included) sought only riches and honor, not the Way (see, e.g., Zhuangzi, “Dao Zhi”). Whatever the immediate impact of this criticism of political involvement, eventually it did fuel the concept of lofty reclusion which proliferated in the early imperial period, allowing its adherents to hold high moral ground (Berkowitz 2000). Facing this criticism of their integrity, some of the followers of Confucius, most notably Mengzi, adopted an uncompromisingly critical stance toward the rulers. This strong criticism was a means to demonstrate to fellow shi (and possibly to the ruler as well) that one serves him only to promote the Way and is not concerned with personal career. Mengzi’s haughty, even confrontational tone adopted in conversations with the monarchs is at times truly astonishing (Pines 2013c: 80–89). If these repeated affronts to the thinker’s employers really took place and were not fabricated by Mengzi’s followers, then we should admit that the rulers of the age were remarkably tolerant. Even when Mengzi raised an issue of an interlocutor’s inadequacy and the possible need to replace the malfunctioning sovereign, the latter just “turned to his attendants and changed the subject” (Mengzi 2.6). Perhaps the fear that persecuting an outspoken advisor will cause massive exodus of gifted shi from the oppressive court and the resultant brain drain moderated the rulers’ response.Yet the confrontational attitude adopted by Mengzi and his like was not a proper way of maintaining relation with the monarchs. It not just alienated the rulers, but, more importantly, it undermined the overarching principle of monarchism, to which Mengzi, as well as other followers of Confucius, adhered. This contradiction was duly noticed by those thinkers who advocated unequivocal monarchoriented political ethics. Shang Yang, Han Fei, and their like conceived of ruler-minister relations as purely political: the ruler was to command, the minister had to obey.The minister could be morally and intellectually superior to the ruler, but this gave him no extra rights. A chapter from a multi-authored text, the Guanzi 管子, says: Hence when one respectfully implements the ruler’s orders, even if he is hurt and defeated, he is not to be punished; while if one implements what the ruler did not order, even if he succeeds he must be punished by death. Thereby the inferiors will serve the superiors as an echo responds to a sound, and ministers will serve the sovereign as a shadow follows the body . . . . This is the Way of orderly rule. (Guanzi XV.45: 912–913 [“Ren fa”]) This strictly authoritarian view which reduced a minister to the ruler’s tool might have fit well the ideology of monarchism, but it was incompatible with the ministerial self-confidence and pride of the Warring States period. It was up for another follower of Confucius, Xunzi, to find a 292

Political thought

middle ground. Xunzi did assert the minister’s right to defy the ruler’s orders, but put clear limits to this defiance: it is legitimate only insofar as the results are beneficent to the ruler and the state: “he who contradicts the orders and benefits the ruler is called loyal.” Echoing Confucius, Xunzi promulgated the ideal of “Follow the Way, do not follow the ruler” (Xunzi XI.13: 250 [“Chen Dao”]), but he was cautious to prevent this dictum from jeopardizing ruler-minister relations. Hence, Xunzi renounced the provocative rhetoric employed by Mengzi (Pines 2013c: 89–94) and also deemphasized the possibility of shifting allegiance to another court whenever a thinker felt himself offended (Pines 2009: 177–180). Xunzi’s reinterpretation of a minister’s obligations to the ruler anticipated the dominant situation under the unified empire. The minister should retain a degree of autonomy and self-respect: but he must be a ruler’s critical servant, neither his friend nor his teacher.

The commoners: “People as a root” One of Mengzi’s most famous statements proclaims: The people are the most esteemed; the altars of soil and grain follow them, and the ruler is the lightest. Hence one who attains [the support of] the multitudes, becomes Son of Heaven; one who attains [the support of] the Son of Heaven, becomes a regional lord; one who attains [the support of] the regional lord, becomes a noble. (Mengzi 14.4) This statement, which read out of context may well resemble a proclamation of the people’s sovereignty, exemplifies the so-called “people as a root” (minben 民本) ideology in early China. This ideology attracted the attention of Chinese thinkers and foreign observers since the early twentieth century as part of debates about the compatibility of Confucianism with Western democratic ideals.3 Yet before its modern implications can be considered, one should first understand this people-oriented ideology in the context of the Warring States-period thought. The notion that the people are the polity’s “root” appeared very early: it can be traced already to the texts associated with the Western Zhou period, and it is fully present in the Spring and Autumn period speeches cited in Zuo zhuan (Pines 2009: 187–197). In the Warring States period, this tendency of emphasizing the people’s importance continued. I shall not focus here on ubiquitous calls for the rulers to address the people’s needs, most prominently their welfare and personal security, because these ideas are a commonplace in political thought worldwide and do not require further discussion. Yet two points deserve our attention: first is conceptualization of “the people” – referring primarily, albeit not exclusively, to the commoners – as the raison d’être of the polity, and second is the emphasis on their role as kingmakers, i.e., as a potentially active political force. The idea that “the people” are the singularly important component of the polity is pronounced in many texts, including those which advocate overtly oppressive policies toward them. For instance, the Book of Lord Shang, attributed to Shang Yang, is notorious for its appalling pronouncements against traditional morality and culture and in favor of blatant militarism; and the authors at the time delight in presenting themselves as people-bashers and supporters of excessively restrictive control (e.g., Shangjunshu 18.2).Yet the authors also remind the ruler that the goal of their policies is to benefit the people and not to benefit the ruler personally. “When [the ancient paragons] Yao and Shun were established in All under Heaven, this was not to benefit privately from All-under-Heaven: they were established for the sake of All-under-Heaven” (Shangjunshu 14.4). War, oppression, and harsh punishments are just the means to attain peace 293

Yuri Pines

and tranquility, demonstrating thereby the ruler’s love of the people. Other texts disagree with the Book of Lord Shang as to the proper means of benefitting the people but share the understanding that their needs should top the ruler’s concern. This view is neatly summarized in a saying from the Lüshi chunqiu: “All under Heaven does not [belong to] a single man, it [belongs] to All under Heaven” (Lüshi chunqiu 1.4: 44). Xunzi reiterates: “Heaven does not give birth to the people for the ruler’s sake; it establishes the ruler for the people’s sake” (Xunzi XIX.27: 504 [“Da lue”]). Proclamations about the people’s ultimate importance for the polity are matched by the assertion that they can influence the outcome of political struggles. Mengzi declared that the ancient tyrants Jie and Zhou[xin] lost All under Heaven through losing the people. They lost the people through losing their hearts. There is a way to attain All under Heaven: when you attain the people, you attain All under Heaven.There is a way to attain the people: when you attain their hearts, you attain the people. (Mengzi 7.9) This emphasis on attaining the people’s hearts, i.e., on making the policies acceptable to them, can be encountered in many other texts. Even the Book of Lord Shang demands of the ruler to pay attention to the people’s internalization of laws and regulations: without this broad approval, the efficiency of laws will decline (Shangjunshu 5.9). The Lüshi chunqiu declares that the success of the former paragons derived primarily from their ability to “be compliant with the people’s hearts” (Lüshi chunqiu 9.2: 478). Xunzi summarizes: “The ruler is a boat; commoners are the water. The water can carry the boat; the water can capsize the boat” (Xunzi V.9: 152 [“Wang zhi”]). Xunzi’s latter saying is often read in the context of legitimating popular rebellion, and similar interpretations are often given to Mengzi’s pronouncements, but this should not be the case (Tiwald 2008). Popular rebellions, which became ubiquitous since the downfall of the first imperial dynasty, the Qin, did not happen in the Warring States or earlier periods. The leverage of the commoners was different: the rulers had to take into consideration that disgruntled people may vote with their feet. If unsatisfied with the ruler, peasants would flee to a neighboring state and deplete the sovereign of human resources, while the conscripts would abscond from the battlefield (see, e.g., Mengzi 1.3). It was primarily to prevent this outcome that the ruler was recommended to be attentive to popular sentiments. Repeated pronouncements in favor of the commoners’ importance and in favor of being attentive to their opinions may create an impression of nascent ideas of popular sovereignty in early Chinese thought, but this is a debatable conclusion. Actually, thinkers who propagated the importance of “attaining the people’s heart” did not advance any mechanism of consulting the people or even verifying their ideas. Only in the Mozi can we distinguish a nascent idea of routinely consulting the commoners and even allowing them to supervise low-level power holders (e.g., Mozi III.12: 116–118 [“Shang tong zhong”]). These ideas are exceptional, though. The other thinkers’ neglect of institutionalizing the people’s political input is not incidental. After all, the very same thinkers who proclaimed the people’s importance appear to have a very low esteem of the people’s mental abilities and morality. Confucius argues straightforwardly: “You can let the people follow [the Way], but not understand it” (Lunyu 8.9). Mengzi adds: “Slight is the difference between men and beasts and birds. Commoners abandon it; noble men preserve it” (Mengzi 8.19). Clearly, intellectually and morally impaired commoners should not be expected to actively participate in decision making. Mengzi is explicit: “Some toil with their 294

Political thought

hearts, some toil with their force. Heart-toilers rule men; force-toilers are ruled by men. Those who are ruled by men, feed men; those who rule men, are fed by men – this is the common propriety of All under Heaven” (Mengzi 5.4). Xunzi echoes these views (Xunzi VI.10: 182 [“Fu guo”]). How can we reconcile, then, the thinkers’ insistence on the people as “the most esteemed” with their disdain toward commoners and rejection of their political participation? A possible answer would be a sinister one, proposed by Han Fei: should the ruler be able to govern simply by reliance on “the people’s heart,” he would have no use of worthy advisors from among menof-service (Han Feizi XX.50: 446 [“Xian xue”]). Indeed, for intellectuals it was advantageous to appropriate the position of the people’s representatives and to voice the people’s grievances to the ruler, which would grant them additional leverage vis-à-vis the sovereigns. This appropriation of what Tu Wei-ming aptly defines as “the most generalisable social relevance (the sentiments of the people)” (Tu Wei-ming 1993: 20) was too important an asset to be yielded to the uneducated masses. It was in the best interest of the self-proclaimed champions of the people from among the educated elite to keep commoners precluded from political processes. Alternatively, a less sinister explanation is possible. In the highly mobile society of the Warring States period, ambitious commoners were able to join the ranks of the elite and become legitimate political players.Those who remained behind evidently lacked either sufficient ambitions, or talents, or both – or so, at least, most shi wanted to believe. As commoners were no longer hermetically excluded from the ruling elite, and as some of them were routinely coopted into the shi stratum, this may have created a kind of “popular representation” from above, which eliminated the need for active political participation from below. It was only with the cessation of this mobility after the imperial unification that alternative modes of political participation from below emerged, namely popular unrest and popular rebellion (Pines 2012: 134–161).

Summary: intellectual legacy of the warring states The imperial unification of 221 bce may be viewed as partial fulfillment of the aspirations of the Warring States period thinkers. Even though the first imperial dynasty, the Qin, was short-lived, its heir, the Han dynasty, brought about a relatively long period of peace, stability, and prosperity to large parts of the East Asian subcontinent. And, while China still had to suffer from longer or shorter periods of fragmentation, rebellions, foreign conquests, and domestic wars, the imperial model proved to be viable enough to remain unchallenged well until the end of the nineteenth century. Among many factors that contributed to the empire’s durability was its unparalleled ideological preparedness. Its fundamental premises were shaped long before the actual imperial unification and retained their appeal for the subsequent two odd millennia. The emperor should be omnipotent and his rule should be universal; the bureaucracy should be staffed by men of proven talent and merit; and the commoners deserve utmost concern but should remain outside policy making.These ideas guided political actors in China from the beginning to the end of the imperial enterprise, from the Qin dynasty to the Qing (1636/1644–1912).They were shared not only by the political and intellectual establishment but even by foreign conquerors and domestic rebels. The fact that no alternative to these fundamental premises was posed throughout more than twenty-one centuries of the imperial rule testifies to the exceptional viability of the ideas shaped prior to the imperial unification. Perhaps the reason for their appeal is that broad consensus in favor of these ideas was formed during the freest period in China’s intellectual history and they were not artificially imposed from above. Later, the empire’s political and intellectual establishment further reinforced the hegemonic status of the imperial ideology, making it the 295

Yuri Pines

glue that held the imperial enterprise intact even during the periods of most profound domestic crises. Taken from this perspective, the intellectual enterprise of the Warring States period thinkers should be considered an exceptional success.Yet this success came at a price. The imperial unification brought an end to the inter-state market of talent which ensured considerable freedom of thought and freedom of expression to men-of-service.With the establishment of the imperial monopoly on power and prestige, first steps were made to subjugate scholars to the state. In the Qin this was done crudely through outright suppression of “private learning” in the wake of the infamous biblioclasm of 213 bce (for which see Petersen 1995; Kern 2000: 183–196; Pines 2009: 180–183). Han rulers preferred cooptation to coercion: by establishing a state-sponsored curriculum based on the Five Classics (Nylan 2001), they brought about elevation of these texts above those of the “Hundred Schools of Thought” (Lewis 1999b: 337–360). In due time, as intellectual orthodoxy came into existence; freedom of thought and expression was significantly curtailed. While intellectual debates continued throughout the imperial millennia, their scope and boldness reduced in comparison to an earlier age. To express themselves more freely, the imperial literati employed other genres than composing political essays. The latter continued to be produced, to be sure, but their intellectual appeal could rarely match that of pre-imperial Masters. Yet the Warring States period legacy of proud and self-confident intellectuals did not disappear entirely. The throne never succeeded to turn all the literati into its obedient tools; nor did it succeed in the long term to determine unilaterally norms of orthodoxy. The ongoing appeal of the Warring States-period texts, especially those that were incorporated into the expanding corpus of Classics (e.g., the Analects and Mengzi), fueled the intellectuals’ self-esteem. Their conviction that in the final account their stratum – rather than the occupants of the throne – represented the Way made them more confident, more willing to criticize the emperors, and more able to influence the government’s policies.This belief of acting on behalf of the Way generated repeated collisions between lofty intellectuals and the throne, but it also greatly benefitted the imperial enterprise. It provided the empire with an extraordinary powerful, confident, and self-aware stratum of scholar-officials who were able to navigate its course even under most inadequate rulers (Pines 2012: 76–103).The persistence of this stratum is one of the major legacies of the Warring States-period intellectual culture, which contributed in the final account to exceptional durability of the imperial enterprise.

Notes * This research was supported by the Israel Science Foundation (grant No. 240/15) and by the Michael William Lipson Chair in Chinese Studies. 1 Note a significant semantic overlap between the Western term “intellectual” and the Chinese shi (Yu Yingshi 1987: 1–3). 2 For views of rulership during the imperial millennia, see Liu Zehua 1996, vols. 2–3 (specifically for Deng Mu and Huang Zongxi, see 3: 412–416 and 600–618). For Huang Zongxi, cf. de Bary 1993 and Jiang Yonglin 2008. 3 For early explorations of the potential relevance of the minben idea to modern democracy, see Liang Qichao 1919 [1996]: 35–44 and 228–234; for modern debates among Western scholars, see, e.g., Murthy 2000; Tan 2003: 132–156.

Works cited Allan, Sarah. 2015. Buried Ideas: Legends of Abdication and Ideal Government in Recently Discovered Early Chinese Bamboo-Slip Manuscripts. Albany, NY: State University of New York Press.

296

Political thought Allan, Sarah. 2016. The Heir and the Sage: Dynastic Legend in Early China, Revised and Expanded Edition. Albany, NY: State University of New York Press. Berkowitz, Alan. 2000. Patterns of Disengagement:The Practice and Portrayal of Reclusion in Early Medieval China. Stanford: Stanford University Press. Brindley, Erica. 2009. “ ‘Why Use an Ox-Cleaver to Carve a Chicken?’ The Sociology of the junzi Ideal in the Lunyu.” Philosophy East and West 59.1: 47–70. Bünger, Karl, 1987. “Concluding Remarks on Two Aspects of the Chinese Unitary State as Compared with the European State System.” In Foundations and Limits of State Power in China, ed. Stewart R. Schram, 313–323. Hong Kong: The Chinese University Press. Chunqiu Gongyang zhuan yizhu 春秋公羊傳譯注. 2011.Annotated by Liu Shangci 劉尚慈. Beijing: Zhonghua shuju. Chunqiu Zuozhuan zhu 春秋左傳注. 1990. Annotated by Yang Bojun 楊伯峻. Beijing: Zhonghua shuju. Csikszentmihalyi, Mark, and Michael Nylan. 2003. “Constructing Lineages and Inventing Traditions Through Exemplary Figures in Early China.” T’oung Pao 89.1–3: 59–99. De Bary, William Theodore. 1993. Waiting for the Dawn: A Plan for the Prince: Huang Tsung-hsi’s Ming-i-taifang lu. New York: Columbia University Press. Di Cosmo, Nicola. 2002. Ancient China and Its Enemies: The Rise of Nomadic Power in East Asian History. Cambridge: Cambridge University Press. Dorofeeva-Lichtman,Vera. 2009. “Ritual Practices for Constructing Terrestrial Space (Warring States-Early Han).” In Early Chinese Religion, Part One: Shang through Han (1250 bc-220 ad), eds. John Lagerwey and Marc Kalinowski.Vol. 1, 595–644. Leiden: Brill. Falkenhausen, Lothar von. 2003. “Social Ranking in Chu Tombs: The Mortuary Background of the Warring States Manuscript Finds.” Monumenta Serica 51: 439–526. Falkenhausen, Lothar von. 2006. Chinese Society in the Age of Confucius (1050–250 bc): The Archeological Evidence. Los Angeles: Cotsen Institute of Archaeology at UCLA. Gassmann, Robert H. 2007. “Die Bezeichnung jun-zi: Ansätze zur Chun-qiu-zeitlichen Kontextualisierung und zur Bedeutungsbestimmung im Lun Yu.” In Zurück zur Freude: Studien zur chinesischen Literatur und Lebenswelt und ihrer Rezeption in Ost und West: Festschrift für Wolfgang Kubin, ed. Hermann Marc and Christian Schwermann, 411–436. Monumenta Serica Monograph Series 57. Sankt Augustin: Steyler Verlag. Gentz, Joachim. 2015. “Long Live The King! The Ideology of Power between Ritual and Morality in the Gongyang zhuan 公羊傳.” In Ideology of Power and Power of Ideology in Early China, eds.Yuri Pines, Paul R. Goldin and Martin Kern, 69–117. Leiden: Brill. Goldin, Paul R. 2011a. “Persistent Misconceptions about Chinese ‘Legalism’.” Journal of Chinese Philosophy 38.1: 64–80. ———. 2011b. “Steppe Nomads as a Philosophical Problem in Classical China.” In Mapping Mongolia: Situating Mongolia in the World from Geologic Time to the Present, ed. Paula L.W. Sabloff, 220–246. Philadelphia: University of Pennsylvania Museum of Archaeology and Anthropology. ———. 2013. “Han Fei and the Han Feizi.” In Dao Companion to the Philosophy of Han Fei, ed. Paul R. Goldin, 1–21. Dordrecht: Springer. ———. 2015. “Representations of Regional Diversity during the Eastern Zhou Dynasty” In Ideology of Power and Power of Ideology in Early China, eds. Yuri Pines, Paul R. Goldin and Martin Kern, 31–48. Leiden: Brill. Gongyang zhuan. See Chunqiu Gongyang zhuan. Graziani, Romain. 2015. “Monarch and Minister:The Problematic Partnership in the Building of Absolute Monarchy in the Han Feizi.” In Ideology of Power and Power of Ideology in Early China, eds.Yuri Pines, Paul R. Goldin and Martin Kern, 155–180. Leiden: Brill. Guanzi jiaozhu 管子校注. 2004. Compiled by Li Xiangfeng 黎翔鳳. Beijing: Zhonghua shuju. Harris, Eirik Lang. 2016. The Shenzi Fragments: A Philosophical Analysis and Translation. New York: Columbia University Press. Henry, Eric, 1987. “The Motif of Recognition in Early China.” Harvard Journal of Asiatic Studies 47.1: 5–30. Jaspers, Karl. 1965. The Origin and Goal of History. Translated by Michael Bullock. New Haven, CT: Yale University Press. Jiang,Yonglin. 2008. “Denouncing the ‘Exalted Emperor’: Huang Zongxi’s uses of Zhu Yuanzhang’s Legal Legacy in Waiting for the Dawn.” In Long Live the Emperor: Uses of the Ming Founder across Six Centuries of East Asian History, ed. Sarah Schneewind, 245–274. Minneapolis: Society for Ming Studies. Kern, Martin. 2000. The Stele Inscriptions of Ch’in Shih-huang: Text and Ritual in Early Chinese Imperial Representation. New Haven, CT: American Oriental Society.

297

Yuri Pines Lewis, Mark Edward. 1990. Sanctioned Violence in Early China. Albany: State University of New York Press. ———. 1999a. “Warring States: Political History.” In The Cambridge History of Ancient China: From the Origins of Civilization to 221 B.C., eds. Michael Loewe and Edward L. Shaughnessy, 587–650. Cambridge: Cambridge University Press. ———. 1999b. Writing and Authority in Early China. Albany: State University of New York Press. ———. 2006. The Construction of Space in Early China. Albany: State University of New York Press. Li, Wai-yee. 2007. The Readability of the Past in Early Chinese Historiography. Cambridge, MA: Harvard University Asia Center. Liang, Qichao 梁啟超. 1919 (1996). Xian Qin zhengzhi sixiang shi 先秦政治思想史. Rpt. Beijing: Dongfang. Liu, Zehua 劉澤華, ed. 1996. Zhongguo zhengzhi sixiang shi 中國政治思想史. 3 vols. Hangzhou: Zhejiang renmin chubanshe. Liu, Zehua. 2014. “The Monarch and the Sage: Between Bifurcation and Unification of the Two,” translated by Yuri Pines. Contemporary Chinese Thought 45.2–3: 55–88. ———. 2015. “Political and Intellectual Authority: The Concept of the ‘Sage Monarch’ and its Modern Fate.” In: Ideology of Power and Power of Ideology in Early China, eds.Yuri Pines, Paul R. Goldin, and Martin Kern, 273–300. Leiden: Brill. Lunyu yizhu 論語譯注. 1992. Annotated by Yang Bojun 楊伯峻. Beijing: Zhonghua shuju. Lüshi chunqiu jiaoshi 呂氏春秋校釋. Compiled and annotated by Chen Qiyou 陳奇猷. Shanghai: Xuelin, 1990. Mengzi yizhu 孟子譯注. 1992. Annotated by Yang Bojun 楊伯峻. Beijing: Zhonghua shuju. Mozi jiaozhu 墨子校注. 1994. Compiled and annotated by Wu Yujiang 吳毓江 (1898–1977). Beijing: Zhonghua shuju. Murthy, Viren. 2000. “The Democratic Potential of Confucian Minben Thought.” Asian Philosophy 10. 1: 33–47. Nylan, Michael. 2001. The Five “Confucian” Classics. New Haven, CT:Yale University Press. Petersen, Jens Østergård. 1995. “Which Books Did the First Emperor of Ch’in Burn? On the Meaning of Pai Chia in Early Chinese Sources.” Monumenta Serica 43: 1–52. Pines,Yuri. 2000a. “ ‘The One That Pervades the All’ in Ancient Chinese Political thought: The Origins of ‘The Great Unity’ Paradigm.” T’oung Pao 86.4–5: 280–324. ———. 2000b. “Disputers of the Li: Breakthroughs in the Concept of Ritual in Pre- Imperial China.” Asia Major 13.1: 1–41. ———. 2002. Foundations of Confucian Thought: Intellectual Life in the Chunqiu Period, 722–453 b.c.e. Honolulu: University of Hawai'i Press. ———. 2005a. “Disputers of Abdication: Zhanguo Egalitarianism and the Sovereign’s Power.” T’oung Pao 91.4–5: 243–300. ———. 2005b. “Beasts or Humans: Pre-Imperial Origins of Sino-Barbarian Dichotomy.” In Mongols,Turks, and Others: Eurasian Nomads and the Sedentary World, eds. Reuven Amitai and Michal Biran, 59–102. Leiden: Brill. ———. 2008a. “Imagining the Empire? Concepts of ‘Primeval Unity’ in Pre-Imperial Historiographic Tradition.” In Conceiving the Empire: China and Rome Compared, eds. Fritz-Heiner Mutschler and Achim Mittag, 67–90. Oxford: Oxford University Press. ———. 2008b. “To Rebel Is Justified? The Image of Zhouxin and Legitimacy of Rebellion in Chinese Political Tradition.” Oriens Extremus 47: 1–24. ———. 2009. Envisioning Eternal Empire: Chinese Political Thought of the Warring States Era. Honolulu: University of Hawai’i Press. ———. 2010. “Political Mythology and Dynastic Legitimacy in the Rong Cheng shi Manuscript.” Bulletin of the School of Oriental and African Studies 73.3: 503–529. ———. 2012. The Everlasting Empire: Traditional Chinese Political Culture and Its Enduring Legacy. Princeton NJ: Princeton University Press. ———. 2013a. “Submerged by Absolute Power: The Ruler’s Predicament in the Han Feizi.” In Dao Companion to the Philosophy of Han Fei, ed. Paul R. Goldin, 67–86. Dordrecht: Springer. –——. 2013b. “Between Merit and Pedigree: Evolution of the Concept of ‘Elevating the Worthy’ in preimperial China.” In The Idea of Political Meritocracy: Confucian Politics in Contemporary Context, eds. Daniel Bell and Li Chenyang, 161–202. Cambridge: Cambridge University Press. ———. 2013c. “From Teachers to Subjects: Ministers Speaking to the Rulers from Yan Ying 晏嬰 to Li Si 李斯.” In Facing the Monarch: Modes of Advice in the Early Chinese Court, ed. Garet Olberding, 69–99. Cambridge, MA: Harvard University Press.

298

Political thought –——. trans. and ed. 2017a. The Book of Lord Shang: Apologetics of State Power in Early China. New York: Columbia University Press. ———. 2017b. “Confucius’s Elitism: The Concepts of junzi and xiaoren revisited.” In Companion to Confucius, ed. Paul R. Goldin, 164–184. Chichester: Wiley-Blackwell. ———. 2017c. “Post-imperial Emperors? Traditional Chinese Political Culture and Its Afterlife.” In: Dynamics of Continuity, Patterns of Change: Between World History and Comparative Historical Sociology: In Memory of Shmuel Noah Eisenstadt, ed. Benjamin Z. Kedar, Adam Klin-Oron and Ilana Friedrich Silber, 94–128. Jerusalem: The Israel Academy of Science and Humanities. ———. 2018. “Chu Identity as Seen from Its Manuscripts: A Reevaluation.” Journal of Chinese History 2.1: 1–26. Pines,Yuri, Lothar von Falkenhausen, Gideon Shelach and Robin D.S.Yates. 2014. “General Introduction: Qin History Revisited.” In Birth of an Empire:The State of Qin revisited, eds.Yuri Pines, Lothar von Falkenhausen, Gideon Shelach and Robin D.S.Yates, 1–36. Berkeley: University of California Press. Puett, Michael J., 2002. To Become a God: Cosmology, Sacrifice, and Self-Divinization in Early China. Cambridge, MA: London: Cambridge University Press. Richter, Matthias L., 2005. Guan ren:Texte der altchinesischen Literatur zur Charakterkunde und Beamtenrekrutierung. (Welten Ostasiens 3). Bern: Peter Lang. ———. 2017. “Roots of Ru 儒 Ethics in shi 士 Status Anxiety.” Journal of the American Oriental Society 137. 3: 449-471. Schaberg, David C. 2001. A Patterned Past: Form and Thought in Early Chinese Historiography. Cambridge, MA: Harvard University Asia Center. Shangjunshu jiaoshu 商君書校疏. 2012. Compiled and annotated by Zhang Jue 張覺. Beijing: Zhishi chanquan chubanshe. Shangshu jin guwen zhushu 尚書今古文注疏. 1998 (1815). Compiled by Sun Xingyan 孫星衍, proof-read by Sheng Dongling 盛冬鈴 and Chen Kang 陳抗. Beijing: Zhonghua shuju. Shelach, Gideon. 2014. “Collapse or Transformation? Anthropological and Archaeological Perspectives on the Fall of Qin.” In Birth of an Empire: The State of Qin Revisited, eds. Yuri Pines, Lothar von Falkenhausen, Gideon Shelach and Robin D.S.Yates, 113–140. Berkeley: University of California Press. Shelach, Gideon and Yuri Pines. 2006. “Secondary State Formation and the Development of Local Identity: Change and Continuity in the State of Qin (770–221 B.C.).” In Archaeology of Asia, ed. Miriam T. Stark, 202–230. Malden, MA: Wiley-Blackwell. Shenzi jijiao jizhu 慎子集校集注. 2013. Collated by Xu Fuhong 許富宏. Beijing: Zhonghua shuju. Smith, Kidder. 2003. “Sima Tan and the Invention of Daoism, ‘Legalism,’ et cetera.” Journal of Asian Studies 62.1: 129–156. Tan, Sor-hoon. 2003. Confucian Democracy: A Deweyan Reconstruction. Albany: State University of New York Press. Tiwald, Justin. 2008. “A Right of Rebellion in the Mengzi?” Dao 7.3: 269–82. Tu,Wei-ming 1993. “The Structure and Function of the Confucian Intellectual in Ancient China.” In Way, Learning and Politics: Essays on the Confucian Intellectual, ed.Tu Wei-ming, 13–28. Albany: State University of New York Press. Vervoorn, Aat. 1990. Men of the Cliffs and Caves: The Development of the Chinese Eremitic Tradition to the End of the Han Dynasty. Hong Kong: Chinese University Press. Wagner, Donald B. 1993. Iron and Steel in Ancient China. Leiden: Brill. Xunzi jijie 荀子集解. 1992. Annotated by Wang Xianqian 王先謙 (1842–1917), ed. Shen Xiaohuan 沈嘯 寰 and Wang Xingxian 王星賢. Beijing: Zhonghua shuju. Yates, Robin D.S. 1999. “Early China.” In War and Society in the Ancient and Medieval Worlds: Asia, the Mediterranean, Europe, and Mesoamerica, eds. Kurt Raaflaub and Nathan Rosenstein, 7–45. Washington, DC: Center for Hellenic Studies, Trustees for Harvard University and Harvard University Press. Yu,Yingshi 余英時. 1987. Shi yu Zhongguo wenhua 士與中國文化. Shanghai: Shanghai renmin chubanshe. Yu,Youqian 虞友謙. 1982. “Fandui junzhu zhuanzhi de sixiang xianqu – Zhuangzi wu jun lun sixiang chu tan” 反對君主專制的思想先驅 – 《莊子》無君論思想初探. Fudan xuebao (shehui kexue ban) 復旦 學報（社會科學版）3: 82–88. Zarrow, Peter. 2012. After Empire: The Conceptual Transformation of the Chinese State, 1885–1924. Stanford: Stanford University Press. Zhu, Fenghan 朱鳳瀚. 1990. Shang Zhou jiazu xingtai yanjiu 商周家族形態研究. Tianjin: Tianjin guji chubanshe. Zhuangzi jinzhu jinyi 莊子今注今譯. 1994. Annotated by Chen Guying 陳鼓應. Beijing: Zhonghua shuju.

299

14 FOOD AND AGRICULTURE

ROEL STERCKXFOOD AND AGRICULTURE

Roel Sterckx

Early China’s fields, waters and forests generated an innumerable range of domesticates and natural assets, including edible crops, livestock and natural resources that could be exploited and processed (wood, hemp, silk, oil and seed crops, indigo, lacquer, etc.). Its people’s relationship to land and grain was central to the emergence of communities, an idea captured in the image of the altars of the soil and grain as symbols for the state. Farming was also a political philosophy since success in managing agrarian society hinged on being able both to generate crops and control the labour force that produced them. But among all the possible benefits that land could bring, the soil was to be tilled first and foremost to produce food: Food is the root of the people, the people are the root of the state, and the state is the root of the ruler. For this reason, the ruler, above, relies on Heaven’s seasons; below, exploits the resources of Earth, and in their midst, deploys the strength of the people. By these means all living things grow to maturity and the five grains flourish in abundance.The ruler instructs his people how to nourish and raise the six domestic animals, plant trees in the proper seasons, work at maintaining the paddy fields and arable fields, start off seedlings and plant mulberry trees and hemp. According to whether the soil is fertile or barren or lies high or low, in each case they cultivate what is appropriate. In hilly and precipitous locations where the five grains do not grow, they plant bamboo and trees. In spring they prune the dry branches. In summer they take the fruits and melons. In autumn they gather legumes and grains. In winter they cut large and small firewood. All these are resources for the people. For this reason the people are not short of supplies while alive, and when dead, their corpses are not abandoned [for lack of a coffin and sacrificial offerings]. (Huainanzi: 9.308) Note how this passage underlines the principle that the populace needs to be taught how to farm by an enlightened ruler or official, that is, agrarian expertise is linked to or at least claimed to hail from the sages, rulers or the political centre.This is a common theme in many texts of the period and it reveals the high stakes held by the state in agrarian life generally. Indeed, the granary de facto was also the treasury. It would be appropriate to describe early Chinese society as a mercantile agrarian society, especially from the Warring States period onwards. Even by imperial 300

Food and agriculture

times officials continued to receive a proportion of their salaries in the form of an allowance of grain. Chinese agrarian society evolved to be highly bureaucratic and commercial in nature, despite musings by philosophers and historians about a past golden age of autarkic agricultural communities. The process of collecting and distributing agricultural surplus and revenue that was initially in the hands of lineages became more administrative in nature over the course of the Zhou period. By the time of the first empires, the picture of Chinese agrarian society was one of nuclear farming households managed by a bureaucratic state that extracted tax and labour via a rural market economy that was already highly monetised. When supervising agricultural activity, the state was not necessarily preoccupied to the same extent with each and every crop. In many ways the balance of agricultural management reflected the composition of people’s diets. Staple starches – mostly cereals and rice – were firmly on the radar of the state as they provided the essential ingredient to keep a population fed and avert famine and social discontent. Complements such as vegetables, meats, spices and condiments did not escape the state’s control either, but their use and availability depended on many other factors such as regional ecology, social and economic status, and the demands of ritual and sacrificial religion. Meat was generally a privilege for elites who could draw on livestock complemented by meats obtained from hunting or at market. The vast majority of commoners however lived on the margins of subsistence and had a simple diet consisting of cereals, beans, mallows and greens. The story of farming life in early China shows little sign of the bucolic bliss or eremitic enjoyment that can be derived from vegetable gardening outside one’s cottage, as intimated in literati poetry of medieval and later times. Likewise, the vast majority of folk were not party to the culinary and gastronomic arrays described in ritual and other texts of the period. Subsistence farmers, even in times of peace and relative political stability, lived a life of chronic hardship. This picture of obedient toil and travail in return for meagre gains is the dominant characterisation of the peasant one finds: “they wear straw rain hats on their heads and clothes of coarse cloth on their bodies. Their bodies become soaked and their feet muddy. Their hair and skin are sunburned, and they exhaust the strength of the four limbs in energetically working the fields. From childhood they become accustomed to it and their hearts are at peace” (Guanzi: 20.401; tr. Rickett 1985–98, vol.1: 325–26).

Sources and scholarly trajectories The study of food and agricultural civilisation in early China has been approached from several angles. Much research has focused on its bio-physical and technological aspects. This includes research on the role of diet, food as medicine, agronomy, ecology, climate and environmental conditions (Amano 1979; Hsu 1980; Bray 1984; Xu Hairong 1999). Another approach has been to understand the cultural-historical dimensions of farming, food production and food consumption (Chang 1977; Anderson 1988; Sterckx 2011; Yue and Tang 2013; Hollmann 2014). Naturally, the technological and socio-political dimensions of food cultivation were closely interlinked. Knowing how to farm had implications for the ways in which one managed a population that largely consisted of people working the land. Success in generating agricultural produce could foment status and power, but failure to secure satisfactory harvests could seal political decline. Technological advances in agriculture could also be a centrifugal force to political power. For instance, introducing iron tools or improving seeding methods results in higher yields and economies of scale, but less labour-intensive farming methods could result in jobless farmers leaving the fields to become a vagrant source for social unrest.The use of bronze or iron advances durability and efficiency but also deprives poor farmers from owning their implements. In short, the history of agriculture and food production is as much a story of social 301

Roel Sterckx

and political developments as it is one of agronomic advances and nutrition. It is best therefore to approach the primary sources with an open mind, as much information can be gauged from different genres of texts, images and artefacts. Let us now turn to the main sources at hand. Aside from the archaeological record, which is most efficiently accessed through reports in the quarterly journal Agricultural Archaeology (Nongye kaogu), the volume of textual and visual evidence that documents the mastery and transmission of agronomic knowledge in early China is relatively small, despite the almost universal claim that it was an agricultural society par excellence. Prior to the late second century ce, one is hard-pressed to find in China topically focused writings on agriculture of the type or scope of Xenophon’s Oikonomikos (The Economist, ca. 362 bce), Cato’s (234–149 bce) De Agricultura (On Agriculture) or Varro’s (116–27 bce) Rerum Rusticarum Libri Tres (On Farming). The age of aristocratic landholding families in China had yet to begin. For much of the twentieth century Marxist historiography treated agriculture almost exclusively as a story of technology and evolving forms of landownership. This, together with the fact that historians of science and technology were the first to take an interest in the study of agriculture in China, has meant that the corpus of primary sources drawn on in scholarly accounts tends to be narrowly defined. The following sources are usually highlighted: a selection of divination inscriptions in Shang oracle bones; some poems preserved in the Book of Odes (Shijing), especially a piece entitled “Seventh Month” (“Qi yue”) that gives a year-round account of agricultural labour; a number of calendrical texts (“Monthly Ordinances”, “The Small Calendar of the Xia”); four agriculturalist chapters in a third-century bce philosophical compendium known as the Spring and Autumn of Master Lü; a number of chapters in a politico-philosophical text known as Master Guan (Guanzi); task descriptions related to agriculture and food handling by officials in the idealised feudal state as laid out in the Rites of Zhou (cf.Wang Xueping 2010); a monthly guide of activities on the farming estate of a mid-level Eastern Han official (Monthly Instructions for the Four Classes of People, cf. Ebrey 1974; Hsu 1980: 215–28); and, finally, fragments of technical information preserved in the sixth century ce Essential Techniques for the Peasantry (Qimin yaoshu) (Xiao Kezhi 2007: 4–14; Xu Ying and Li Changwu 2013). Some entries in Han lexicographical works also contain valuable technical information. To this day, however, the Essential Techniques (completed ca. 540 ce) remains the earliest and most comprehensive agronomic manual and therefore the starting point of most scholarly treatments of agriculture and food. But we must remember that this so-called canon of agronomic literature postdates the oldest songs describing farm labour attested in the Book of Odes by more than fifteen centuries. Unfortunately all but one of the nine titles of works on agronomy listed in the Han imperial bibliography are now lost (Hanshu 30.1742–43; Zeng Xiongsheng 2008: 210–20). The only surviving Han farming manual, the Book of Fan Shengzhi (late first century bce), at a length of approximately 3700 characters, is a relatively short text (Shi Shenghan 1959). Early Chinese writings on agriculture seem as much managerially oriented as they are driven by technological or agronomic concerns. Hence most information on agricultural knowledge is transmitted in normative and prescriptive texts that take the form of calendars, administrative laws and statutes, or ritual instructions. The most detailed transmitted corpus of administrative rules regulating agriculture dates to imperial times. These are the statutes dating to 217 bce uncovered at Shuihudi (Yunmeng county, Hubei province). Regulations are grouped under the headings of “Agriculture”, “Stables and Parks”, and “Granaries” (Hulsewé 1985: 21–46). Legal texts excavated at Zhangjiashan (Jiangling county, Hubei province), dating to 186 bce, indicate that these statutes on agriculture were fairly stable and had not changed much in early Han. These so-called legal texts however are addressed to officials and administrators. We know as yet very little about the dynamics between prescriptive statutes on agriculture and actual farming 302

Food and agriculture

practice. For the historian of agriculture, any statistic on crop yields, food allotments, measurements of surface areas and volume, etc. are to be handled with great caution. For example, the very fact that target quantities of seeds per surface unit are specified in law (e.g. Hulsewé 1985: 41-A27) can indicate that real practice fell short of the norm. Agronomic knowledge tends to be formulated as a menu of instructions or a catalogue of rules that need to be enforced, reinforced or sanctioned.We learn about beginnings and outcomes, that is, what the farmer should or should not do on the one hand and, on the other hand, how the farmer should be rewarded or punished for the outcome of his labour. However, the process in-between, that is, the doing itself, is poorly documented. It also remains uncertain to what extent officials whose main task it was to collect crop revenues and taxes were instrumental in stimulating agronomic innovation. Farmers tend to be more innovating than their paymasters. Just as gardeners today rarely follow instructions on a seed packet to the letter, to many working the land in early China, innovation must also have meant putting aside the rulebook or manual. A great deal of information on agrarian life and the edible domesticates it produces can also be garnered from other texts, including ritual compendia and literary, philosophical, medical and other types of sources. Information on fields and food comes to us through analogies, allegories and metaphors coined by poets and philosophers. The greatest share of information on food production in sources from pre-imperial and early imperial China is concerned with the issue of when to engage in agricultural activities rather than how to farm. The issue of “timing” (sowing and harvesting, selecting the right crops according to the seasons, etc.) is dominant, and, together with this, there is a strong emphasis on matching appropriate soils with crops and location to ensure the efficiency of labour and output. Visualising agricultural life in great detail based on the sources currently available remains difficult. We have impressions in late Warring States and Han mural paintings and on decorated bricks, many from the Sichuan area. These depict farm labour, stables, granaries, ponds, paddy fields, ploughing scenes, etc. (Xia Henglian 1996). Roof end tiles that can be linked to granaries or other farmyard buildings have been found, some containing inscriptions (He Kewei 2009). Other artefacts that help us imagine the architecture and activity on a Han farmyard take the form of miniature terracotta figurines that were buried in tombs, including models of buildings, animal pens, livestock, etc. (Sun Ji 2008: 9–25, 247–50; Guo Qinghua 2010). Interpreting these, both within and beyond the funerary context in which they were found, is not without its problems, but they are informative vignettes of daily life nevertheless. Just as our textual sources are mostly prescriptive, we must assume that much of the visual evidence that has come to us represents idealised or didactic images of farm life.

Climate and ecology Studies of agriculture in recent decades increasingly take note of findings and hypotheses in historical climatology. Estimates on climate variation in early China are arrived at using various methods. These include proxy records delivered by geophysical research, zoo-archaeology and archaeobotany but also the use of historical texts, an approach that was pioneered by meteorologist Zhu Kezhen (1890–1974). While there is no doubt among paleo-climatologists that temperatures changed over the course of the two millennia or so leading up to Qin unification, matching these data to observable changes in agricultural practice and food production as documented in texts remains a tentative exercise. The mining of texts for records on droughts, floods, plagues or frosts can lend useful support to scientific modelling, but there are also methodological pitfalls. The most obvious problem is that our data are drawn from texts with a progeny, dating and context that often remain uncertain. Another problem is that the reporting and 303

Roel Sterckx

explanation of natural disasters or famines, including the terminologies used, were often steeped in political, moral and ideological motives. The Mozi for instance illustrates how elastic terms for a failed harvest could be: “Failure to collect one grain is called dearth, failure of two grains is called scarcity; failure of three grains is called calamity; failure of four grains is called want; and failure to collect all five grains is called famine” (Mozi jiangu, 7.25–26). Finally, charting out the distribution of fauna and flora based on the occurrence of plants and animals in texts is a useful exercise, but it has its limits: the correspondence of nomenclature in literary texts to actual genii of plants and animals can be unstable and hard to corroborate without conclusive archaeological data. On the other hand, cycles of climate change stretch over longer periods than the potential margins for the approximate dating of texts allow for. Zhu Kezhen’s work produced in the 1920s and 1930s continues to be quoted as authoritative (cf. Zhu Kezhen 1979), although scholars have refined some of his findings. Zhu proposed a warm age stretching from around 3000 to 1100 bce, followed by a colder spell that lasted until the mid-eighth century bce. Shang and Western Zhou period temperatures in north China were still around two degrees Celsius higher than today. Temperatures warmed up from the Spring and Autumn period through to Western Han before cooling down again in the Eastern Han (Wang Qizhu 1994: 125–29).The hottest three decades of the Qin-Han period were likely between 210–180 bce, when temperatures for the greater part of the year were around one degree Celsius higher than today (Ge Quansheng 2010: 106). Han average temperatures were slightly lower than in Warring States times, but disagreement exists on the causes for these developments and their regional distribution. On average, then, Eastern Zhou, Qin and Western Han temperatures would be 1.5 to 2 degrees Celsius higher than present-day temperatures. By Eastern Han times they dropped to around 0.7–0.4 degrees Celsius below today’s levels (Wang Yong 2004: 4–5; Wang Yuanlin 2005: 6–10; Wang Zijin 2007: 15–17; Xu Weimin 2011: 15). Southern parts may have been less susceptible to temperature variations, but rainfall extended further north and over longer periods. Higher levels of precipitation, together with milder temperatures, suggest that during most of the Bronze Age and into the Han, growing seasons therefore may have been slightly longer than today.

Diets, place, time Archaeologists increasingly concur that the transition from hunting-gathering to agriculture (in prehistoric China some 10.000 years ago) represented a process rather than an abrupt shift. It is now also accepted that, rather than being mutually exclusive modes of existence, nomadic ways of life and sedentism more often complemented each other. The shift from the use of perennial types of cereals to annual crops was gradual. It is important to keep in mind therefore that, despite transitions over time and significant regional differences, and despite increasing sophistication in soil modification, plant cultivation, animal domestication and the use of tools, communities in early China rarely fed themselves by one means only. Most early Chinese were farmer-hunter-gatherers who would complement their lot by means of gift exchanges and, later on, transactions of a more or less commercial nature. The primary domesticated animals in Neolithic times were the pig, dog, water buffalo and domestic chicken. Hunted wild animals, fish, shellfish and wild plants formed a major part of the diet, and this continues into historical times. Some of the earliest millets in the world were discovered at sites in north China, foxtail (Setaria italica) and broomcorn (Panicum miliaceum) being the main species. Among the best-researched millet farming settlements are the Yangshao culture (5000–3000 bce) in the Yellow River valley and Dawenkou settlements (5000–3000 bce) in Shandong. In the Yangtze region, rice became the most important staple food from the 304

Food and agriculture

fifth millennium bce onward. In south China wild animals (bovines and deer) complemented a diet that included various tubers (yam, taro, lotus) and aquatic plants (Flad and Chen 2013: 185–87; Shelach-Lavi 2015: 118–21; Chang 1999: 42–54; Luo Yunbing 2012; Xu Wangsheng 2009; Liu Li and Xingcan Chen 2012: chapter 4). The distinctive nature of plants, animals and diets in the northern plateaus and central plains versus those in the south in pre-agricultural times will come to define the two main agricultural traditions that would develop in historical times: the dryland cultivation of cereals (millets, barley, wheat) in the (loess-based) north and wetland rice agriculture in the south (the Huai river being the boundary). Up until roughly the tenth century ce, the north would remain the agriculturally most productive region. Whilst the north-south divide is the most articulate in terms of crops and farming methods, increasingly scholars are now studying micro-regions within these areas (Wang Yong 2004; Zhu Hongbin 2010; Han Maoli 2012: vol. 3). The late Shang was a largely agricultural society, with millet as its staple. Shang elites would have offered up in sacrifice and consumed substantial quantities of meat, including game. The Shang also produced ales from cooked grain and rice for human consumption and sacrificial purposes. Oracle-bone divinations show the Shang king to be very inquisitive about the fate of the harvest and crops across his royal domains. He conducted regular inspections, commanded communities to engage in planting and requested participation in harvest-seeking rituals. Hunting and farming remained closely linked: in oracle-bone records the graph for “field” or “taking to the fields” (tian) also means hunting to open up new lands and clear fields from wild animals to protect the crops. The Shang also cultivated mulberries and practiced sericulture (Keightley 1999: 277–81; Keightley 2012: 161–66). The Shang royal house set up breeding grounds on fertile and safe lands within its domain as well as on the periphery and edges of its polity. “Herdsmen” (mu) were officials of considerable status, possibly even related to the royal clan. Their knowledge of local terrain also made them useful as guides in military campaigns (Cai 2009). Much larger areas of China were forested at the time, ranging from deciduous hardwood and broadleaf forests (elm, oak, maple) in the north to semitropical forests in the Yangtze area (and evergreens in the tropical south). Scholars estimate that the first significant environmental pressures of expanding agriculture by the clearing of forests became evident around 1000 bce. This is when a number of animal species (such as the elephant and tiger) start disappearing from the zoo-archaeological record for the north China plain (which, by the mid-Han, had been deforested). The idea that much of farming activity at the time was centred on clearing forests transpires in the following Zhou hymn: “They clear away the grass and trees, their ploughs open up the ground; in a thousand pairs they tug at weeds and roots, along the low grounds, along the ridges” (Waley 1996: 304). Increased reclamation of land at the expense of wooded areas, pastures and marshes in the centuries following the Shang meant that staple diets became increasingly crop based. The soya bean appears late in the record during the late Western Zhou period (Liu and Chen 2012: 85–91). Millet (setaria) continued to be a main staple, and the Zhou designated Lord Millet (Hou Ji) as their founding ancestor (Waley 1996: 244–47; Cao Shujie 2006). Rice was reserved for the elites. In texts of the period, the term “the five cereals” (wu gu) denominates all grains in general (including the two main millets, rice, wheat and barley, and legumes, or sometimes hemp and beans). A rich array of vessels was designed to boil and steam cereals. Meats were roasted, baked, broiled or boiled. By Eastern Zhou times pulses became a significant part of the everyday diet. Several types of bean, lentils, peas and seeds were common in the diet of north China. The only preserved agronomic treatise dateable to the Han notes that the ancients were aware of the usefulness of soya crops, as they produce a harvest in adverse years and could be grown as a provision against famine (Shi Shenghan 1959: 4.6). The soya bean was used in its 305

Roel Sterckx

fermented form in sauces, relishes or, later on, as bean curd. According to some scholars a scene on an engraved mural in an Eastern Han tomb at Dahuting (Mixian county, Henan) represents the production process of a prototype of doufu (Huang 2000: 302–16). Others argue it depicts wine-making (Sun Ji 1998). Warring States and Han farming was mainly crop-focused. The overall trend was one of turning cleared lands into fields for intensive farming. Surface areas available for pasture were limited, and intensive cropping on permanent fields could produce yearlong yields. Beyond animals kept on farmyards for domestic use (poultry, ducks, fish, pigs), livestock farming took a minor share. Some roof end tiles recovered from the remains of the Ganquan palace (in presentday Chunhua, Shaanxi province) contain formulas seeking the fecundity of the six domestic animals (Fu Jiayi 2002: nos. 732–34). We cannot infer however that a mixed farming model did not develop at all. Jia Sixie’s Essential Techniques contains several chapters on livestock breeding, including some that are attributed to figures known in pre-imperial times. By Western Han times, the state had a definite interest in managing stock-breeding. The management of pasture lands was tightly regulated. Strict accountancy rules applied to government herds, including the use of passports for horses. Horse branding seals were used to indicate ownership (Sun Weizu 2010: 134 fig.14). Punishable offenses ranged from injuring livestock, to being unable to prevent cattle or horses from trespassing on crops, to misreporting cadavers or animal parts (Cao Lüning 2002: 143–56; Ding 2009: 99–100). Shepherds were no free-roaming agents, and the Han statutes stipulated that they needed a tally that authorised exit or entry to the fords and passes at the frontier. Owners of horse and cattle herds were not allowed to exit the border at all (Er nian lüling yu Zouxian shu: 305). Even with farming mostly crop-focused, the surface area of cultivable soil in early China was relatively small (in the People’s Republic today, less than 12 percent of land is arable). Just over a quarter of the total area of demarcated land accounted for in a census held in 1–2 ce was specified as actual or potentially arable land (Hanshu: 28B.1640). More research is needed to establish how non-arable environments that supplied food, such as rivers, parks, coastal areas and forests, were managed. Judging by the volume of early imperial prohibitions and regulations that govern access to forests, parks and rivers, it is clear that no amount of attention devoted to plant-tending or animal husbandry deterred people from foraging for small game, acorns, nuts, berries and fruits. For instance the Yunmeng marshes in the old state of Chu, which covered a large lake area in the middle reaches of the Yangtze along the confluence with the Han River, produced large quantities of fish, game and timber (Ma Biao 2013). Parks and enclosures could cover large swathes of land.The Shanglin Park occupied an area of around 100 square kilometres south and southwest of Chang’an. For rulers to open up parks and pasture lands to the people in times of famine was a way to show largesse. The Chinese diet reflected the importance of crop-farming and was therefore less carnivorous than that in Europe. Staples were complimented with a variety of greens and fruits including cabbages, radishes, mallows, gourds, mustard greens, bamboo shoots, plums, peaches, pears, apricots and jujube. Foodstuffs were conserved and preserved by drying, salting and the use of ice (Sterckx 2011: 14–19). By Han times the diet of elites had become more varied. For some, if ritual records are to be believed, it included a wide range of exotica (Du Qingyu 2010: 124–42). Noodles first occur in the written record around 100 ce, but it is likely that they were consumed much earlier. A bowl with remnants of millet noodles has been uncovered from a Late Neolithic site in northwestern China (Serventi and Sabban 2002: 276–83, 294–95; Lu Houyuan et al. 2005).The emergence of floured wheat products such as dumplings and dough-based pancakes reflected an increasing Central Asian influence. Dairy products did not belong to the mainstay diet, and reference to fermented or soured milk almost invariably 306

Food and agriculture

links its consumption with pastoral nomads (e.g. Hanshu 96B.3903). Tea may have served as a medicinal drink in southwest China before the Han, and reference to brewing tea is made in a mock slave contract (ca. 58 bce) (Hsu 1980: 231–34). However, for more reliable evidence of tea drinking we must wait until early medieval times (Benn 2015: 21–41). The earliest mention of sugarcane syrup as a condiment is in one of the poems in the Songs of the South (Chuci jin zhu: 233). Prior to the use of cane sugar, malt sugar and honey were known as sweeteners, although firm evidence for bee-keeping does not occur before the second century ce (Daniels 1996: 55–9, 67–71; Pattinson 2012). Hands and spoons had for centuries been the primary feeding tools, and although the philosopher Han Fei (ca. 280–ca. 233 bce) claims that the last king of the Shang used a pair of ivory chopsticks, none have been archaeologically recovered from Shang sites so far (Han Feizi jishi: 7.438). Some Bronze Age sites however contained sticks which archaeologists believe were used as feeding sticks. By the late Warring States period chopsticks were commonly used (Ōta 2001: 1–19; Liu Yun 2006: 70–152; Wang 2015: 16–26, 41–54). Han tombs contain a rich array of miniature models of the stove, kitchen and cooking equipment (including barbecue grills). The most intact post-mortem dining room excavated to date is a compartment of Mawangdui tomb no. 1 in Changsha (Hunan province). It contained dozens of bamboo boxes, pottery vessels with foodstuffs, and lacquerware table utensils (Pirazolli-t’Serstevens 1991). The biological expansion of empire that followed the establishment of a corridor to Central Asia and trade routes out of its southernmost parts resulted in an expansion of the Han’s dietary range and its agricultural and horticultural horizons. Grapes, pomegranates, and alfalfa entered across the trade routes alongside the Taklamakan desert (a segment of the corridor also referred to as the Silk Road). Emperor Wu (reigned 141–87 bce) attempted to grow these new plants in soil around his palaces (Shiji: 123.3173–74). In the south, one commandery (Ba in Sichuan) established two agencies to deal with fruit orchards (Hanshu: 28A.1603). The quest for exotica also posed a burden on the coffers. In 32 bce Emperor Yuan agreed that the imperial greenhouse in which out-of-season vegetables were farmed (including scallions and leeks) was to be closed down, resulting in an annual saving of “tens of millions of coins” (Hanshu 89: 3642–43).The idea of home-grown produce was even criticised by some on the grounds that it deprived farmers from economic opportunity. Sima Qian (ca. 145–ca. 86 bce) tells the story of an erudite from the state of Lu who, upon tasting home-grown vegetables, found them so tasty that he immediately pulled up all vegetables in his garden. He also threw out all fine cloth made by his maids and burned their looms. If one grows vegetables and weaves cloth at home, “how would farmers and weaving girls have a place to sell their goods?” (Shiji: 119.3102).

Agronomic developments Agronomy – that is, the technical understanding of the science behind the growth of crops, husbandry and the management of their environments – evolved significantly over the period. While the transition from slash-and-burn agriculture to fallow-rotation systems and the continuous tilling of fertilised fields did not develop in clear-cut stages, tillage techniques generally focused on improving yields per surface area through more intensive cultivation methods. The Shang used wooden or stone tools and implements to clear the land, turn over or break open soil. Bronze was used for some types of tools such as spades, picks, axes, knives, chisels and sickles. Wooden digging sticks and spades made of stone, jaw or shoulder blades would continue to outweigh the use of bronze implements in the Western Zhou. Bronze was too precious, prestigious, sacred or costly to be used for agricultural tools. Only industrialised iron production in Warring States times would turn metal into a significant tool (von Falkenhausen 2006: 409–10). 307

Roel Sterckx

Innovation in agriculture centred on moving away from field-rotation and fallowing to intensive multi-crop rotation and annual or year-round use of the land. This included the use of organic fertiliser, advanced irrigation and water storage, improved seeding methods and more efficient techniques to break up soil and weed in between sprouting crops. The Shang already used fertilisers to improve settled farmlands. Both animal waste and human night soil were used as manure (Yang Shengnan and Ma Jifan 2010: 153–5). Applying fertilisers to improve soil quality was common by Warring States times, including the use of green waste and ashes (Hsu: 6–7; Li Yaguang 2009: 115–17, 139). The science of fertilising soil using various concoctions of animal waste became more sophisticated by early imperial times. Records mention manure adapted according to soil type as well as a technique whereby one dips seeds in liquid manure to repel insects and improve growth (Sterckx 2002: 107). Several Han murals depict farmers collecting dung and urine behind their horses, a practice possibly first mentioned in an analogy in the Zhuangzi (Zhuangzi jishi: 4.168). Replica earthenware latrines constructed above or annexed to a pig sty or pen appear frequently in tombs of the late Western and Eastern Han period (e.g. Henan bowuguan 2002: plates 62–72, 124–156). There has been some discussion as to whether these human-pig privies were designed in order to accumulate all manure together or whether they were meant to facilitate pigs to feed on human night soil (Wang Qizhu 1994, vol.1: 502–6). The Zhou started to exploit animal traction more functionally. Ox-drawn wooden and stone ploughs increased the surface areas that could be cultivated and freed up manpower to open up new land. As a result, wheat and barley increasingly supplemented millet as the staple.When iron became widely available around the late sixth century bce, the introduction of cast or wrought iron tools became the main propellant for advances in agricultural production (Li Yaguang 2009: 2–22; Zhou Xin 2010: 192–204, 276–304). At the same time, knowledge of soil types and soil modification techniques developed. The first detailed taxonomies of soil appear in the “Tribute of Yu” chapter in the Book of Documents. By the third century bce soil and plant ecology was described in detail according to terrain: plains, hills, mountains (Hsu 1980: 91–96; Zeng Xiongsheng 2008: 84–90, 142–47). Prior to the advent of large-scale irrigation in the late Spring and Autumn period, water was mostly sourced from wells, ponds, reservoirs and rivers. In the multi-state world that emerged, several figures are associated with advances in irrigation technology. In the state of Wei, Li Kui (fifth century bce) championed improvements in water conservation and irrigation. In Qin, Zheng Guo started work in 246 bce on a canal to supply the alkaline loess plateau north of the Wei River, irrigating 450.000 acres. Qin also constructed the Dujiangyan dam system, still functional today, to divert the Min River and significantly increase rice yields in the Sichuan basin (Tan Xumin 2004). Western Han rulers invested massively in water conservancy projects and canals to provide the capital area of Chang’an with water and grain supplies. Han Wudi’s irrigation projects just north of the Wei River alone created a network of man-made canals that watered over 30.000 hectares of land (Nylan 2015: 103–6). In the south, on a smaller scale, households also engaged in aquatic agriculture. Murals from Sichuan depict ponds for raising fish, lotus plants, ducks and turtles. For northern climes, the Book of Fan Shengzhi describes how to use snow to the best possible effect. Snow kills insects and preserves moisture for the soil. Rolling down snow prevents it from drifting away on the wind. Snow was stored in vessels and buried underground, and used to soak excrement for fertiliser. To protect sprouting crops from snow and frost injury, the manual advises, dew should be scraped off overnight until sunrise has set in (Shi Shenghan 1959: 4.1.3, 4.2.2). In addition to scaled-up irrigation projects, other innovations during the Western Han greatly enhanced productivity. Most important was the use of a metal plough with two shares 308

Food and agriculture

pulled by two oxen. This enabled a new seeding method known as the “alternating fields system”, attributed to Zhao Guo (ca. 100 bce) (Hanshu: 24A: 1138; Swann 1950: 184–91). Instead of planting seeds on wide ridges between furrows, seeds were now planted inside the furrow. In the process of weeding, earth would fall down into the furrow, strengthening the root system of the newly sprouting crops so that by midsummer ridges and furrows would be level. The next year, the positions of the furrows and ridges would be “alternated” to preserve nutrients in the soil and reduce the need for fertiliser. Ploughing scenes are depicted frequently on pictorial stones and mural paintings of the Han (Zhou Xin 2005: 333–45; Hu Zexue 2006: 75–84). However, given the cost of iron tools and draught cattle, hand-worked tools continued to be used. Another improved seeding method was to sow seeds in shallow pits. This improved the retention of moisture and fertiliser around the plants and could be practised on sloping ground (Shi Shenghan 1959: 7.1).Wetland rice agriculture in the south is referred to in several sources as “ploughing with fire and weeding with water”.This method consisted of burning off weeds and flooding the fields in which rice seeds are directly sown. A second watering during the period of growth was aimed at getting rid of weeds. Reference to rice transplantation is first recorded for north China towards the end of the Eastern Han, from where the practice spread to central and southern areas (Shiji: 30.1437; Hanshu: 6.182; Nishijima 1986: 568–74). Miniature terracotta paddy fields and ponds have been uncovered from several Han sites in the Sichuan area (Liu Wenjie 1983; Zhang Feng 2009). Other technological improvements in Han included the use of pedal-operated hammers to pound grain and improvements to wells and pipes (Zhou Xin 2010: 332–37, 352–62). Together with increasingly intensive cultivation, grain storage facilities improved. Prior to the use of above-ground granaries, grain was kept in storage pits. Large numbers of terracotta model granaries have been found in Han tombs, some containing labels of the food supplies kept within (mostly foxtail millet, wheat, barley and beans). These miniature granaries and roofed depots reveal interesting architectural and technical features. Granaries were round or square, raised up from ground level, and supplied with one or more airing vents. Some were door-less or had bamboo latticed windows to prevent vermin and sparrows from entering. Legal statutes contain precise requirements for grain storage: the door-bar was to be tight so that nobody unauthorised would be able to steal or scoop grain; when cleared the granary had to be swept empty, rat holes were to be prevented, insects had to be cleared, chickens were to be reared away from the granaries, etc. (Cai Wanjin 1996). In imperial times, grain was stored in huge quantities. Over twenty granary complexes are attested in or around the Western Han capital Chang’an alone (Nylan 2015: 110–113). State granaries were erected in heavily guarded walled complexes. The granary was, as some texts intimate, like the stomach of the empire. Between 57 and 44 bce, designated granaries known as “ever level granaries” were set up to stabilise the price of grain. The principle was that in times when grain was cheap it could be bought up at an increased price by the government to benefit the farmers; conversely, in times of high grain prices, the government would release grain from storage at reduced prices to benefit the people (Liu Jiapeng 2010: 87–94). Knowledge of how to fight agricultural pests became increasingly important as large segments of the population became dependent on centrally stored harvests. In addition to droughts and floods, locusts are among the most damaging agents on record. Insect plagues and the surface area they affected were to be meticulously reported, and cash could be on offer in return for catching locusts (Hulsewé 1985: 21-A1; Hanshu: 12.353, 99B.4176). Mole crickets could damage dikes, and the fight against agricultural plagues is mentioned in a wide variety of texts from poetry to calendars (Xiao Kezhi 2007: 173ff; Zhang Yihe 2007: 15–24; Bu Fengxian 2006: 59ff.; Duan Wei 2008: 39–41; tables 267–350). 309

Roel Sterckx

Silviculture Technical information on silviculture beyond comments on the growing of mulberry for silk is sparse. But it is clear from texts such as Master Guan that the value of wood as a medium- to long-term return was well acknowledged (wood taking ten years on average to become timber, according to the same text; Guanzi: 3.55). By Han times, when large areas had been deforested and demand for wood for building went up, there is more reference to it. In economic discussions there is an acknowledgement that, unlike wood, cereal agriculture ultimately serves the short term: one produces and consumes over one season, stores it perhaps for another, but that is as long as one can project benefits from it. Sima Qian captures the idea in a proverb: “You don’t go one hundred miles to peddle firewood; you don’t go one thousand miles to deal in grain. If you are going to be resident in a place for one year, then seed it with grain. If you are going to be there for ten years, plant trees. And if you are going to be there one hundred years, provide for the future by means of virtue” (Shiji: 129.3271–72). Managed woodlands included forests that produced timber woods, bamboo groves and lacquer trees. Campaigns to plant up areas of land or set up parks are documented for the Qin and Han when large areas in the north and central part of the empire had been cleared of their natural vegetation and woodlands. The notorious highways built by Qin Shihuang were flanked by pine trees.The Wei River valley was dotted with plantations of bamboo. They yielded a higher profit than agricultural crops, notes Sima Qian, suggesting that bamboo was grown in this region for timber (Shiji: 129.3272). The most popular fruitless trees grown in Qin and Han times included the elm, locust tree (sophora japonica), white poplar and willow. Elm was the more exploitable, fit to be used in buildings after five years and ready to be made into wooden implements after ten years. The value of one 15-year-old elm tree equalled that of nine bales of unworked silk. Fruit trees included the date or jujube tree, tangerine, peach, chestnut and plum (Yu Kunqi 2012: 75–76; Wang Zijin 2007 18–22, 333–44; Hara 2013). Woodland management and other tasks in the forest economy were supervised by officials. Taboos on felling trees were part of religious belief. “Daybooks” (calendars) record divinations for cutting down specific tree species and list the unpleasant consequences that could result from chopping down the wrong type of tree on the wrong day (Liu Tseng-kuei 2009: 905–6).

The farmer and peasant life The social history of farm life in early China still remains to be written. It is a difficult task since most of our sources have little to say about the mass of nameless and archaeologically invisible peasants that made up society. The story of the peasant in early China is told by actors who may have never tilled the soil. Already in Shang divination records, the farmer, as opposed to those serving as soldiers, is notably absent. David Keightley notes: “There was much concern with the harvest, much of it involved the king himself, but not generally for the peasants who carried it out” (Keightley 2012: 136). Discussions of the farmer’s lot stretched across the domains of the political, economic and technological. As much as fields were there to generate produce, the agricultural economy was also a political economy: fields tilled laboriously also shaped the character and psychology of those who worked the land. Those who govern, one sixthcentury bce statesman claimed, should be like those engaged in agricultural labour, by which he meant they should be relentlessly dedicated to the task from dawn to dusk without overstepping their responsibilities (Zuozhuan, Xiang 25: 1108). The peasant was to be obedient, amenable to instruction and tied to his plot of land.The writings of the Warring States masters of philosophy deal with questions such as how farmers should behave, what place farming should occupy 310

Food and agriculture

among other professions and, most importantly, how rulers should control a farming population and ensure political stability.Whilst every philosopher shares views on the imperative of the state to produce food and organise farm labour, the worthies and sages themselves, however, are rarely seen to put their own hands to the plough (Sterckx 2015a). Rhetorically, farming was granted pride of place among the professions. By Qin and Han times the court self-consciously started to extol farmers over merchants as the preferred constituents of the economically productive state. The empire was to be agrarian, and any activity that pulled people away from the fields or allowed some to accumulate land was a concern for the court.This outward ideological emphasis on the importance of farming, and the discriminatory measures the Han court took to control the power of merchants, originated from the single most important variable that determined economic life in early China: the relationship of people to land. The most defining factor in the peasant’s work and livelihood was his relationship to land, which in turn determined his economic, social and political agency. The history of landownership and the concept of property in early China remain highly contentious and problematic. Marxist historiography has been influential in depicting the transition of the farmer’s fate from the Spring and Autumn period to early Qin as nothing short of a social revolution in which the toiling peasant threw off the shackles of feudal serfdom to become the head of a free and landowning peasant household. It is true that, over the course of the Eastern Zhou period and certainly as populations grew, land became economically more valuable than labour. But the transition from a society run by landholding aristocratic lineages that drew on peasants as vassals to one in which small farming households increasingly carved out agency over a plot of land they held in usufruct, long-term tenancy or, possibly, in freehold is a story as yet marked by a lot of unknowns. Reference to “public” and “private” landholding in the sources ought to be read with caution. It is probably best to conceive of “public” fields throughout the period as lands that were somehow managed directly or by proxy by an overlord, the court or the bureaucratic state. “Private” land then was worked and managed by tenant farmers or more or less independent farming households but usually still under the very tight supervision of the state. We have as yet very little evidence of land changing ownership through contractual purchase prior to Qin. Even during Han times farmers could “own” land by various means. Once the “private” purchase of land for agricultural purposes had been established in late Qin and early Han, all non-agricultural natural resources (mountains, marshes, forests, waterways, parks, etc.) were considered to be the emperor’s private property. They were managed by the privy treasurer and therefore not an automatic source of income for the grand minister of agriculture, who was the government’s treasurer. The most idealised system of land use associated with the Zhou is that of the so-called wellfield system, described by Mencius (fourth century bce) (Mengzi: 10.348–63). This model envisions a collective of eight farming households distributed on a rectangular grid that copies the Chinese graph for a well (jing). Each household gets usufruct of one hundred mu of farmland and some extra space on which to rear animals, grow mulberry trees, and put up a cottage. The overlord then draws income from a ninth plot jointly worked by all eight households. In Han times the well-field system was often invoked as the symbol of an ideal agrarian past. This past had been swept aside, it was claimed, when Shang Yang (around 350 bce) in Qin instituted land reforms that abolished the old fixed land tenure system with a grid of pathways dividing land into blocks and managed directly by farming households. But given that there is little or no evidence of the well-field system in other pre-imperial sources, we can assume that this ideal of the farming commune was largely invented by Mencius (Cao Yuying 2005). After the decline of the old aristocratic lineages and before the (re-)emergence of great landowning families in the Eastern Han, the agrarian economy of early China was predominantly 311

Roel Sterckx

based on production by nuclear farming households that intensively tilled relatively small plots. Both technology and the ecological environment were conducive to small-scale farming. Households might organise themselves into cooperative units for larger works such as the maintenance of dykes and irrigation channels, but the state envisioned the smallholding household rather than the landed estate as the basic labour unit. Among kinsmen belonging to wealthier families, the inheritance system that subdivided land and real estate among all sons also limited the size to which individual household farms could grow. Farming did not necessarily imply living a life in rural remoteness and cut off from towns. By Western Han times peasants did not so much live in isolated farms dotted across the land but in walled residential areas. Such rural villages would consist of around one hundred families whose relationships were governed by a system of ranks. Each family owned a small plot of land. Theoretically, nothing prevented a farmer from aspiring to an official career, but in practice a system of recommendations meant that upward social mobility was the exception. One government measure to promote the profile of farmers was the introduction by Emperor Hui in 191 bce of a title for model farmers, who were to be known as “diligent tillers” (li tian).These locally recommended farmers would be granted exemption from taxes and corvée and receive imperial gifts.Yet the establishment of an honour of rank for farming achievement did little to improve the lot of the farming population as a whole (Hsu 1980: 25; Huang Fucheng 2006; Wang Yong 2008: 20). Although most texts speak of neatly divided gender roles with men tilling the fields and women engaged in sericulture, this picture is prescriptive or ideological at best. In practice, in the small-scale farming household both men and women would be toiling over the same chores. Work patterns were dictated by the seasons rather than gender (Yu Kunqi 2012: 36–39). The intensifying volume of discussion devoted to the fate of the farmer in Western Han sources reflects increasing concern about the need to feed a growing population. In the census of 2 ce the Han Empire counted just over twelve million registered households, making up a total population just short of sixty million. Statistics on agricultural output and consumption in the Han farming household need to be handled with caution, as figures vary from source to source and depend on time and locality. But it is possible to arrive at a rough profile.The average farming household would consist of four to five members (father, mother, two to three children, one grandparent), out of which two to three would be economically active (for the purpose of registration and taxation, adulthood started at 15). On average throughout the Han a small farming household had at its disposal around 50 mu of land, one mu equalling around 0.11 acre. This corresponds to around 14 mu of land per head. Average grain yields were between 2.5 and 4 shi per mu (one shi equalling around 20 litres). An adult male would consume around 3 bushels of grain per month; a family would require just over 11 bushels per month, or around 140 bushels per year. Sixty percent of a household’s revenue from farming and side occupations would be required to feed and clothe the family; 12 percent was due in the form of taxes (land, property, poll tax), leaving the remainder to fund special occasions such as weddings and funerals, fulfil obligations in community sacrifices and festivals, care for orphans or cover for illness. In reality many families had less land at their disposal or saw their revenue siphoned off by landowners. Fluctuations in weather and variations in annual yields meant that the farming household was more likely to run a chronic deficit rather than surplus (Hsu 1980: 67–80;Yu Kunqi 2012: 130–135, 147–48). The biggest draw on the farmer’s energy beyond his own labour and the taxes that were levied from him in cash and kind was statutory corvée labour. This periodic requisition of labour in the service of public works and large-scale building and irrigation projects goes far back, possibly even to Shang times, but it became increasingly regulated and administratively organised during the late Warring States and early empires. When the state was in control and able to 312

Food and agriculture

finance flood control and irrigation, it could draw on registered households for labour. As soon as the state’s grip weakened and powerful families stepped in to fund and manage water and roads, its taxable cohort of farmers greatly reduced. Farmers could also be summoned to war. The discipline and social bonds forged on the fields made men well prepared for the battlefield, and some passages metaphorically compare farming tools to weapons: Repair agricultural implements as though they were military weapons. Conduct agricultural work as though it were making war. Push and pull grubbing tools or hoes as though they were swords and halberds. Don rain capes as though they were coats or armour. Weave straw to make rain hats as though they were shields. This is because once agricultural implements are fully cared for, military weapons will then be in plentiful supply, and once one has become thoroughly familiar with agricultural matters, one will become skilful in the construct of war. (Guanzi 53: 1016–17; Rickett 1985–98, vol.2: 220) Given that the farming household was relatively small, it was particularly fragile when labour was requisitioned by the state. Corvée labour duty was one month per year, but military duties could last much longer. Registered farming households were the best guarantee for the state to secure and account for adequate revenue and control the movement of people and goods.The Han even pushed this agrarian model out to the pasture lands on the northwestern edges of the empire, where it set up state-sponsored agricultural colonies populated by soldiers and conscripts (Hanshu: 96A.3873). While ideologically it was claimed that all professions should be separate and stringent measures were put in place to control mercantile activity, in practice farming households engaged in small-scale commercial activities that are part of the rural economy.The slave boy in the contract mentioned earlier is busy buying and selling goods at rural markets and is told to display good manners when peddling his goods. The fate of farming households became more precarious in Eastern Han, when increasingly indebted farmers were forced to sell their land and services to emerging great families in return for patronage and social security. The combination of a high poll tax and low tax on land encouraged land sales. The wealthy accumulated land and the small-scale farmer had to abandon or rent his plot, or sell his labour to large estate holders. We do not have sufficient data to establish the share of hired labour and tenancy as it evolved during Han, but one estimate puts it at one-fifth of the population under Han Wudi (Hsu 1980: 66). An attempt by Wang Mang (reigned 9–23 ce) to bring all land back under state control was short-lived (Hanshu: 99B.4110–11). As the Han went on, the peasant increasingly became associated with a discourse on dispossession, displacement, poverty and the malcontent.

Food, farming, religion To no great surprise, discussions of food consumption and dietary choices often focused on the distinctions between the rich and the deprived. Access to food and the sparing or overindulgent use of food supplies was a prominent marker for moral judgement. Mencius condemns rulers who allow their dogs and pigs to feed on food destined for humans or keep their granaries full in times of famine and natural calamity (Mengzi: 2.59, 5.158). Food thus served as a marker of identity in several ways. Dietary habits distinguished the familiar from the non-familiar, and texts produced by the Central States often simplistically divided the world into a civilised and sedentary heartland of farming folk who eat grain and cooked meats versus a barbarian periphery 313

Roel Sterckx

where people who follow their animals have a diet devoid of cereals and where food is consumed raw (Sterckx 2011: 20–1). Diet marked rank and hierarchy, both among the living and among the dead who received food offerings. Diet also marked one’s moral character. Competent rulers, it was claimed, were able to hold the mean between fasting and feasting or were willing to make economies at court and in their personal regime. The distribution of meat and ale to the infirm, weak and elderly demonstrated compassion by the powerful. The use of alcohol was particularly prone to moral commentary. Alcohol was a dubious substance. On the one hand, rule-guided drinking was a requirement in social life and a key element in sacrificial ritual. On the other hand, ale invited excess. Our texts judge rulers, officials and their regimes by the way they handled alcohol and ritualised drinking: debauched drinkers such as the late Shang kings were pitched against exemplary rulers or sages, such as the Confucius figure, who were praised as masters in the art of moderate consumption (Sterckx 2015b). The art of cooking also provided an endless trove of metaphors to describe the art of rulership: butchers are exemplary craftsmen because of their sense of proportion; cooks and stewards figure as political advisers, and the art of blending ingredients into a balanced stew evoked the idea of harmony and social concord (Sterckx 2011: 65–76). Feeding the ancestors and the spirit world by means of sacrificial food offerings was at the heart of religious practice. For those producing food, the blessing of the spirit world was equally important.The farming calendar was punctuated by ritual and religious observances. Besides the Zhou founding ancestor Lord Millet, other tutelary spirits and deities were worshipped to watch out over crops and livestock. As much as the farmyard and the fields were a theatre of hard labour, they were also a numinous landscape inhabited by spirits and baleful powers that could influence the crops, weather and the peasants’ health and fortune. The Book of Odes mentions sacrifices offered to the “Ancestor of the Fields”. Other incantations invoke spirits to protect the health and fertility of pasture animals or keep vermin and wild animals away from the crops and yard (Sterckx 1996). Dikes and irrigation channels that offered flood protection were also addressed in prayer: “May the soil return to its dwelling place. May the water return to its riverbed. May insects and vermin not arise, and grasses and woods remain in the fens” (Liji jijie, 25.696). The Spirit of the Soil (she) who in Zhou times, together with the spirit of the grain, received blood sacrifices (to seal political power through land entitlement) evolved into a focal deity for local communities. Religious observances related to the mulberry and silkworm harvest were held in farmsteads and households long before the state turned such practices into an annual state ritual in Han times. One tradition identified the wife of the legendary Yellow Emperor as the First Sericulturalist (Kuhn 1988: 248–57). But the most important patron deity of agriculture was Shennong, the First or Divine Tiller. In Warring States and Han iconography, Shennong is depicted working the fields pushing a forked digging stick. In early Chinese literature, the figure of Shennong is associated with the untainted original stages of civilisation, an era unspoiled by violence, military rule and human artifice. As the legendary inventor of agriculture, he appears as a patron deity of farmers, protector of cereal crops, a proto-pharmacist who distinguished edible from poisonous plants, and a patron sage embodying the philosophical ideal of hermitage, reclusion and the autarkic agrarian society (Graham 1979). Shennong was the recipient of cult at state level. An annual ploughing ceremony, well studied by Derk Bodde (Bodde 1975), was held in his honour at the outset of the agricultural season. Sacrifices to Shennong and cognate spirits were also conducted at the local level. Recent archaeological finds have given us more information about the cult surrounding First Tiller in the localities. A large cache of Qin bamboo and wooden documents found in well J1 at the town of Liye (Longshan district, Hunan province; dated between 222 and 208 ce) contained storage records for sacrificial offerings reserved for 314

Food and agriculture

the cult of First Tiller. Around two dozen slips so far have been identified that relate to it. They record the offerings to be made (including salt, male goats and millet) by the granary manager and his assistant as well as the distribution of the offerings after the sacrifice (Zhang Chunlong 2007). In a ritual prayer from another tomb in Hubei, First Tiller is called upon during a ritual conducted at the granary before the village farmers set out the seeds (Zhoujiatai tomb no. 30, dated ca. 209–206 bce; cf. Hubei sheng Jingzhou shi 2001: 51–2, 132). It is clear then that Shennong’s association was primarily with grain agriculture, as on several occasions he is said to have invented agriculture because animal domesticates did not suffice to feed the people. However, equally noteworthy is the fact that Shennong was also associated with the invention of the market, where “at midday, he brought together all people under Heaven and assembled together all goods under Heaven. [The people] exchanged [their goods] and then retired, each obtaining what one wanted” (Zhouyi zhengyi: 8.5a). Shennong’s association with the origins of the rural market may well have been rooted in a utopian ideal that saw self-sufficient farmers in charge of the market forces behind their own produce. Yet in acknowledging that Shennong, who “followed the seedlings as his teacher” (Huainanzi: 1.16), also facilitated the market exchange, we find perhaps a tacit acknowledgement of the ultimate political risk, namely, that those who laboriously worked the soil in early China were not always the same as those who enjoyed its fruits.

Works cited Amano, Motonosuke 天野元之助 (1979; rpt. 1989) Chūgoku nōgōshi kenkyū 中國農業史研究. Tokyo: Ochanomizu shobō. Anderson, E.N. (1988) The Food of China. New Haven:Yale University Press. Benn, James A. (2015) Tea in China: A Religious and Cultural History. Honolulu: University of Hawai’i Press. Bodde, Derk (1975) Festivals in Classical China: New Year and other annual observances during the Han dynasty, 206 B.C.–A.D. 220. Princeton: Princeton University Press. Bray, Francesca (1984) Science and Civilisation in China, vol.VI, part 2 “Agriculture”. Cambridge: Cambridge University Press. Bu, Fengxian 卜风贤 (2006) Qin Han Jin shiqi nongye zaihai he nongye jianzai fanglue yanjiu 秦汉晋时期农 业灾害和农业减灾方略研究. Beijing: Zhongguo shehui kexue chubanshe. Cai, Wanjin 蔡万进 (1996) Qin guo liangshi jingji yanjiu 秦国粮食经济研究. Nei Menggu renmin chubanshe. Cai, Zhemao 蔡哲茂 (2009) ‘Wan Shang xumuye de jingying – yi mu yu chu wei zhongxin’ 晚商畜牧 業的經營 – 以牧與芻為中心, Gu wenzi yu gudai shi 古文字與古代史 (Academia Sinica) 2: 143–63. Cao, Lüning 曹旅宁 (2002) Qin lü xin tan 秦律新探. Beijing: Zhongguo shehui kexue chubanshe. Cao, Shujie 曹书杰 (2006) Houji chuanshuo yu Ji si wenhua 后稷传说与稷祀文化. Beijing: Shehui kexue wenxian chubanshe. Cao,Yuying 曹毓英 (2005) Jingtian zhi yanjiu 井田制研究. Wuhan: Huazhong shifan daxue. Chang, Kwang-chih (ed.) (1977). Food in Chinese Culture: Anthropological and Historical Perspectives. New Haven:Yale University Press. ———. (1999) ‘China on the eve of the historical period,’ in Michael Loewe and Edward Shaughnessy (eds.) The Cambridge History of Ancient China: From the Origins of Civilization to 221 B.C. (Cambridge and New York: Cambridge University Press), 37–73. Chuci jin zhu 楚辭今注 (1997). Edited by Tang Bingzheng 湯炳正 et al. Shanghai: Shanghai guji. Chunqiu Zuozhuan zhu 春秋左傳注 (1995). Annotated by Yang Bojun 楊伯峻. Beijing: Zhonghua. Daniels, Christian (1996) ‘Agro-Industries: Sugarcane Technologies,’ in Science and Civilization of China, vol. VI, pt. 3. Cambridge: Cambridge University Press. Ding, Bangyou 丁邦友 (2009) Handai wu jia xin tan 汉代物价新探. Beijing: Zhongguo shehui kexue chubanshe. Du, Qingyu 杜庆余 (2010) Handai tian zhuang yanjiu 汉代田庄研究. Jinan: Shandong daxue. Duan, Wei 段伟 (2008). Rang zai yu jian zai. Qin Han shehui ziran zaihai yingdui zhidu de xingcheng 禳灾与 减灾. 秦汉社会自然灾害应对制度的形成. Shanghai: Fudan daxue chubanshe.

315

Roel Sterckx Ebrey, Patricia B. (1974) ‘Estate and Family Management in the Later Han as seen in the Monthly Instructions for the Four Classes of People,’ Journal of the Social and Economic History of the Orient 17.2: 173–205. Er nian lüling yu Zouyan shu 二年律令與奏讞書. Edited by Wuhan daxue jianbo yanjiu zhongxin and Jingzhou bowuguan. Shanghai: Shanghai guji chubanshe, 2007. Flad, Rowan K. and Pochan Chen (2013) Ancient Central China. Centers and Peripheries along the Yangzi River. Cambridge: Cambridge University Press. Fu, Jiayi 傅嘉儀 ed. (2002) Zhongguo wadang yishu 中國瓦當藝術. Shanghai: Shanghai shudian. Ge, Quansheng 葛全胜 (2010) Zhongguo lichao qihou bianhua 中国历朝气候变化. Beijing: Kexue chubanshe. Graham, Angus (1979) ‘The Nung-chia ‘School of Tillers’ and the Origins of Peasant Utopianism in China,’ Bulletin of the School of Oriental and African Studies 42.1: 66–100. Guanzi jiaozhu 管子校注 (2004). Edited by Li Xiangfeng 黎翔鳳. Beijing: Zhonghua. Guo, Qinghua (2010) Mingqi Pottery Buildings of Han Dynasty China: 206 bc–ad 220. Brighton: Sussex Academic Press. Han Feizi jishi 韓非子集釋 (1991). Attributed to Han Fei 韓非 (d. 233 bce); annotated by Chen Qiyou 陳奇猷. Gaoxiong: Fuwen. Han, Maoli 韩茂莉 (2012) Zhongguo lishi nongye dili 中国历史农业地理, 3 vols. Beijing: Beijing daxue chubanshe. Hanshu 漢書 (1962). Beijing: Zhonghua shuju. Hara, Motoko 原宗子 (2013) ‘Kodai Chūgoku ni okeru jumoku e no ninshiki no hensen,’ 古代中国にお ける樹木への認識の変遷, Tōyō bunka kenkyūjo kiyō 163, 1–45. He, Kewei 贺科伟 (2009) ‘Cong wenzi wadang kan Handai nongye xisu’ 从文字瓦当看汉代农业习俗, Nongye kaogu 3: 152–55. Henan bowuguan ed. (2002) Henan chutu Handai jianzhu mingqi 河南出土汉代建筑明. Zhengzhou: Daxiang chubanshe. Hollmann, Thomas O. (2014) The Land of the Five Flavors. A Cultural history of Chinese Cuisine. New York: Columbia University Press. Hsu, Cho-yun (1980). Han agriculture. The formation of early Chinese agrarian economy (206 B.C.–A.D. 220). Seattle and London: University of Washington Press. Hu, Zexue 胡泽学 (2006). Zhongguo li wenhua 中国犁文化. Beijing: Xueyuan chubanshe. Huainanzi honglie jijie 淮南子鴻烈集解 (1992). Edited by Liu Wendian 劉文典. Taipei: Wenshizhe. Huang, Fucheng 黄富成 (2006) ‘Luelun Handai xiangcun nongguan – li tian’ 略论汉代乡村农官- 力田, Nongye kaogu 4: 79–82. Huang, Hsing-tsung (2000) ‘Fermentations and Food Science’, in Joseph Needham (eds) Science and Civilisation in China, vol. 6, Part V (Biology and Biological Technology). Cambridge: Cambridge University Press. Hubeisheng Jingzhou shi Zhou Liangyuqiao yizhi bowuguan (eds.) (2001) Guanju Qin Han mu jiandu 關 沮秦漢墓簡牘. Beijing: Zhonghua shuju. Hulsewé, Anthony F.P. (1985) Remnants of Ch’in Law. Leiden: E.J. Brill. Keightley, David N. (1999) ‘The Shang: China’s First Historical Dynasty,’ in Michael Loewe and Edward Shaughnessy (eds.) The Cambridge History of Ancient China: From the Origins of Civilization to 221 B.C. (Cambridge and New York: Cambridge University Press), 232–91. ——— (2012) Working for his Majesty. Research Notes on Labor Mobilization in Late Shang China (ca.1200– 1045 B.C.), as Seen in the Oracle-Bone Inscriptions, with Particular Attention to Handicraft Industries, Agriculture, Warfare, Hunting, Construction, and the Shang’s Legacies. Berkeley: Institute of East Asian Studies (China Research Monograph 67). Kuhn, Dieter (1988) ‘Textile Technology: Spinning and Reeling,’ in Joseph Needham (ed.) Science and Civilisation in China, vol 5, part 9. Cambridge: Cambridge University Press. Li,Yaguang 李亚光 (2009) Zhanguo nongye shigang 战国农业史纲. Changchun: Jilin daxue. Liji jijie 禮記集解 (1995). Edited by Sun Xidan 孫希旦. Beijing: Zhonghua. Liu, Jiapeng 刘甲朋 (2010) Zhongguo gudai liangshi chubei tiaojie zhidu sixiang yanjin 中国古代粮食储备调 节制度思想演进. Beijing: Zhongguo jingji chubanshe. Liu, Li and Xingcan Chen (2012) The Archaeology of China: From the late Paleolithic to the Early Bronze Age. Cambridge: Cambridge University Press. Liu, Tseng-kuei (2009) ‘Taboos: An Aspect of Belief in the Qin and Han,’ in John Lagerwey and Mark Kalinowski (eds.) Early Chinese Religion. Part one: Shang through Han (1250 bc-220 ad) (Leiden: E.J. Brill), 881–948.

316

Food and agriculture Liu, Wenjie 刘文杰 and Xu Dezhang 徐德章 (1983) ‘Sichuan Handai potang shuitian moxing kaoshu’ 四 川汉代陂塘水田模型考述, Nongye kaogu 1: 132–35. Liu,Yun 刘云 (2006) Zhongguo zhu wenhua shi 中国箸文化史. Beijing: Zhonghua shuju. Lu, Houyuan,Yang Xiaoyan,Ye Maolin, Liu Kam-Biu, Xia Zhengkai, Ren Xiaoyan, Cai Linhai,Wu Naiqin and Liu Tung-Sheng (2005) ‘Culinary archaeology: Millet noodles in Late Neolithic China,’ Nature 437: 967–68. Luo, Yunbing 罗运兵 (2012) Zhongguo gudai zhulei xunhua, siyang yu yishixing shiyong 中国古代猪类驯 化，饲养与仪式性使用. Beijing: Kexue chubanshe. Mao, Biao 馬彪 (Ba Hyō) (2013) Shin teikoku no ryōdo keiei: Unbo Ryūkō Shinkan to Shikōtei no kin’en 秦 帝国の領土経営 : 雲夢竜崗秦簡と始皇帝の禁苑. Kyōto: Kyōto Daigaku Gakujutsu shuppankai. Mengzi zhengyi 孟子正義 (1996). Edited by Jiao Xun 焦循. Beijing: Zhonghua. Mozi jiangu 墨子閒詁 (2001). Edited by Sun Yirang 孫詒讓. Beijing: Zhonghua. Nishijima, Sadao (1986) ‘The Economic and Social History of Former Han,’ in Michael Loewe (ed.) The Cambridge History of China, vol. 1,The Ch’in and Han Empires (Cambridge: Cambridge University Press), 545–607. Nylan, Michael and Griet Vankeerberghen (eds.) (2015) Chang’an 26 bce. An Augustan Age in China. Seattle: University of Washington Press. Ōta, Masako 大田昌子 (2001) Hashi no genryū o saguru: Chūgoku kodai ni okeru hashi shiyō shūzoku no seiritsu 箸の源流を探る：中国古代における箸使用習俗の成立. Tokyo: Kyūko shoin. Pattinson, David (2012) ‘Pre-modern beekeeping in China: a short history,’ Agricultural History 86:4: 235–55. Pirazolli-t’Serstevens, Michèle (1991) ‘The Art of Dining in the Han period. Food Vessels from Tomb No. 1 at Mawangdui,’ Food and Foodways 4: 209–19. Rickett, W. Allyn (1985–1998) Guanzi. Political, Economic, and Philosophical Essays from Early China, 2 vols. Princeton: Princeton University Press. Serventi, Silvano and Françoise Sabban (2002). Pasta: The Story of a Universal Food. New York: Columbia University Press. Shelach-Lavi, Gideon (2015) The Archaeology of Early China: From Prehistory to the Han Dynasty. Cambridge: Cambridge University Press. Shi, Shenghan (1959, rpt.1982) On “Fan Shêng-chih shu”: An Agriculturist Book of China Written by Fan Shêngchih in the First Century B.C. Beijing: Science Press. Shiji 史記 (1959) Begun by Sima Tan 司馬談 (d. 112 bce), primarily composed by Sima Qian 司馬遷 (ca. 145- ca. 86 bce). Beijing: Zhonghua. Sterckx, Roel (1996) ‘An Ancient Chinese Horse Ritual,’ Early China 21: 47–79. ——— (2002) The Animal and the Daemon in Early China. Albany: State University of New York Press. ——— (2011) Food, Sacrifice, and Sagehood in Early China. New York: Cambridge University Press. ——— (2015a) ‘Ideologies of the peasant and merchant in Warring States China,’ in Yuri Pines, Paul Rakita Goldin, and Martin Kern (eds.) Ideology of Power and Power of Ideology in Early China: Studies in Early Chinese Political Thought (Leiden: E.J. Brill), 211–48. ——— (2015b) ‘Alcohol and Historiography in Early China,’ Global Food History 1.1: 13–32. Sun, Ji 孙机 (1998) ‘Doufu wenti’ 豆腐问题, Nongye kaogu 3: 292–96. ——— (2008) Handai wuzhi wenhua ziliao tu shuo 汉代物质文化资料图说. Shanghai: Shanghai guji chubanshe. Sun, Weizu (2010) The History and Art of Chinese Seals. Beijing: Foreign Languages Press. Swann, Nancy Lee (1950) Food and Money in Ancient China. Princeton: Princeton University Press. Tan, Xuming 谭徐明 (2004) Dujiang yan shi 都江堰史. Beijing: Kexue chubanshe. von Falkenhausen, Lothar (2006) Chinese Society in the Age of Confucius (1000–250 bc): The Archaeological Evidence. Los Angeles: Cotsen Institute of Archaeology, University of California. Waley, Arthur (1996) The Book of Songs. New York: Grove Press. Wang, Q. Edward (2015) Chopsticks. A Cultural and Culinary History. Cambridge: Cambridge University Press. Wang, Qizhu 王啟柱 (1994) Zhongguo nongye qiyuan yu fazhan 中國農業起源與發展.Taipei: Guoli bianyi guan. Wang, Xueping 王雪萍 (2010) Zhouli yinshi zhidu yanjiu 周禮飲食制度研究.Yangzhou: Guangling shushe. Wang,Yong 王勇 (2004) Dong Zhou Qin Han guan zhong nonye bianqian yanjiu 东周秦汉关中农业变迁研 究. Changsha:Yuelu shushe. ——— (2008) ‘Qin Han difang nong guan jianzhi kaoshu’ 秦汉地方农官建置考述, Zhongguo nong shi 3: 16–23.

317

Roel Sterckx Wang, Yuanlin 王元林 (2005) Jing Luo liuyu ziran huanjing bianqian yanjiu 泾洛流域自然环境变迁研究. Beijing: Zhonghua shuju. Wang, Zijin 王子今 (2007) Qin Han shiqi shengtai huanjing yanjiu 秦汉时期生态环境研究. Beijing: Beijing daxue chubanshe. Xia, Henglian 夏亨廉 and Lin Zhengtong 林正同 (eds.) (1996) Handai nongye huaxiang zhuanshi 汉代农 业画像砖石. Beijing: Zhongguo nongye chubanshe. Xiao, Kezhi 肖克之 (2007) Nongye guji banben cong tan 农业古籍版本丛谈. Beijing: Zhongguo nongye chubanshe. Xu, Hairong 徐海榮 (ed.) (1999) Zhongguo yinshi shi 中國飲食史, 6 vols. Beijing: Hua Xia chubanshe. Xu,Wangsheng 徐旺生 (2009) Zhongguo yang zhu shi 中国养猪史. Beijing: Zhongguo nongye chubanshe. Xu,Weimin 徐卫民 (2011) Qin Han ducheng yu ziran huanjing guanxi yanjiu 秦汉都城与自然环境关系研 究. Beijing: Kexue chubanshe. Xu,Ying 徐莹 and Li Changwu 李昌武 (eds.) (2013) Jia Sixie yu Qimin Yaoshu yanjiu lunji 賈思勰与齐民 要術研究论集. Jinan: Shandong renmin chubanshe. Yang, Shengnan 楊升南 and Ma Jifan 馬季凡 (2010) Shangdai jingji yu keji 商代經濟與科技. Beijing: Zhongguo shehui kexue chubanshe. Yu, Kunqi 于琨奇 (2012) Zhanguo Qin Han xiao nong jingji yanjiu 战国秦汉小农经济研究. Beijing: Shangwu yinshuguan Yue, Isaac and Siufu Tang (eds.) (2013) Scribes of Gastronomy. Representations of Food and Drink in Imperial Chinese Literature. Hong Kong: University of Hong Kong Press. Zeng, Xiongsheng 曾雄生 (2008) Zhongguo nongxue shi 中国农学史. Fuzhou: Fujian renmin chubanshe. Zhang, Chunlong 张春龙 (2007) ‘Liye Qin jian ci Xiannong, ci Yin?, he ci Di jiaojuan’ 里耶秦簡祠先 農，祠窨(?)和祠隄校劵, in Wuhan Daxue jianbo yanjiu zhongxin (eds.) Jianbo 簡帛, vol.2 (Shanghai: Shanghai guji), 393–96. Zhang, Feng 张凤 (2009) ‘Cong kaogu ziliao kuice Handai han, shui tian de shengchan jingying’ 从考古 资料窥测汉代旱，水田的生产经营, Sichuan wenwu 6: 54–58. Zhang,Yihe 章义和 (2007) Zhongguo huang zai shi 中国蝗灾史. Hefei: Anhui renmin chubanshe. Zhou, Xin 周昕 (2005) Zhongguo nongju fazhan shi 中国农具发展史. Jinan: Shandong kexue jishu chubanshe. ——— (2010) Zhongguo nongju tongshi 中国农具通史. Jinan: Shandong kexue jishu chubanshe. Zhouyi zhengyi 周易正義 (1985). Annotated by Kong Yingda et al. Shisanjing zhushu edition Taipei: Xinwenfeng. Zhu, Hongbin 朱宏斌 (2010) Qin Han shiqi quyu nongye kaifa yanjiu 秦汉时期区域农业开发研究. Beijing: Zhongguo nongye chubanshe. Zhu, Kezhen 竺可楨 (1979) ‘Zhongguo jin wu qian nian lai qihou bianqian de chubu yanjiu’ 中国近五 千年来气候变迁的初步, in Zhu Kezhen wenji文集. Beijing: Kexue chubanshe. Zhuangzi jishi 莊子集釋 (2010). Annotated by Guo Qingfan 郭慶籓 (1844–1897 ce). Taipei: Guanya.

318

15 WARFARE

WICKY W.K. TSEWARFARE

Wicky W.K. Tse

Unlike their Greco-Roman counterparts, the veteran soldiers in early China left no first-hand account of what they had experienced and how they had fought on the battlefield – at least none has come down to us. Although ancient military writers of China had produced an array of treatises on strategies, tactics, doctrines, generalship, stratagems, and military philosophies, those are mainly handbooks and theoretical texts without clear demarcation between description and prescription in nature and are therefore nothing comparable with Thucydides’ The Peloponnesian War, Xenophon’s The Expedition of Cyrus, and Julius Caesar’s The Gallic War and The Civil War, as well as the three accounts of the wars of Alexandria, Africa, and Spain attributed to him, in terms of providing vivid records of lives in barracks and campaigns and giving us a glimpse of the real face of battle. In other words, voices from the combatants are lacking in the Chinese written tradition, and the real practice of Chinese warfare cannot be understood by referencing the military classics alone. Furthermore, the obscurity and difficulty of understanding early Chinese warfare have been aggravated by the brevity and standard metaphorical modes of description in the official histories; a battle would be narrated only in a few words, such as ‘X and Y armies engaged in somewhere’, ‘the arrows fell like rain’, and ‘X defeated Y, and the latter was routed’ (Graff 2009: 143–49). Pageantry and policy debates held before and after the battle usually attracted more attention than combat from the chroniclers, who were generally literati and civil officials without hands-on experience of fighting on the front line (Creel 1935: 343); sometimes the chroniclers repeated a stereotype instead of reporting what they actually saw or even used flowery writing as a way to compensate for their ignorance of actual happenings (Graff 2002: 7–8). While acknowledging these obvious caveats and limits, it is still possible for us to grasp certain features of early Chinese warfare based on the patchy evidence extracted from both received texts and archaeological findings. This chapter aims to provide an overview, as comprehensive as possible, of early Chinese warfare in three sections.1 The first section offers a general account of how war was conducted in early China in chronological order, covering the pre-imperial period and the age of the Qin, Former Han, and Latter Han empires, whereas the second focuses on how war was prepared, especially on the mechanisms for recruiting and preparing soldiers. The first two sections thus together depict the picture of warfare, as the activity conducted in the state of war, at an operational level. The third section, however, discusses the abstract concept of how war was perceived among early Chinese thinkers and statesmen, with 319

Wicky W.K. Tse

special emphasis on their pursuit of the justification and legitimation of military fighting. Early China was a period witnessing a flowering of military writings which revealed the spectacular development of military affairs of the times. Since much scholarly effort has been devoted to the authorship, intellectual affiliation, nomenclature, and thoughts of ancient military classics (Yates 1988: 211–48 and 2005: 65–79; Sawyer 1993; Lewis 2005: 1–15), with no end in sight, this chapter therefore only consults the classics for clues of the practice of warfare. To place warfare in a larger context of early Chinese history in general, this chapter also sets the timeless interactive relationship between warfare and state/empire formation, as well as that between war and society, as an important dimension.

Doing war War was at the core of the history of early China. As the oft-quoted saying of the sixth century bce goes, ‘the great affairs of a state are sacrifice and war’ (國之大事在祀與戎) (Yang 2000: 861),2 contemporary ruling elites acknowledged well the crucial role that war played in governance. War served not only as a tool to establish and sustain a regime but also as a weapon to destroy it. Being the most organized form of violence, the largest in scale, and most devastating in effect, war would devour all resources available and incur huge costs to the participants – even to the victors. Early Chinese thinkers therefore generally called for serious caveat before launching a war. Nevertheless, war was still ubiquitous in early China. There was no concrete evidence of when the earliest war happened in the region known as China today. War, however, as organized lethal violence between two groups (Ferguson 2003: 28), was present in China at least by the second millennium bce and was associated with the emergence of the early state. Before that, only obscure evidence of war was left by the prehistoric people: deep ditches believed to be for defensive purposes were dug around the Neolithic Yangshao 仰韶 villages in the Yellow River valley around 5000 bce; later, rammed earthen walls and scalped and decapitated skeletons from around 2600 bce were discovered throughout the Longshan 龍山 cultural sites in North China and are said to be traces of war (Ferguson 2003: 32; Chang 2005: 125–26); a pile of skeletal remains bearing the wounds of weapons were excavated in the site of Taosi 陶寺 dating back to around 2000 bc and is suggested to be the result of a rebellion (Xu 2014: 2–3). These evidences, however, can also be explained in another way. The ditches and walls would be constructed to keep out not only human enemies but also predators, skeletal trauma could be caused by interpersonal or other forms of violence apart from war, and the post-mortem bone damage might be part of mortuary rituals or other presently unknown belief practices rather than collateral consequences of war. It is therefore difficult to determine whether those cases were primitive warfare without new and further concrete evidence. Nevertheless, primitive warfare – happening in the pre-state era – was vividly portrayed retrospectively in the early Chinese historical accounts. Those accounts were more of a product of the mixture of imagination and reified memories passed down through generations than an objective reality; the narrated fighting was identified as proto-state warfare with the advent of Chinese civilization. For example, the Yellow Emperor, a legendary hero who was honoured as the progenitor of the Chinese people, was recorded in the Shiji 史記 as using war as a means to punish tyrannical chieftains, consolidate his own authority, and impose order upon chaos. Although the wars between the Yellow Emperor and his nemesis were supposed to have happened in the incipient stage of Chinese civilization, they were depicted as sophisticated and well organized. The Yellow Emperor was said to have provided substantial training to his troops, formed military alliances, set up military institutions, and invented new technological tools to assist his war effort. Different from the ancient Greek epics of warfare, the accounts of the 320

Warfare

Shiji and other historical records downplayed the role of divine intervention – though certain obscure sources narrated that both the Yellow Emperor and his archrival Chiyou 蚩尤 had divinities joining in the fighting. The refined version of the Yellow Emperor’s campaigns in the early historical accounts provided an orthodox answer to the question of how far back war went in time and a linkage between warfare and the emergence of the early state. The classical texts clearly show that war came hand in hand with the development of state and kingship in the early era of Chinese civilization (Li 1997: 215–48; Chang 2005: 125–28). The Xia 夏, presumably the first dynasty in Chinese history after the reigns of the legendary sage kings, is said to have consolidated rule by means of war. There is a war speech attributed to an early Xia king in the Shangshu 尚書 or Book of Document, and duplicated in the Shiji as well, in which the Xia king defined his military action as a punitive campaign against an unruly subordinate. War speeches attributed to the early rulers like this constituted a genre in the Shangshu and indicated the compilers’ perception of the importance of warfare in the early states and their efforts in tracing the historical origin of legitimate casus belli. In contrast with the Xia’s lack of contemporary written documents, the succeeding Shang 商 dynasty provided both literary and non-literary evidence of war. The oracle-bone inscriptions and artefacts such as arms and other military equipment excavated in the Shang sites since the early twentieth century have provided us with a valuable glimpse into certain kinds of military activities of the Late Shang period, covering the reigns from King Wu Ding 武丁 (ca. ?–1189 bce) to King Di Xin 帝辛 (ca. 1086–1045 bce).The Late Shang frequently engaged in wars with polities or tribal peoples at its peripheries for the purpose of defense or of expansion initiated by ambitious or aggressive rulers – in fact, these were two sides of the same coin, as the ‘Shang dynastic expansionism was aimed not at creating a unified empire, but rather at establishing frontier control regimes to secure resources for the Shang war machine, sacrificial practice, divination, and bronze making’ (Wu 2012: 26). Expansion was therefore a means to consolidate and strengthen the security of Shang, especially its heartland. Networks of military alliance with rulers of distant zones were also weaved for the same strategic purpose, though little is known about the detailed division of labour between Shang and its allies (Keightley 2012: 176). With its rather sophisticated state system, Late Shang was able to raise armies of about 3,000 to 5,000 men, in addition to supplying them with chariots, horses, weapons, food, and fodder (Keightley 2012: 99 and 178). The Shang deployed both chariots and infantry in war, with the latter as the mainstay. Chariots were excavated from the Shang tombs in Anyang 安陽, and the Shang elites or nobles were supposed to ride in chariots and use them as mobile command and firing platforms, but references to their practical use in warfare are few (Keightley 2012: 187). Some scholars even suggested that the newly imported chariot was primarily an emblem of royal status and actually played an extremely limited role in most of the Shang times (Shaughnessy 1988: 199; Wu 2013: 49–50; Bulliet 2016: 115 and 122). Also found in the tombs were weapons such as dagger-axes and spears – generally made of bronze – bows and quite a large number of arrowheads. The Shang troops took captives for slavery or ancestral worship, and one of the alien peoples called Qiang 羌 were frequently mentioned in the oracle-bone inscriptions as being beheaded for human sacrifices (Creel 1935: 340; Keightley 2012: 48, 66–69, and 184). Women were also seized in war and, in certain circumstances, submitted to the Shang rulers as a token of surrender and peace negotiation. Daji 妲己, the legendary luscious consort of the last Shang king, is said to be sent to the Shang harem as a form of paying homage by a defeated polity. Meanwhile, women did take an important part in warfare, which would be rarely seen in later times when war was only the work of men; a consort of King Wu Ding named Fu Hao 婦好, for example, assumed the commandership of the Shang army. 321

Wicky W.K. Tse

Although the oracle-bone inscriptions were essentially scripts of divination, the information of Late Shang warfare derived therein manifests a feature of ritualization of war. Ancestral divination was not only a prerequisite for the Shang rulers in making a decision to have war but also a necessary ceremony at the departure and return of the army and the king’s reception of war captives. War was thus built into the Shang culture and the world of the ancestors, which echoed the aforementioned, saying that ‘the great affairs of a state are sacrifice and war’. It gave the Shang warfare a certain form of legitimacy and indicated that, after the divination, war was not random violence. Although the ceremony would vary in later eras, pre-war consultation and proclamation and post-war presentation of trophies at the ancestral temple consistently served as necessary rituals in any formal military campaign. The Shang sources would only tell the Shang side of the war stories, not to mention that they scarcely recorded things of little concern to the rulers. The enemies of the Shang were mostly silenced in the extant sources, except for the Zhou people, who finally replaced Shang as the hegemon of North China. The Shang-Zhou military confrontation peaked at the Battle of Muye 牧野. The battle was depicted as an Armageddon between the benevolent Zhou rulers and the evil Shang tyrant, a battle of virtue versus vice, in the received texts of the Zhou tradition; it was also regarded as an archetypical battle often quoted as an allegory in war speeches and accounts in pre-modern Chinese history. Well known as it was, the details of the battle are in fact rather vague, and even its date lacks unanimity among modern scholars, with a wide range between 1130 bce and 1018 bce; the number of troops involved, especially on the Shang side, is generally believed to be highly exaggerated in the historical texts. It is also difficult to prove if the Zhou troops really deployed 300 war chariots and 3,000 elite warriors called huben 虎賁 (literally tiger warriors) in the battle as the sources say. But some scholars suggested that the success of the Zhou might be attributed to their employment of war chariots as a shock-troop (Shaughnessy 1988: 228–31). On the other hand, in a widely read pre-battle speech,3 the Zhou king is said to order his men: ‘do not advance more than six or seven steps, and then stop and adjust your ranks . . . do not exceed four blows, five blows, six blows, or seven blows, and then stop and adjust your ranks’ (Legge 2000: 304). This might give a glimpse of how the Zhou foot soldiers were supposed to be arranged in serried ranks and move in unison with others during the battle. To re-shape the political landscape of the former Shang ecumene, the western-based confederation under the Zhou leadership steadfastly pushed their forces eastward and carried out a policy of military colonization. Members of the Zhou ruling lineages and their allies led forces to the east, relentlessly suppressed local resistance of the residual Shang forces and their sympathizers, secured strategic points, and established their new states. The Zhou control over North China was thus established through a defense network connecting the Zhou royal domain and various outlying states. Once the surge of early expansion was over, the Zhou army transformed its function from conquest to garrison and turned their focus to the real or perceived threats of the peoples who were recalcitrant to comply with the Zhou political order, resulting in pre-emptive military campaigns. King Zhao 昭, the fourth monarch of the dynasty, personally led troops against the Huai Yi 淮夷 tribes scattering along the southeastern frontier. The king is said to have drowned in the Han River in the midst of an operation, and his unfinished expedition thus marked the southern reach of the Zhou realm. While in the north, the Zhou states had to cope with the harassment from various Rong 戎 peoples, with the most pressing menace from the Xianyun 玁 狁 residing in the northwestern highlands adjacent to the Zhou royal domain (Li 2006: 141–92). Contemporary sources like the bronze vessels inscriptions and poems collected in the Shijing 詩 經 or Classic of Odes provided some information about the Zhou military campaigns, though from the aristocratic point of view, by praising the merits of the martial elite or lamenting the 322

Warfare

onerous duties. The confrontations between the Zhou and non-Zhou peoples were further depicted in the later historical writings as the struggle between the civilized and the savages, which subsequently became a recurring theme in pre-modern Chinese histories. In 771 bce, the Xianyun sacked the Zhou capital and killed the king. The uprooted Zhou royalty then moved east to the heretofore ancillary capital Luoyang and maintained the dynastic mandate under increasingly adverse circumstances. The royal relocation not only signified a symbolic rupture of the Zhou dynasty but also marked the beginning of a new epoch that was characterized by the intensification of war. Capitalizing on the decline of royal power, some strong Zhou and non-Zhou states expanded territorially at the expense of their weak neighbours. In the Spring and Autumn period (772– 468 bce/770–404 bce), the number of local states that once were scattered over North and Central China dwindled, and only seven large regional kingdoms and a handful of minor states survived the frequent internecine warfare and entered the Warring States period (403–221 bce). The various tribal peoples who once inhabited the areas between the Zhou states were also absorbed into the expanding powers or squeezed out of the Zhou ecumene. The scale and intensity of war in the Warring States period – as its name tells – escalated and ended in the destruction of all but one state, which created the first empire in Chinese history in 221 bce. Over the course of these turbulent centuries leading to the imperial age, changes in political and social circumstances triggered certain transformations in the military sphere, which significantly shaped the features of Chinese warfare thereafter. The first and foremost transformation was the gradual falling out of chariotry on the battlefield. The chariot was commonly used on the battlefield during the Western Zhou dynasty,4 and reached its pinnacle in the Spring and Autumn period. The number of chariots in one’s possession was customarily counted as the metrics of a state’s military strength; even in the Warring States period, when the military importance of chariots had receded, terms like a state of a thousand chariots (qian cheng zhi guo 千乘之國) or a state of ten thousand chariots (wan cheng zhi guo 萬乘之國) were used to rank a state (Wu 2013: 112). The scale of a military campaign was also assessed by the number of chariots being fielded. In the Battle of Chengpu 城濮 (632 bce), for example, the state of Jin 晉 deployed an unprecedented number of 700 chariots to defeat Chu and its allies, marking the battle hitherto one of the largest scale.5 Chariot warfare was conventionally portrayed and widely accepted as the prevailing pattern of Spring and Autumn warfare, following the accounts of the Zuo zhuan 左傳 – the most important textual sources of the period – and the modern interpretation of the unearthed chariot remains (Lu 1993: 824–38). But the real function of war chariot is in fact far from certain. There are little details about how a chariot was actually used on the battlefield and how chariots organized in a formation with each other and other sections of an army.6 Nevertheless, based on the scant information, we can still sketch the character of chariot warfare and how and why certain developments in warfare finally marginalized the use of war chariot. The fortunes of chariots were in fact closely associated with those of the combatants involved. Since it required a lot of resources and time in training to make a qualified charioteer or chariot archer, it was unsurprising that only the aristocracy would be able to afford chariots; it was also a common feature across ancient Eurasia as a hereditary caste of warriors riding in chariots to suit and uphold their elite status and outlook (Bulliet 2016: 121–22, 137). The Spring and Autumn aristocracy shared a common literary and martial culture; most of them planned and were indoctrinated in their life trajectories to govern and to go to war; they trained from their youth in driving a chariot and shooting on and off it. The high-born warriors also observed certain rules of ritual etiquette in chariot warfare for upholding their social and cultural status; for example, one had to halt the chariot, take off his helmet and salute the enemy of senior or 323

Wicky W.K. Tse

superior status encountered on the battlefield. Knightly combat was thus portrayed in the Zuo zhuan as one of the defining characteristics of the Spring and Autumn noble warfare – though the depiction was mainly from the point of view of the ruling elite and somewhat in idealized form, with details of pageantry and rituals other than real combat (Kierman 1974: 27–56). As long as the hereditary aristocracy controlled the Spring and Autumn politics, fighting on the chariot was a privilege of the noblemen, and the chariot dominated the battlefield; when the military importance of the chariot-centred warriors declined, so did the vehicle. The inherent technical limitations of chariot on the battlefield also sowed the seeds of its gradual military obsoleteness. In theory, a chariot battle would begin with two lines of charioteers facing off on a flat plain, with each side widely spread and of only one or two vehicles deep in order to get a good view of the enemy and have enough room to maneuver. Each chariot would have a driver in the middle, an archer on the left and a dagger-axe fighter on the right, whereas infantry presumably presented only in small numbers and accompanied the chariots.7 When combat started, the two lines drove at each other and shot arrows. If the opponent’s chariots were not stopped during the crossfire, then it would be the dagger-axe fighters’ turn against the opposing counterparts when the vehicles drew close. The side with the upper hand would then exploit the break of the opposing line and drive a wedge into it or even split to go around it on one or both sides. The obvious limitation of riding chariot was, therefore, terrain. Obstacles on the battlefield and rugged landscape would deter or offset the momentum of the charging chariot – a main function of the infantry advancing alongside was probably to remove barriers and help the chariot out from ruts and holes. Difficult physical environments such as mountains, ravines, and woods, as well as poorly maintained roads, would also severely restrict the room of maneuver of chariots. Even on a flat plain, a line of chariots would still find it difficult to move forward in an even and disciplined fashion and to change its direction freely as the battle progressed.8 Furthermore, the surge in the wall-building movement during the Warring States period made siege warfare – which most armies tried to avoid in the preceding era – a commonplace, in which the chariot could hardly play a role (Lu 2005: 210–23). Coming hand in hand with the military obsoleteness of chariots was the second transformation – the resumption of infantry ascendancy.9 For sure, foot soldiers were the primary type of military forces in history and had been active on the battlefields long before the emergence of chariots, but they were overshadowed by the vehicles and were assigned only ancillary roles during the heyday of chariot warfare; they were also politically and socially inferior to the chariotborne aristocratic warriors. The dominance of chariots, however, did not last long in China and elsewhere in the world. As Lan Yongwei’s classic study on the Spring and Autumn infantry shows, new infantry tactics were increasingly adopted at the time, and particularly in conflicts with peoples of non-Zhou origins positioned in mountains or other rugged terrain who fought on foot (Lan 1979: 37–53 and 72–102). When the inherent limitations of the chariot prevented it from operating on a new battleground, innovative commanders turned to the infantry and even ordered the warriors to get off their chariots and join the foot soldiers – despite meeting some reluctance at the very beginning (Yang 2000: 1215–16). The relative flexibility of infantry enabled them to move more effectively in difficult landscape than the chariots and also made them a vital element in siege warfare, which adopted swarming infantry tactics. Moreover, different from charioteers and cavalrymen, not to mention the navy, foot soldiers were the easiest type of forces that could be massively trained and produced. They could therefore meet the rising demand of the enormous input of military manpower with the unprecedented scale, duration, and distance of military operations in the Warring States period, as well as in the early empires. The massive infantry was generally deployed into three formations (left, middle, and right) in open pitched battle, but they could also be flexibly 324

Warfare

deployed to fit any specific circumstances. Infantry became the mainstay of forces and met no challenge in the early empires, although the cavalry and the navy would perform the leading role in specific operations. The cavalry also gained gradual ascendancy as an independent force, as the third transformation, while the chariot was gradually going out of use in combat and the horses were released from the driving yokes.10 It was arguably between the sixth and fourth centuries bce that the northern Warring States gradually adopted the style of fighting on horseback from the pastoral nomads, who recently emerged along and increasingly threatened their northern frontiers, with cavalry becoming a regular unit of the army by the mid-fourth century bce (Yates 2002: 40, 49). Until then horses were used for charioteering rather than for riding, though they had been domesticated and employed in the military in the Late Shang period (Yuan 2003: 110–4). The Warring States horsemen also adopted a kind of nomad’s outfit that allowed them to sit astride the horse – the trousers;11 a debate about the ‘barbarization’ of clothing style subsequently arose in the royal court of the northern state of Zhao, but the tactical benefit finally silenced their critics. It is commonly asserted that the cavalrymen were not capable of doing close-quarter combat before the invention and application of the stirrup to help them sit tightly on horseback; therefore the cavalry generally took up the tactical tasks of shooting, reconnaissance, and scouting in the advantage of their speed and flexibility (Yates 2002: 57–66). But it might underrate the capability of light cavalry in ancient warfare. As some scholars pointed out, Alexander the Great and his Macedonian troops, the contemporary of the Warring States Chinese, already made spectacular use of light cavalry during his campaigns without the stirrup, and for well-trained horsemen, the saddle was in fact more important for shock cavalry than the stirrup (Bulleit 2016: 125; Roland 2016: 32).12 The employment of infantry and cavalry provided combined arms and naturally enriched the array of tactics being used. As a matter of fact, the Spring and Autumn armies already skillfully adopted some sophisticated tactics, based on the examples preserved in the Zuo zhuan, such as the outflanking maneuver (Yang 2000: 45), ambush (Yang 2000: 65–6, 1104), deception (Yang 2000: 109–11, 134, 241–42, 434–5, 495–96, and 1038), feigned flight (Yang 2000: 125, 1104), interdiction (Yang 2000: 498), nocturnal assault (Yang 2000: 737–43), spying (Yang 2000: 241–42 and 435), and poisoning the drinking water (Yang 2000: 1009). With the advantages enjoyed by the cavalry, more new tactics were able to be executed. A novel one was deploying the cavalry to sever the enemy supply lines, which was first mentioned in the Warring States military treatises such as the Liutao (The Six Secret Teachings) (Sawyer 1993: 80) and the Sun Bin bingfa (Sun Bin’s Art of Warfare) (Lau and Ames 2003: 114–15 and 179), and henceforth became one of the most commonly used tactics in later eras (Tse, 2017). Further revealing the increasing complexity on the battlefield during the period leading to the imperial age was siege warfare. Siege was generally a form of attritional warfare, in which the army laying siege and the besieged army tried all means to capture or defend the fortified area, and it would end in the exhaustion of either side or even both. The Spring and Autumn armies always chose open-field battle and avoided siege warfare. The escalation of expansionistic wars, however, made siege a commonplace among the Warring States; fortified settlements would have to be taken as a proof of the occupation of new territories. Fortifications were strengthened, whereas various siege technologies, such as scaling ladders, movable towers, earth and wood ramps, and battering rams, and methods like mining and digging subterranean tunnels, were invented, refined, and utilized. Both fortification and techniques of siegecraft, as dueling technologies that developed dialectically in response to each other’s evolving capabilities, always marked the most advanced level of technology in the ancient world (Roland 2016: 24–5). And there was a school of specialists, namely the Mohist, who travelled across the Warring States and applied 325

Wicky W.K. Tse

their knowledge primarily to the defense of cities. Fragments of their instructions and manuals were preserved in the last twenty chapters of the Mozi, after which the Chinese techniques of siegecraft did not change much until the early modern period, when powerful stone-throwing artillery and later cannons were introduced (Yates 1988: 238–46; Needham and Yates 1994). With the scale in all aspects of both pitched and siege battles escalating in the Warring States period, an increasing number of people were involved and mobilized, of which a direct consequence was the large casualties that would have been impossible in the preceding ages. Not only were combatants killed in battle, but also non-combatants were slain in the ensuing massacre or were starved to death during the siege. The death toll of the Warring States period recorded in the histories was generally high, certainly not without exaggeration. The most controversial case was the Battle of Changping 長平 (260 bce), in which the Qin army defeated the Zhao troops after an exhausting campaign and is said to have slaughtered 400,000 surrendered Zhao soldiers. The accuracy of the number is of course debatable, but the presumed high death toll arguably revealed the large-scale militarization that dragged as many people as possible into wars; the massive involvement would not only contribute to the creation of the early empires but also shape their mode of mobilization. Large numbers of farmers-cum-soldiers were fielded, with the infantry as numerically the mainstay, in the Qin and Former Han empires. Since the last quarter of the second century bce, however, a series of outward expeditions launched by Emperor Wu of the Former Han necessitated military adaptation to new circumstances. The militia was no longer able to afford the long-range campaigns, temporally and spatially, against the Xiongnu and other neighbouring polities or peoples; professional soldiers were thus in need. Professional cavalry – not a small number of them were surrendered or recruited non-Han horsemen – played a particularly important role as the main strike force in the steppe operations, as far as to Central Asia (Loewe 1974: 67–104). The use of professional soldiers was more extensive in the Latter Han times, when the empire was troubled by the various peoples scattering along its long frontiers. On the other hand, the navy, in a rudimentary sense, was also trained for Emperor Wu’s campaigns in the southern tip of China and the Korean peninsula; the emperor even constructed a large lake in the capital area for the maneuver training of warships.The navy, however, was mainly operated on rivers and along the coast instead of engaging in the sea. One of its major functions was to deliver troops via waterway rather than fighting naval battles or, more precisely, riverine warfare, since the belligerent parties in early China, both internal and external, were usually land-based. As for weapons, apart from the conventional types being used and passed down from the preceding period, with iron having increasingly replaced bronze, two new inventions merit particular mention. The first is the single-edged sword or single-bladed sabre called dao 刀. Scholars suggest that the wide adoption of dao was a result of the increasing popularity of riding on horseback, as the weapon was more effective for a cavalryman slashing down the infantry (Trousdale 1975: 64–9). Another is the crossbow. Since it required less training in mastering it than the composite bow, the crossbow could significantly facilitate the undertrained farmer-soldiers and subsequently increased the firepower of the Chinese armies, especially when they were facing the steppe nomads, who were good archers. The crossbow was indeed a weapon associated with the massive employment of infantry. This linkage was part of the big picture of the rising dominance of commoner-infantry, which constitutes the main subject of the next section.

Preparing for war The earliest groups of warriors, in a broad sense, were probably the hunter-gatherers who picked up weapons and organized themselves to hunt game; hunting henceforth played an integral part 326

Warfare

in military exercise or drill in early China – besides its military function, its functions of ritual and ceremony advanced as time passed by – with textual evidence showing it at least by the Shang times (Fiskesjö 2001: 145–46). The primitive armed bands naturally extended the use of weapons from hunting animals to intertribal conflicts. A warrior class mainly constituted by the clan members and allies of the leaders also gradually emerged.The existence and prevalence of a hereditary martial aristocracy was a salient feature during most of the pre-imperial era. According to the extant textual sources, the Late Shang military force comprised two core components, namely the zhong眾, who were the cadres of the army and served as a permanent standing force, and the ren 人, who, as the militia, constituted the rank and file (Keightley 2012: 62). The armies were generally recruited in groups of 3,000 and further assigned in units of 1,000 each to the Right, Centre, and Left divisions, collectively called the Three Armies (san shi 三師), with each subdivided into a number of regiments (lü 旅). The largest extent of mobilization was about 3,000–5,000 troops.13 The Late Shang already organized its army on command lines and assigned military officers with particular duties; for example, the Duo She 多射 (Many Archers) and the Duo Ma 多馬 (Many Horse-chariots) were supposed to lead the archery units and chariot units, respectively (Keightley 2012: 183).14 As a colonial regime, the early Zhou garrisoned its core and standing troops, namely the Six Armies (liu shi 六師) and the Eight Armies (ba shi 八師), in the western capital and the eastern capital, respectively (Li 2008: 78–83). Meanwhile, the royal clan members and their allies led forces to strategic areas and established local states, which initially followed the regulations of the royal court – for instance, the number of troops of each state was capped according to its rank in the Zhou hierarchy. But the local states gradually developed their own military institutions and practices in response to the changing circumstances. While chariot warfare was the prevailing mode of fighting, the war vehicles were theoretically organized into squadrons of five, and five such units formed a brigade (twenty-five chariots); ten to twenty-five foot soldiers – presumably increased to seventy-five later – accompanied a chariot (Lan 1979: 90–100). Besides the aristocratic charioteers, the rank and file was constituted by the people living within the city wall, accordingly known as Guoren 國人, who were usually the collateral clansmen of the nobility and were influential stakeholders of state politics (Tu 1992: 476). In contrast, there were people called Yeren 野人 who resided outside the city gates and were ineligible for serving in the military. Nevertheless, to assure the sufficient number and continuous supply of foot soldiers, the expanding states introduced measures to recruit as many as possible and consequently removed the traditional distinction between the Guoren and the Yeren. In the early Warring States period, foot soldiers of commoner background were massively conscripted, thus produced an armed force of new social composition (Tu 1990: 49–96; Lewis 1990: 54–65). A large number of able-bodied male farmers were called up and were regarded by the state as the best soldiers – a perception that would be influential throughout the imperial age. With the eclipse of the hereditary martial aristocracy and the increasing number of commoners being recruited and trained for warfare, individual martial prowess honoured by the aristocracy in the old days was now replaced by obedience and action in unison; individual heroism would bring no reward but penalty – as the contemporary military treatises and numerous anecdotes repeatedly emphasized. Soldiers were trained to become familiar with signals given with flags, pennants, drums, and gongs, to act together and to know what the others were likely to do so as to adjust their actions accordingly. Besides punishment, the state also provided material incentives and rewards to encourage the commoner-soldiers to fight for its cause. The Qin was highly and yet notoriously successful in institutionalizing a meritorious ranking system based on the number of heads cut from the enemy, for the entry level, and the achievement of one’s commanding unit, for the 327

Wicky W.K. Tse

senior level. As one’s merit was in direct proportion to the number of heads captured, the brutality of warfare increased accordingly.15 Coming together with the massive conscription of the commoner-infantry was the employment of the general of humble background, which gave a new definition of generalship. A cluster of commoners who had studied strategies, tactics, and relevant military knowledge were appointed as commanders of the new Warring States armies. Their ascendancy undermined the general’s traditional role as a physical combatant taking his place at the front and fighting in battle. The new type of general was now a battle manager whose knowledge of operational management was far more important than his physical strength and fighting skills.The production of military treatises and manuals was also flourishing in this period, partly due to the aspirants who tried to catch the rulers’ eyes, earning personal advancement, and to transmit their one-time esoteric knowledge (Lewis 1990: 97–132 and 2005: 1–17). The mass mobilization served as an underpinning for the Qin and the early Former Han empires.The universal military service was the levy of nearly every able-bodied adult male, who had to spend one year for training in his home commandery and another for garrisoning at the capital or the frontiers and was on the list of reserves until the age of retirement – around 56. The conscripts were farmers who worked on the lands for most of the time, which restricted their availability for various military tasks. In an agrarian society like early China, the farmersoldiers could not serve long in both time and distance away from their home, given that the farming schedule would allow only limited days for military training, even though infantry was in fact the type of force that required the least training time but produced as many soldiers as possible, compared with charioteers, cavalrymen, and the navy. When the military campaigns were mainly within the empire, the limits of the farmer-infantry could be considerably eased. However, with the enlarged ranges, spatially and temporally, of campaigns against new enemies beyond the imperial frontiers since the reigns of Emperor Wu, the Former Han empire needed to deploy more volunteers and convicts who could join the expeditionary forces to fight further afield and for longer periods or even stay along the frontiers as long-term garrisons. Both volunteers – those who for whatever reason were not staying on farmland – and convicts were always regarded by the state as potential troublemakers if they were not disciplined, nor was their energy diverted to the imperial service. The recruitment of them was therefore also a pragmatic means of social control. The volunteers were usually depicted as unruly martial men who were tough and physically fit for military mission. A case in 99 bce was rather revealing: General Li Ling 李陵 (d. 74 bce) led an army of 5,000 swordsmen recruited and trained to participate in an expedition against the Xiongnu. Li and his men, however, were outnumbered and chased by the enemy. But they were still able to resist and deter the enemy for quite a few days, which indicates the strength of the volunteers; on the other hand, Li discovered women were kept among the troops who weakened the men’s energy and morale, which signifies the lack of discipline among the volunteers – similar to a Roman anecdote that Scipio Aemilianus expelled all the traders, prostitutes, and soothsayers hidden in his army in 134 bce. As for the convicts, they not only took part in fighting but also went to the frontiers or newly conquered regions with their families as a garrison or colonisers. In addition, the Han empire employed surrendered alien peoples to serve in the military, especially horsemen, who usually maintained their own socio-military organizations as a way to preserve their fighting ability and skills. Training, of course, was vital for whichever type of military forces. Although charioteering was no longer the most essential item in the training repertoire of the Qin-Han armies, there were still several cases of military officers gaining their reputation and winning the imperial favor with their spectacular driving skills – even not on the battlefields. For sure, there was a wide range of training items, and their importance would be varied according to the type of 328

Warfare

force and the nature of the tasks assigned. But the training of obeying order and discipline, acting in unison, situating oneself in tactical formation, and maneuvering in response to signals, etc., were core courses. In the Former Han, there were state-sanctioned annual military reviews at central and regional levels to evaluate the outcomes of training; the Latter Han, however, imposed restrictions on those exercises because of several precedents of ambitious military leaders, including the dynastic founder himself, hijacking the exercise and leading the troops assembled at the regional headquarters to launch revolts. Thanks to the unearthed Han wooden slips, we know that the Han military conducted routine archery tests in the frontier garrisons, highly likely also in the interior regions, with a clearly defined scale of reward and punishment based on the shooting accuracy. Along with the training of mastering an arsenal of weapons and assorted equipment and acting in tactical formation, the soldiers honed their physical and close-quarter combat techniques through sport games such as rock throwing, wrestling, boxing, running, and long jump (Loewe 1994: 236–48; Lorge 2012: 66–72). Besides martial exercises, serving in the military also provided many commoners the primary form of literacy education (Yates 2011: 360–64). The Latter Han empire, however, downsized the formal military training of the commoners and abolished the universal military service (Lewis 2000). The empire chose to maintain a standing force of those who, once enlisted, would stay in – though untrained conscripts would still be summoned in an emergency. The separation between civilians and military men subsequently became more pronounced than in the Qin and Former Han periods; the military would gradually become a specialized society within society as a whole with the collapse of Latter Han. Meanwhile, the non-Han proportion of troops in the imperial army was enlarged, as foreign horsemen were increasingly recruited. The reliance on professional armies – of both Han and non-Han origins – presaged the prevalence of private military forces that triggered the disintegration of the empire and further contributed to the political disunion in the ensuing centuries.16 Logistics and military finance, which supply material and cover related costs for maintaining the army and its fighting capability, were of vital importance to the early Chinese armies. The two systems or mechanisms, however, were not easily reconstructed in detail since they were not well documented nor fully discussed in ancient literature; they are seldom studied in modern scholarship either. Nevertheless, some scattered evidence can still be assembled for further research, as I have touched upon elsewhere (Tse, 2017). In general, logistics became more of a problem when the armies became larger, the duration of the campaign longer, and the theatre of operation farther; the Warring States period therefore marked a new page of logistics in early Chinese warfare. And the system of military finance also became more complex and sophisticated under the empire, which controlled a vast territory and even dispatched troops beyond its frontiers. Due to limited space, I would save this topic for another study. But the late Edward Dreyer’s remarkable research on the Former Han general Zhao Chongguo 趙充國 (137–52 bce), who left precious discussions with an emperor about the logistic support of a border campaign, will suffice to provide a practical example (Dreyer 2008: 665–725).

Thinking about war War was ubiquitous in early China, from which people were seldom spared, but for most of those who took part in war, whether they were attackers or defenders, commanders or rank and file, combatants or non-combatants, it was not meaningless, if not vital, to give them a certain motivation for risking their own lives. Besides merely fighting for survival or for material gains, war leaders and the state needed a primary casus belli to package the war aim(s) and to give a 329

Wicky W.K. Tse

sufficient justification for the decision of resorting to arms, which in turn might buttress the legitimacy of the regime. For much of early China, there were considerable interests in the discussion of the justification of war. Quite a number of statesmen, strategists, and political thinkers had put forward their own views for promoting and denouncing war in their times, which further provided essential references for discussions on the same topic throughout the imperial age. A common salient feature among the various views was the pragmatism of accepting the inevitability of war, and generally no pure pacifism existed – the so-called Daoism might come close to it, though not exactly. Therefore, what one would try was to define the principles and impose constraints on using military violence instead of trying to abandon war in an absolute manner. No later than in the Spring and Autumn period, the ruling elite had already recognized war as a kind of necessary evil and recognized its symbiotic relationship with the development of human society by saying: Heaven gives rise to the Five Resources, and the people use all of them. Discarding even one of them will not do. Who can remove weapons? Weapons have been in use for a long time; such is the means for using authority to forestall transgressions and for making manifest the virtue of culture. The sages rise by them, and those who foment disorder fall by them.The ways to determine rise and fall, preservation and destruction, darkness and illumination, all originate in weapons. (Yang 2000: 1136; Durrant 2016: 1203–5) Since removing weapons – military violence in a broad sense – was impossible, a realistic alternative was to regulate and restrain its use; otherwise disaster would strike. As an oft-quoted Spring and Autumn caveat says, ‘Warfare is akin to fire. If one does not contain it, one will be consumed oneself ’ (Yang 2000: 36; Durrant 2016: 31). The Laozi, the least prone to violence among the various pre-imperial philosophical texts, pointed out that weaponry was an inauspicious tool and should only be used in unavoidable circumstances, which still did not rule out the necessity to resort to military force. Preparing for war was therefore a crucial concept in Chinese military thinking. The Zuo zhuan, among other early Chinese texts, repeatedly underscored the ideas of ‘thinking about danger while residing in peace’ 居安思危 (ju an si wei) and ‘with sufficient preparations there will be no calamities’ 有備無患 (you bei wu huan), which echoed the Roman dictum of si vis pacem, bellum para (If you desire peace, prepare for war). Based on these caveats, modern scholar Alastair Johnston analyzed that there was a parabellum (prepare for war) paradigm in traditional Chinese strategic culture (Johnston 1995: 61–108). Coming along with the material preparedness was the theoretical foundation for using military violence; the just war theory or its associated notions were proposed and discussed over time in early China. Roughly classified, there were two perspectives of viewing and defining what made a war just or unjust. One was retrospectively analytical that identified the nature of the war only after the outcome; in other words, it would be the winner legitimatizing the military conducts. On the other hand, the prospectively prescriptive one tried to promote the ideas of a just war or righteous army as norms to guide the conduct of the belligerents. These two perspectives are parallel to the consequentialists, who justify a war act based on whether the consequences of not doing it are worse, and the intrinsicists, who believe certain acts are just or unjust in themselves, respectively (Echevarria 2017: 87). In fact, the two perspectives were not distinctively exclusive but often intertwined in the same discourse. Moreover, the righteousness of a war hinged largely on the moral issues before it was treated as a legal concern in early China. 330

Warfare

As shown earlier in this chapter, the Yellow Emperor’s wars against his enemies at the dawn of Chinese civilization were retrospectively depicted in later historical writings as righteous; the righteousness was closely associated with the legitimacy of the orthodox rulers or dynasties. A crucial notion of just war in Chinese history was invented by the Zhou people. The early Zhou conquerors, on one hand, denounced the vices and crimes of the last Shang king and, on the other hand, championed the benevolence of their own leaders, so as to manifest their bestowal of the Mandate of Heaven and to justify not only their violent takeover but also their governance of the former Shang subjects. Thereafter the Mandate of Heaven constituted a core doctrine, if not a cliché, in ancient Chinese political ideology. Following the Zhou precedent, a just war must be in response to the wrongs committed by the other side – usually an evil tyrant. Needless to say, the one who launched a war would list the wrongdoings of the other side and claim the operation as punitive in nature, whether pre-emptively or retrospectively, and the winner certainly had the last words. After destroying all the rival states in 221 bc, the First Emperor of Qin issued an edict to justify his military campaigns, claiming that the extermination of the six Warring States was incurred by their own misdeeds, such as betraying the alliance with the Qin, violating the accords with the Qin, severing the diplomatic relations with the Qin, and even sending assassin against him, all leaving him without alternatives but to resort to violence as a means of rectification and self-defense. In response, the courtiers praised the fact that the emperor had launched the ‘righteous armies’ (yibing 義兵) and successfully unified the Chinese realm (Lewis 2006). Self-defense was indeed a very strong argument for the legitimacy of war, which was generally shared by different schools of thought. Among them, the Mohists condemned wars for expansion and other aggressive campaigns but championed the rights of self-defense, and enthusiastically applied their knowledge of siegecraft to help defend the weaker side. In theory, a military operation that infringed on others’ rights of self-defense would be hardly justified as a righteous one. But the reality was more complicated. The principle of self-defense was responsive and passive in nature, and the theorists needed to explain why some of the military campaigns that were pre-emptive and aggressive in nature were also righteous. To save the people’s lives from tyranny and protect or improve their welfare, some political philosophers such as Mencius and Xunzi acknowledged the righteousness and appropriateness of taking initiative and employing military forces. A leader who placed the well-being of the people as top priority would gain wide and wholehearted support from the people, which could be translated into invincible military power – and even achieved victory without going to war. A story told in the Mencius was frequently referred to as a model: If a ruler of a state is fond of benevolence, he will have no match All-under-Heaven. When his army advanced to the south, the northern barbarians complained; when his army advanced to the east, the western barbarians complained. They said, “Why does he make us last?” When King Wu [of the Zhou] waged punitive war against the Yin [Shang], he had three hundred chariots and three thousand brave warriors. The King told the Shang people, “Do not be afraid! I bring you peace. I am not to be an enemy of the people.” The people then bowed their heads like the toppling of a mountain. To launch a punitive war is to rectify. Everybody wishes to rectify himself. What use is fighting? (Yang 1984: 325) The prime aims of proponents of this line were certainly to regulate and direct the use of military violence in a moral way and to highlight the well-being of the people, although its 331

Wicky W.K. Tse

rhetorical function also served as a useful tool of political propaganda for those launching war in the name of the faceless multitude. In addition, since the ultimate aim of the righteous leader to wage a war was to stop all wars and therefore save the people from suffering, military violence was therefore regarded as a means or tool, though not ideal, to achieve long-lasting peace. The Book of Lord Shang, among others, tells its reader ‘in order to eradicate war with war, even waging war is permissible; to eradicate murder with murder, even murdering is permissible; to eradicate punishments with punishments, even making punishments is permissible’ (Pines 2017: 215). Hence, a noble intention to use military violence would legitimatize and justify the actions and allow the temporary suffering. It was in fact vague to delineate what was in the interest of the common people, but with the birth of the empire, a new meaning of serving the common good emerged. The establishment of the first empire under Qin, though short-lived, realized the ideal of a unified empire and provided a justification for war, aggressive and defensive, accordingly. Wars for rulers’ ambition or self-interest were criticized, but not those in the name or for the sake of the empire – the common good. After the downfall of Qin, the fact that the two Han dynasties lasted in total for four hundred years greatly strengthened and fostered the idea of a unified empire. The pursuit of protecting, maintaining, and restoring – after rupture – a unified empire became a legitimate cause of both external and domestic military actions. Liu Xiu, the founding emperor of the Latter Han, for instance, made use of the restoration of the Han regime and the re-unification of the empire to win support and legitimatize his mandate. Furthermore, the moral principles of benefiting the people and filial piety also produced causes for the emperor waging wars abroad. A famous example was that Emperor Wu of Former Han launched a series of expeditions against the Xiongnu with a publicly manifested reason, among others, of avenging the humiliating defeat that the steppe power inflicted on the dynastic founder, who was also his great-grandfather. In sum, the early Chinese had already recognized that a world without war might be a noble dream, not to be easily achieved, if not unrealistic. Instead, preparing for war was emphasized and practised. As a consequence, different theorists and pragmatic statesmen proposed notions of legitimatizing the use of military violence and tried to produce a cover-all justification for war.17

Acknowledgement The author would like to thank the Research Grants Council of Hong Kong for providing financial support via the General Research Fund (GRF) scheme to the writing of this chapter (project no. 15609617).

Notes 1 A general overview of war in early China before this chapter is Yates 1999: 7–45. 2 Edward Shaughnessy, however, suggests that the line only indirectly concerned ‘war’, since the word rong 戎 in the context refers to the ‘war sacrifices’ held before the troops set off. See Shaughnessy 1996: 159. 3 Though it is in doubt if the Zhou king really gave such a speech on site given the practical difficulties in making himself heard by a large army, scholars generally accepted that the speech reveals certain contemporary information of the Shang-Zhou transition; see Tu 1992: 311–30. 4 In fact, not only the Zhou army but also their enemies, such as the Xianyun, employed chariots in warfare. See Li 2006: 144. 5 The number of chariots that the Spring and Autumn states could field increased after the Battle of Chengpu, but the 700-chariot had become a benchmark for the Jin army thereafter. In 589 bce, for instance, when Jin was going to war with Qi, the Jin commander asked for 800 chariots instead of

332

Warfare the originally given 700. The commander explained that 700 was the number used in the Battle of Chengpu when the Jin leaders were the most capable and brilliant cohort; 100 more were therefore needed to compensate for the inferiority of the present incumbent (Yang 2000: 789). 6 It is also a widely discussed question in world history. Three major views have been put forward; the first suggests that the chariot was used to transport the warriors to the battlefield, where they dismount and then fight on foot; the second takes the chariot as a mobile missile platform that the charioteers drove into range of the enemy to shoot arrows at the target, but saw no close combat; the third proposes that the charioteers would drive directly to the enemy and engage in close combat with long spears or lances. See Archer 2010: 58–61 and Roland 2016: 16–17. Among Chinese historians, some have tried to provide a hybrid explanation of the use of chariot in early Chinese warfare. Edward Shaughnessy, for example, suggested that the role of the chariot in battle evolved from a mobile command platform in the Shang times to a tactical weapon employed by the Zhou people, who therefore gained decisive military advantage over the Shang troops. See Shaughnessy 1988: 212–227. Herrlee G. Creel, however, asserted that the chariot mainly assumed a symbolic role and served as a command post, whereas the real fighting was done and won by the infantry. See Creel 1970: 262–76. 7 The role of the Spring and Autumn foot soldiers in battle is not very clear. Some scholars have made admirable attempts to elucidate this matter, but there is not sufficient evidence in the sources to say anything concrete and definitive. See Creel 1970: 284–93 and Lan 1979. 8 Raimund Kolb has calculated theoretically the space needed for chariot warfare. He found that, for instance, as the Jin army fielded 700 chariots in the Battle of Chengpu and if the opposing Chu followed suit, given that a chariot occupied about 16 square metres, then the two sides would need a field of battle at least 22.4 kilometres in size just to array the 1,400 chariots. See Kolb, 1991: 185. 9 Some scholars of military technology place the alternative dominance of mounted (chariot and cavalry) and infantry warriors in warfare – especially in the West – as a recurring mounted vs. infantry cycle that took place in world history. See Roland 2016: 20–30. 10 Chariotry and cavalry are always put in a polarity that they are exclusive or antagonistic to each other; that means the latter replaced the former in ancient warfare. Instead of viewing them as a separate development, I am in line with Robin Archer’s proposition that cavalry evolved directly from chariotry, and the early development of cavalry was actually an attempt to develop a ‘rough terrain chariot’. See Archer 2010: 66–71. 11 For the early relationship between trousers and horseback riding, see Beck 2014. 12 The late Peter Connolly, a scholar of ancient military equipment, conducted a practical experiment and proved that even in the absence of stirrup, a professional horseman could use a lance or long spear effectively and hit a target with 100% success in a certain way. See Connolly 2000: 103–12, especially 107–9. 13 Some Chinese researchers believe that the Shang state was capable of mobilizing a huge army of 170,000 men under the last Shang king; for example, see Liu 1992: 43, 49. But the figure seems too large to be acceptable given the contemporaneous demographic and technological conditions. 14 But Keightley also pointed out that ‘the men who led the armies were not purely military officers; various officers served a variety of roles, both civilian and military’ (Keightley 2012: 176). 15 The Qin’s counting of enemy heads as metrics of personal military achievement was not unique or novel in world history, as similar stories could be found in other civilizations; for more examples beyond Chinese cases, see Larson 2014. 16 There are two recent overviews of the military affairs of the Han empires; see Loewe 2009: 65–89 and de Crespigny 2009: 90–111. 17 A recent discussion of just war in different Chinese intellectual traditions, especially in the pre-imperial era, can be found in Lo 2015.

Works cited Archer, R. (2010) ‘Chariotry to cavalry: developments in the early first millennium’, in Garrett Fagan and Matthew Trundle (eds.) New Perspectives on Ancient Warfare, Leiden: Brill: 57–79. Beck, U., Mayke Wagner, Xiao Li, Desmond Durkin-Meisterernst, and Pavel E. Tarasov (2014) ‘The invention of trousers and its likely affiliation with horseback riding and mobility: a case study of late 2nd millennium bc finds from Turfan in eastern Central Asia’, in Quaternary International, 348: http://dx.doi. org/10.1016/j.quaint.2014.04.056 Bulliet, R.W. (2016) The Wheel: Inventions & Reinventions, New York: Columbia University Press.

333

Wicky W.K. Tse Chang, Kwang-chih (2005) ‘The rise of kings and the formation of city-states’, in Kwang-chih Chang, Xu Pingfang et al. (eds.) The Formation of Chinese Civilization: An Archaeological Perspective, New Haven:Yale University Press: 125–40. Connolly, P. (2000) ‘Experiments with the Sarissa – The Macedonian pike and cavalry lance – A functional view’, in Journal of Roman Military Equipment Studies, 11: 103–12. Creel, H.G. (1935) ‘Soldier and Scholar in Ancient China’, in Pacific Affairs, 8.3: 336–43. Creel, H.G. (1970) The Origins of Statecraft in China, Volume One: The Western Chou Empire, Chicago: University of Chicago Press. de Crespigny, R. (2009) ‘The military culture of Latter Han’, in N. Di Cosmo (ed.) Military Culture in Imperial China, Cambridge, MA: Harvard University Press: 90–111. Durrant, S., Wai-yee Li, and David Schaberg (trans) (2016) Zuo Tradition, Zuozhuan: Commentary on the “Spring and Autumn Annals”, Seattle and London: University of Washington Press. Dreyer, E.L. (2008) ‘Zhao Chongguo: A professional soldier of the Former Han dynasty’, in The Journal of Military History, 72: 665–725. Echevarria, A.J. (2017) Military Strategy: A Very Short Introduction, Oxford: Oxford University Press. Ferguson, R. (2003) ‘The birth of war’, in Natural History, 112: 28–34. Fiskesjö, M. (2001), ‘Rising from blood-stained fields: royal hunting and state formation in Shang China’, in Bulletin of the Museum of Far Eastern Antiquities, 73: 48–192. Graff, D.A. (2002) Medieval Chinese Warfare, 300–900, London: Routledge. Graff, D.A. (2009) ‘Narrative maneuvers: the representation of battle in Tang historical writing’, in N. Di Cosmo (ed.) Military Culture in Imperial China, Cambridge, MA: Harvard University Press: 143–64. Johnston, A.I. (1995) Cultural Realism: Strategic Culture and Grand Strategy in Chinese History, Princeton: Princeton University Press. Keightley, D.N. (2012) Working for His Majesty: Research Notes on Labor Mobilization in Late Shang China (ca. 1200–1045 B.C.), as Seen in the Oracle-Bone Inscriptions, with Particular Attention to Handicraft Industries, Agriculture,Warfare, Hunting, Construction, and the Shang’s Legacies, Berkeley: Institute of East Asian Studies, University of California. Kierman, F.A., Jr. (1974) ‘Phases and modes of combat in early China’, in F.A. Kierman, Jr. and J. K. Fairbank (eds.) Chinese Ways in Warfare, Cambridge, MA: Harvard University Press: 27–66. Kolb, R.T. (1991) Die Infanterie im Alten China-Ein Beitrag zur Militärgeschichte der Vor-Zhanguo-Zeit, Mainz: Zabern. Lan,Yongwei 藍永蔚 (1979) Chunqiu shiqi de bubing 春秋時期的步兵, Beijing: Zhonghua shuju. Larson, F. (2014) Severed: A History of Heads Lost and Heads Found, London: Granta. Lau, D.C. and Ames, R. (2003) Sun Pin:The Art of Warfare, New York: Ballantine Books: 114–15 and 179. Legge, J. (2000) The Chinese Classics Vol. III: The Shoo King or the Book of Historical Documents, Taipei: SMC Publishing INC. Lewis, M.E. (1990) Sanctioned Violence in Early China, Albany: State University of New York Press. Lewis, M.E. (2000) ‘The Han abolition of universal military service’, in Hans van de Ven (ed.) Warfare in Chinese History, Leiden: Brill: 33–75. Lewis, M.E. (2005) ‘Writings on warfare found in ancient Chinese tombs’, in Sino-Platonic Papers, 158: 1–17. Lewis, M.E. (2006) ‘The Just War in early China’, in T. Brekke (ed.) The Ethics of War in Asian Civilization – A Comparative Perspective, London: Routledge: 185–200. Li, Feng (2006) Landscape and Power in Early China: The Crisis and Fall of the Western Zhou 1045–771 bc, Cambridge: Cambridge University Press. Li, Feng (2008) Bureaucracy and the State in Early China: Governing the Western Zhou, Cambridge: Cambridge University Press. Li, Xueqin 李學勤 (ed.) (1997) Zhongguo gudai wenming yu guojia xingcheng yanjiu 中國古代文明與國家 形成研究, Kunming:Yunnan renmin chubanshe. Liu, Zhan 劉展 (ed.) (1992) Zhongguo gudai junzhi shi 中國古代軍制史, Beijing: Junshi kexue. Lo, Ping-Cheung and S.B. Twiss (eds.) (2015) Chinese Just War Ethics: Origin, Development, and Dissent, London: Routledge. Loewe, M. (1974) ‘The campaigns of Han Wu-ti’, in F. A. Kierman Jr. and J. K. Fairbank (eds.) Chinese Ways in Warfare, Cambridge, MA: Harvard University Press: 67–118. Loewe, M. (1994) ‘The chüeh-ti games: a reenactment of the battle between Ch'ih-yu and Hsüan-yüan’, in M. Loewe (ed.) Divination, Mythology and Monarchy in Han China, Cambridge: Cambridge University Press: 236–48.

334

Warfare Loewe, M. (2009) ‘The Western Han army: organization, leadership, and operation’, in Nicola Di Cosmo (ed.) Military Culture in Imperial China, Cambridge, MA: Harvard University Press: 65–89. Lorge, P. (2012) Chinese Martial Arts: From Antiquity to the Twenty-First Century, New York: Cambridge University Press. Lu, Liancheng (1993) ‘Chariot and horse burials in ancient China’, in Antiquity 67: 824–38. Lu, Liancheng (2005) ‘The Eastern Zhou and the growth of regionalism’, in Kwang-chih Chang, Xu Pingfang et al. (eds.) The Formation of Chinese Civilization: An Archaeological Perspective, New Haven:Yale University Press: 200–47. Needham, J. and R.D.S.Yates (1994) Science and Civilization in China, vol. 5: Chemistry and Chemical Technology, part 6, Military Technology: Missiles and Sieges, Cambridge: Cambridge University Press. Pines,Y. (trans. and ed.) (2017) The Book of Lord Shang: Apologetics of State Power in Early China, New York: Columbia University Press. Roland, A. (2016) War and Technology: A Very Short Introduction, Oxford: Oxford University Press. Sawyer, R. (1993) The Seven Military Classics of Ancient China, Boulder: Westview Press. Shaughnessy, E.L. (1988) ‘Historical perspective on the introduction of the chariot into China’, in Harvard Journal of Asiatic Studies, 48: 189–237. Shaughnessy, E.L. (1996) ‘Military histories of early China: a review article’, in Early China 21: 59–182. Trousdale, W. (1975) The Long Sword and Scabbard Slide in Asia, Washington, DC: Smithsonian Institution Press. Tse, W.W.K. (2017) ‘Cutting the enemy’s line of supply – the rise of the tactic and its use in early Chinese warfare’, in Journal of Chinese Military History 6.2: 131–156. Tu, Cheng-sheng 杜正勝 (1990) Bianhu qimin: chuantong zhengzhi shehui jiegou zhi xingcheng 編戶齊民： 傳統政治社會結構之形成, Taipei: Lianjing chuban shiye gongsi. Tu, Cheng-sheng 杜正勝 (1992) Gudai shehui yu guojia 古代社會與國家,Taipei:Yunchen wenhua chuban. Wu, Hsiao-yun. (2013) Chariots in Early China: Origins, Cultural Interactions, and Identity, Oxford:Archaeopress. Wu, Shu-hui (2012) ‘Fighting for His Majesty (I)’, in Journal of Chinese Military History, 1: 24–60. Xu, Hong 許宏 (2014) Heyi Zhongguo: gongyuanqian 2000 nian de zhongyuan tujing 何以中國：公元前 2000年的中原圖景, Beijing: Sanlian shudian. Yang, Bojun 楊伯峻 (1984) Mengzi yizhu 孟子譯注, Beijing: Zhonghua shuju. Yang, Bojun 楊伯峻 (2000) Chunqiu Zuozhuan zhu 春秋左傳注, Beijing: Zhonghua shuju. Yates, R.D.S. (1988) ‘New light on the ancient Chinese military texts: notes on the evolution and the development of military specialization in Warring States China’, in T’oung Pao, 74: 211–48. Yates, R.D.S. (1999) ‘Early China’, in K. Raaflaub and N. Rosenstein (eds.) War and Society in the Ancient and Medieval Worlds: Asia, the Mediterranean, Europe, and Mesoamerica, Washington, DC: Center for Hellenic Studies, Trustees for Harvard University: 7–45. Yates, R.D.S. (2002) ‘The horse in early Chinese military history’, in Huang Ko-wu (ed.) Junshi zuzhi yu zhanzheng 軍事組織與戰爭 (Military Organization and War: Papers from the Third International Conference on Sinology, History Section), Taipei: Institute of Modern History, Academia Sinica: 1–78. Yates, R.D.S. (2005) ‘Early modes of interpretation of the military canons: the case of the Sunzi bingfa’, in Ching-I Tu (ed.) Interpretation and Intellectual Change: Chinese Hermeneutics in Historical Perspective, New Brunswick, NJ: Transaction Publishers: 65–79. Yates, R.D.S. (2011) ‘Soldiers, scribes, and women: literacy among the lower orders in early China’, in Li Feng and D. P. Banner (eds.) Writing and Literacy in Early China: Studies from the Columbia Early China Seminar, Seattle: University of Washington Press: 339–69. Yuan, Jing and Rowan Flad. (2003) ‘Two issues concerning ancient domesticated horses in China’, in Bulletin of the Museum of Far Eastern Antiquities 75: 110–126.

335

16 CURRENCY

FRANÇOIS THIERRYCURRENCY

François Thierry

The traditional Chinese monetary system is based on the idea that money is an instrument similar to others, definable by its function, i.e. to exchange goods and services, and hence to be used as a measure of value (Thierry 2001, 2002). Scholars have established a list of objects mentioned in the early times in oracle-bones inscriptions, in bronze ritual vessels and in ancient literature, indicating that these objects or matters had acquired a particular value which made them objects for exchange: sea-cowries, tortoise-shells, pearls, jade stones, skins, horns, food grains, textiles and stone tools for hunting (knives, axes) or for farming (hoes, spades, sickles). Among these natural or man-made products, grain, textiles and tools persisted and continued over a long period to serve as money, or played by default the role of money.

Birth of money Origin of the monetary system: sea-cowries The first object for which the monetary use is clearly established is the sea-cowrie (bei 貝, Figure 16.1). During the last period of the Shang dynasty (twelfth–eleventh century bc) and especially during the Western Zhou (1045–771 bc), as seen on the bronze inscriptions, cowries were used in the form of the double-string (peng 朋). Analysis of the character peng shows two strings (xi 系) linked together by a small knot. The number of cowries in each peng is not clearly established: scholars have different opinions, and there is no archaeological evidence. In a Western Zhou tomb excavated in 1981 at Jiangjia village, near Fufeng (Shaanxi), archaeologists found a peng consisting of two strings xi of which the knot was decorated with two stone beads; each xi counts five cowries (Luo 1985: 2, 9; Peng 2007: 11–12).1 Divinatory inscriptions on bones or tortoise shells as well as on bronze vessels of the Shang (ca. 1600–1050 bc) and Zhou (1045–256 bc) dynasties regularly mention “gifts of cowries”, “putting cowries in the treasure”, “seizure of cowries”, “use of cowries” or “rewards of double-strings of cowries”. At first, rewards of double-strings were, for the king, more or less a way of rewarding in return the faithfulness of lords and vassals. But the role of cowries was not restricted to rewarding loyalty or meritorious acts; it also extends to the payment of services and the measure of value (“general equivalent”), as mentioned

336

Currency

Figure 16.1 Sea-cowrie Collection Wanxuanzhai 2151

on bronze inscriptions. On a bronze vessel dating from King Cheng of the Zhou (1042/35– 1006 bc), we can read: “Huan, earl of Ju, made this precious zun vessel using ten and three peng of cowries”; on another zun of the same period, we find: “He was rewarded with thirty peng of cowries that he used to make this [?] precious zun vessel” (Wei 1936: 27; Wang 1951: 64; Wu Zhenfeng 1987: 4; Liu-Chen-Dong 1989: 73–76; Thierry 1997: 40); thus cowries were used to pay the craftsman for his work, the metal and the casting materials; double-strings were rewards and gifts, but also – even since the early Zhou – the means to pay the cost of manufacturing ritual vessels (Wei: 23–29; Zhu 1991: I, 7). The most important inscription for monetary history and economic history is that of the bronze vessel known as the he of Qiu Weì 裘衛盉, dating to the third year of King Gong of the Zhou (915 bc)2 : The commoner Ju Bai received from Qiu Wèi a jade tablet which was worth 80 peng in exchange for ten tian from his estate; Ju also received two pieces of red coral, a pair of kid knee-caps and a parade belt, all worth 20 peng, in exchange for three tian from his estate. (Tang 1976; Wu Zhenfeng 1987: 13–14; Liu-Chen-Dong 1989: 108–111; Ma 1986: 128–130, 132–133; Huang 2001: 37) In this case, the double-string is a measure of value, a general equivalent, because there is an exchange of objects for land, and the respective value of the goods is based on the peng as a unit of account; this deal has no monetary nature because nobody use a monetary instrument for the exchange. Moreover, the text gives cai bashi peng 才[財]八十朋, “worth 80 peng”, and cai nian peng 才[財]廿朋, “worth 20 peng”, and not the classical formula bei bashi peng 貝八十朋, “80 peng in cowries”, or bei nian peng 貝廿朋, “20 peng in cowries”. Thanks to this fundamental text, we can

337

François Thierry

understand that, at least in the second half of the tenth century bc, the Chinese had “invented” currency as a means of exchange and general equivalent, although they did not cast or strike money. There are some quotations about the use of cowries during the Shang and Zhou dynasties in literary sources, but these texts are too late to be considered as contemporary evidence: in the Yijing, dating back to the early Western Zhou; in the Shijing, back to before 600 bc; in one of the parts of the Shujing considered as dating from the early Western Zhou (1045–771 bc); in a later text, the Liji, there is frequent use of the expression huobei 貨貝,“money-cowries”. From the bronze inscriptions and literature, we know that sea-cowries, probably under the Shang already and certainly under the Western Zhou, were a form of real money: they were used to reward loyalty and to exchange goods, but also to pay for services and labour and to measure the value of goods. Archaeological studies confirm the monetary role of the sea-cowries. Numerous discoveries have been made all over in China in sites of different periods. In the tombs dating from the first period of the Shang (ca. 1600–1300 bc) excavated at Anyang 安陽 (Henan), relatively few sea-cowries have been found, but in some other sites of the early Shang, cowries were found in greater number. As demonstrated by the Yinxu 殷墟 excavations (Henan) and by discoveries made in sites of other regions in China, cowries become most numerous in the graves of the late Shang (Dai 1983: 74; Erlitou: 262–263; Wang 1988: 1142–1143; Zhu 1983: 12–13; Wang 1965: 26). Precise studies of these finds have shown that during that period (ca. 1300–1050 bc), the back of cowrie shells was pierced; the holes pierced at the end of the shell are more or less large; from the late Shang tombs there begin to appear, albeit in small quantity, cowries completely open on the back. This open-back type is the most numerous in the Western Zhou tombs and then in the Eastern Zhou tombs (770–256). Cowries were very common during the Western Zhou and their number increased further during the Eastern Zhou. The distribution of findings coincides with the progressive expansion of Chinese territories during the Zhou: cowries have been found in the regions belonging to the royal House of the Zhou, to the feudal states of Qi 齊, Lu 魯,Yan 燕, Jin 晉, Zheng 鄭, Weì 衛, Qin 秦, Chu 楚, Cai 蔡, Ceng 曾 and Wu 吳, as well as in the territories of the neighbouring barbarian confederacies such as Xianyu 鮮虞, Shanrong 山戎, Donghu 東胡, Di 翟, and Rong 戎 (Zhang 1991: 28; Zhu 1991: I, 7–9; Shao 1993: 4). The geographical distribution and the chronology of the numerous finds from elsewhere in China clearly show that – especially from the middle of the Spring and Autumn period (770–481 bc) – the increase of cowries in the tombs is correlated with the declining role of the cowries as currency. The majority of these cowries are cypraea moneta.3 There was much questioning concerning the provenance of the cowries found in China, but now two zones are considered the probable suppliers in cowries: the coasts and islands of the East China Sea (present Jiangsu and Zhejiang, Japan, Ryu-Kyu and Formosa) and the shores of the South China Sea (present Guangdong, Indochina, the Malaysian peninsula and the Philippines). It is certain that cowries came from eastern coasts through the intermediary of barbarian peoples: numerous bronze inscriptions tell of military expeditions against barbarians of the east and southeast – especially that of the Huai region – who possessed cowries and refused to give them for nothing. The eastern origin is also suggested in several ancient documents: in both bronze inscriptions and the Bamboo Annals, there is mention of cowries tribute from vassals settled in the Huai and the Chu regions (Jiangsu, Zhejiang, Anhui); as clearly shown by the bronze inscriptions, most of the feudal states procured cowries by war, by looting directly in the treasures of their neighbouring lords. The long practical experience of the use of cowries during many centuries is the origin of practices and ideas on which is based the Chinese notion of money: first, because cowries of different sizes have been used for the same value, the value of the monetary instrument is not

338

Currency

connected to its size or its weight; second, the increase of the exchange value is made by the increase of the number of monetary instruments and/or of the account units, not by a change of the monetary instrument; and lastly, there is no state guarantee, since the monetary instrument (i.e. the cowrie) is a natural object without mark or stamp. It follows a series of corollaries: clear lack of a monopoly in coin casting and issuing, complete or partial free-coining, private controls, use of the coins one by one or in string, and a habit of changing the material for the monetary instrument. Thus, the Chinese invented fiat money (fiduciary money), since the acceptance of both the monetary instrument (“currency”) and its exchange value by the different economic players depends on a tacit agreement between members of the society from the king to the poorest farmer and not on the actual value of the material of the monetary instrument.

A revolution: men are able to produce money The development of a more productive and more diversified economy under the Western Zhou brought with it an expansion of the trade and a massive increase in means of payment; the main – and paradoxical – consequence was a progressive devaluation of the cowries, as clearly shown by the bronze inscriptions: under the Shang dynasty and the early Western Zhou, rewards were generally of the order of one or two double-strings; a reward of over twenty peng was quite exceptional; but from the reign of King Mu 穆 (956–918 bc) onward, we can observe a regular increase of the number of peng up to the amount of thirty peng, and even fifty peng. The consequence of the growth of demand for means of payment was the production of imitation cowries and the adoption of other materials for money making. Imitation cowries were made in various materials, as bone, horn, nacre, stone, clay, lead and bronze.This production expanded rapidly only from the Zhou. The nacre cowries are relatively few in number; the main discoveries were mostly unearthed in the vicinity of Luoyang (Henan), in Hebei and Anhui provinces. On the other hand, there are a great variety of bone cowries, usually sculpted from bones of cattle. The monetary use of these bone cowries is not certain, and it is highly probable that they were used mainly for burial and adornment purposes with the goal of avoiding the immobilization of sea-cowries in the graves. The findings distribution shows that these cowries were being made in all the feudal states during the Zhou period, and that these bone cowries, alongside seacowries, stone and bronze cowries, were used as tomb deposits. Stone cowries probably appeared at the same time as bone cowries: there are various types of stone cowries, some of which imitate quite perfectly, on both the front and back, the original sea-cowries. Imitation cowries in jade or in gold (or gold covered) never served as money for circulation and were only one of the forms of hoarding (reserve of values): in China, contrary to the western classical theories of money, hoarding is not one of the functions of money. The most important imitation cowrie is the bronze cowrie (Figure 16.2), the use of which constitutes a significant step in Chinese monetary history. For the first time, a man-made material was used to make monetary instruments, and it was possible to reuse the raw material to recast monetary instruments. The first bronze cowries can be dated from the late period of the Shang dynasty, known as Shang-Yin: five bronze cowries have been unearthed in the site of Yin metropolis, and 109 bronze cowries have been discovered in a late Shang grave excavated near Baode 保德 in Shanxi province (Zhu 1985: 5–6; Thierry 1997: 46). But it is only from the Eastern Zhou (770–256 bc) onwards that the manufacture of bronze cowries grew enormously. The total amount of bronze cowries recorded by Chinese archaeologists is over 5000. The most important discoveries include that of Huixian 輝縣 in Henan; that of Shangma 上

339

François Thierry

Figure 16.2  Bronze cowrie, dukedom of Jin (seventh–sixth c. bc). Bronze, 23 mm, 1,66 g Collection Wanxuanzhai 2151

馬, near Houma 侯馬 in Shanxi; and that of Linqian 林前 in Shandong (Zhu 1985: 16–17). The dukedom of Jin 晉 occupies a special place in the development of this kind of money: for since 1949, 80% of the bronze cowries found in China have been unearthed in Shanxi province, which corresponds roughly to the center of the Jin state. At Houma, in this metropolitan zone of Jin, which lies at the confluence of the Fen and Kuai Rivers, Chinese archaeologists discovered in 1959–1961 a bronze workshop where vessels, tools and weapons as well cowries were cast in clay moulds. According to archaeological evidence, this workshop dates from a period from the late seventh to the early sixth century bc. The features of the moulds for bronze cowries are already very close to the two types of cash moulds of the Qin and Han periods: on the first type, the cowries were placed to the right and the left of a central canal into which the bronze was poured, on the second one the cowries were arranged in vertical lines. For the population of the Great Plain in north China, the main source of food and wealth was farming. In this activity, hoes and spades (bu 布, bo 鎛, chan 鏟 or qian 錢), were indispensable instruments; for this reason they acquired a special value in exchanges: progressively spades were used both as exchange instruments and as farming tools. Afterward, following the model of real spades, others were cast solely for the purpose of exchange. These spades which one can consider monetary instruments were still very close to the hoes of the late Shang and of the early Western Zhou, and it is sometime difficult to tell the difference between farming tools and exchange instruments: they are of large size, with round or flat shoulders, and retained the socket to accommodate a wooden handle (Chen 1982; Wang 1988: nos. 23–29; Shanxi 1989: nos. 1–2). Most of them do not bear inscriptions, but there are several types dated from the early Spring and Autumn (eighth century bc) with characters such as peng, shan “mountain”, or numerals; only one toponym is known, that of the Lu clan, Lu shi 盧氏 (Wang 1988: no. 29). Primitive spades have been unearthed from Shang tombs and in several Shang-Yin period sites, such as Anyang, Luoyang, near Zhengzhou and in southern Shanxi. There have also been more numerous discoveries in the same regions but in sites and tombs dating from the Western Zhou and early Eastern Zhou. 340

Currency

From that time onward, people understood that it is possible to manufacture money and to cast it in metal in the same way as for other daily tools. In China, since the main technique to produce metal vessels, tools, instruments or weapons was by casting them in moulds, thus money would be made by casting.

Ancient spades and ancient knives From the Eastern Zhou onwards, and particularly during the second period of the Spring and Autumn period (sixth and early fifth centuries bc), the development of large trade and economic activities necessitated an increase in means of payment and stimulated a great change in the monetary domain already perceptible with the manufacturing of bronze cowries. New forms of money, hollow-handled spades and pointed knives, appeared and were both cast in great quantity. The “hollow-handled spades” (kongshoubu 空首布), named in this way because they retain a socket to accommodate the wooden handle, are merely a means of exchange: they are very thin, and the bronze alloy used for casting is too brittle.4 These spades are usually classified in three main groups according to their general shape. About 210 different inscriptions have been recorded by epigraphists; many of them are fief names or identifiable toponyms (Dai-Dai 2014). The most common are the “flat-shoulder hollow-handled spades” (Figure 16.3); they are 9–10 cm high and weigh between 25 and 35 g (Wang 1988: nos 32–566); they usually have a

Figure 16.3  Hollow-handled spade, Henan (sixth c. bc). Bronze, 101 mm, 31,69 g Zeno-Oriental Coins Database n° 125311

341

François Thierry

single character, very occasionally two. Epigraphists have recorded over 180 inscriptions, comprising numerals, cyclical characters, words associated with goods and wealth (e.g. “sheep”, “richness”, peng, “cowrie”, “value”, “grain”), names of places, cities, clans or fiefs (e.g. Zhou, Shang, Song, Wu, Zhu, Wu), or titles (e.g. “king”, “earl”, “duke”, “marquis”, officer”), and several other characters which have yet to be identified or deciphered. This is the earliest type and can be dated from the second half of the seventh to the mid-sixth century bc. Archaeological discoveries and spade inscriptions demonstrate that the region in which they were cast and the circulation era correspond to the royal domain of the Zhou, and to Central China fiefs, mainly the dukedoms of the Jin, Zheng, Song and Wèi. The flat-shoulder spades of the second type are smaller (7 cm, ca. 19 g) and bears various inscriptions such as Wu 武, An cang 安藏, An Zhou 安周 and Dong Zhou 東周 (Wang 1988: nos 616–680). These are later and date from the early Warring States period (late fifth–early fourth century bc), as confirmed by the archaeological context of the finds in the vicinity of modern Luoyang (Henan). The second group consists of the so called “slanting-shoulder spades” (Wang 1988: nos 567– 615). The spots where they were found, as well as the toponyms, allow us to attribute this currency to fiefs of southern central China (present-day Henan) as Wu, Lushi or Sanchuan, vassals of the Zhou royal house and then of the dukedoms of Jin and Zheng; they are dated from the middle and late part of the Spring and Autumn period (end of the sixth, fifth and early fourth centuries bc). There is at least a third type known as “pointed-shoulder and pointed-feet hollow-handled spades” (Figure 16.4); some of these large spades have an inscription of one character, rarely more, while others do not bear a single character (Wang 1988: nos 681–712). The main discoveries of these spades took place in three regions: Western Hebei, northeast of Henan and the lower valley of the Fen River (Shanxi). They provided important evidence for the chronology and attribution of this type. Near Houma (Shanxi), between 1959 and 1961, pointed-shoulder and pointed-feet spades have been found in the ruins of the workshop where they had been cast (Zhang 1960: 12–14; Houma 1962: 58–60). The discovery of Houma is particularly important because the excavation site corresponds with the ancient city of Xintian 新田, where Duke Jing of Jin installed his new capital in 585 bc. The Houma workshop, the official foundry under the control of officials at the ducal palace, provides us with important information about the monetary production status in sixth-century China: considered a manufactured product, as well as the other instruments, tools or weapons, money was cast in the palace foundries in the same way and by the same workers as for vessels, tools and objects of daily life. On the other hand, this discovery provides a very important chronological step, allowing a fairly certain dating of these spades from the early sixth century bc. The distribution of other finds of pointed-shoulders\ and pointed-feet spades confirms that this type of currency is characteristic of Jin dukedom. As the feudal principalities of the Central Plain began to use spade-money, a similar process took place in northern and northeastern China among the nomadic and sedentary shepherds, hunters and fishing populations. As essential tools to cut up cattle, big game or fish and to scrape skins, knives played a special role in the exchanges. In north and east China, i.e. from the loop of the Yellow River to the Shandong peninsula, passing through southern Mongolia and Manchuria, knives used in daily life began to be used as means of exchange.There is a clear link between the shape of South-Siberian knives and that of knives discovered in Shang tombs at Anyang: both are curved knives, blade and handle made in one piece, handle with three or more grooves and a final ring, sometimes zoomorphic, sometimes simple. Mutatis mutandis, this shape persisted down to the last period of the Zhou dynasty in the third century bc. From the sixth and fifth century, the main northern and northeastern states, i.e.Yan, Qi, Zhongshan 中山 and even Zhao 342

Currency

Figure 16.4  Hollow-handled spade, dukedom of Jin (sixth c. bc). Bronze, 146 mm, 37 g Courtesy Numismatic Guaranty Corporation, Auction 96 no 1074

趙, as well as the barbarian confederacies of the Shanrong, Rong and Di peoples, used “pointed knives” jianshoudao 尖首刀 (Figure 16.5), which, like the contemporary hollow-handled spades, bore inscriptions; but the variety of inscriptions is clearly lesser and consists mainly of numerals, cyclical characters and words such as “sheep”, “fish” and so on (Wang 1988: nos 2662–2873). All these coins are fiduciary money because their exchange value is not determined by their metallic composition nor by their weight: the element that matters is the size that must be somewhat the same. For example, the 111 “flat-shoulder” kongshoubu of the Shanghai Museum collection have quite regular sizes, 90 millimeters tall and 50 millimeters wide, but their weights are very variable and range from 22,2 grams to 39 grams, i.e. they vary by twice as much (Shanghai 1994–95: I, nos 6–117). The copper content of kongshoubu analysed by Zhou Weirong ranges from 54,31% to 73,38% without a clear average copper content; the copper content of the fiftysix jianshoudao knives in the same study ranges from 21,68% to 62,36%, with an average group between 40% and 55% (Zhou 2004: 5–6, 17–20). Even if coinage grew very rapidly during the Spring and Autumn period, its use both geographically and socially remains low, and it is difficult to consider the actual economy a monetary economy; on the other hand, cowrie shells continue to play an important role in the exchanges. The acceleration of the monetarization process is only the result of the social, economic, political and intellectual changes at the extreme end of the Spring and Autumn and in the early Warring States periods (ca. 500–400 bc). Beside the workshops of the lords’ palaces, there appears a strata of entrepreneurs (e.g. mine operators, salt producers, ironmasters, copper 343

François Thierry

Figure 16.5  Pointed knife, northeast China (sixth c. bc). Bronze, 158,4 mm, 14,16 g (Zeno-Oriental Coins Database n° 41864)

founders) and big merchants who move around the Chinese world. And, likewise, the dukes and princes of the largest states attempt to climb to the power position of the Zhou king; the vassal earls and lords inside these great states rebel against their suzerain and sometimes succeed in supplanting him and dismembering his territory. Each lord, in his own domain, considers himself the master of men and goods, and he wants to control the money, because, as the Guanzi says, “the very lord who will take in his hand the balance between grain and money, will be able to conquer the world” (Guanzi: lxix, 351; Thierry 2001: 140–141).

The monetarization of the society The currency of the Warring States period The complexity and the variety of the coins of Warring States are the reflection of the complexity of the social structure of these different states: the kingdoms in which vassals are numerous and powerful have an abundant and heterogeneous coinage, whereas the kingdoms whose royal house is powerful have a quite uniform coinage, if not unified. It is advisable to consider separately the case of Chu, a kingdom that does not belong to the Chinese universe, strictly speaking, and that has a completely different monetary system based on the metal value (“intrinsic value”). 344

Currency

The Kingdoms of Yan, Zhao, Wei 魏 and Han 韓 covered broad territories, often tangled, and their boundaries were changing according to victories and defeats; the sharing of the former Dukedom of Jin (453 bc) by Zhao, Wei and Han was not done according to local or regional consistency but according to the feudal ties born between the different lords during the battle for power during the early fifth century bc and to the breaking up of fiefs from the vanquished feudal houses. For that reason, each state had in its own territory different feudal traditions: in monetary matters, some lords were following the Great Plain tradition and issued spades, others used knives; some were attached to the traditional use of the coins, while others were influenced by the monetary conceptions of Chu or Qin. Most of the lords of central China, vassals of the Zhou royal house and of the kingdoms born from the sharing of the former states of Jin,Wèi or Song, issued small spades called “flat-handled spades” (pingshoubu 平首布) according to their shape. Among these spades, square-feet spades are the most important group, on account of the large number of fiefs which made them as well as on account of their widespread circulation or the large quantity that was issued (Wang 1988: nos 1457–2345). Square-feet spades were issued by vassal fiefs of Zhao in the extreme north of actual China, as well as by the southernmost fiefs of Han and the fiefs of Wei; Chinese numismatists listed more than sixty fiefs which issued this type; the spades of Anyang 安陽 (Figure 16.6), Pingyang 平陽 and Zhaiyang 宅陽 were cast in enormous quantity. In the upper valley of the Fen River, in the area situated between there and the Yellow River, and in the eastern foothills of the Taihangshan Mountains 太行山, from the high valley of the Sanggan River to Handan (i.e. present-day northern Shanxi and western Hebei), special spades known as “pointed-feet spades” were issued (Wang 1988: nos 713–1212). These spades derived from the ancient “pointed-shoulder and pointed-feet spades”. The earlier ones are larger than the square-feet spades. But by a coinage process of homogenization, pointed-feet spades of the same size as the square-feet spades were issued; at first these small spades bore the inscription ban 半, “half ”, indicating their value against that of the big ones. About thirty fiefs are known as pointed-feet spades issuers: the most numerous are the spades of vassals of Zhao, Cishi 兹 氏, Pingzhou 平州, Dayin 大陰, Pingyang 平陽 and Jinyang 晉陽. Spades with round feet and round shoulders are quite rare; only five fiefs – Lin 藺, Lishi 蘺石, Cishi, Dayin and Jinyang – are known to have issued them (Wang 1988: nos 2346–2455). As these five fiefs are situated in the

Figure 16.6  Square-feet spade, Zhao, city of Anyang (fourth–third c. bc). Bronze, 47,2 mm, 5,66 g Collection Wanxuanzhai 2258

345

François Thierry

middle reaches of the Fen River and between this area and the Yellow River,“round-feet spades” are clearly a local money typology. These spades were cast in two sizes, large and small, which in some ways links them with the pointed-feet spades of the Fen valley. There are some extremely rare “round-feet spades” which are pierced with one round hole in the head and two in the feet, known as “three-holes spades” (Wang 1988: nos 2456–2489), and some square-feet spades have an enlarged head (Wang 1988: nos 1215–1244). With the exception of the earlier pingshoubu of the Fen valley, all these spades have the same typology: around 5 centimetres tall, weighing about 5 or 6 grams and with rims in relief on both obverse and reverse. On the obverse, two characters (more rarely three) give the name of the issuer, the fief or the town; there are three lines on the plain reverse, and sometimes we find an indication related to the casting, as a numeral.These coins are always cast in bronze in moulds, mainly in clay, engraved one by one with a stylus; for this the reason the writing is not homogenous and is related to local calligraphic tradition; but, even on spades of the same towns, for example Pingyang spades, we can see a great variety of calligraphy (Wang 1988: nos 1730–1798; Thierry 1997: nos 188–201). In terms of monetary circulation, the main characteristic of north and northeast China is the use of the knife-money deriving from the jianshoudao. The kingdom of Yan issued in great quantity Yi dao 易刀, “knives of Yi”, also known as mingdao 明刀 ,5 measuring about 14 cm in length, with the single character Yi 易 on the obverse (Figure 16.7); Yi refers to the name of the Yan kingdom metropolis.6 The main reverse inscriptions such as zuo 左, “left”, and you 右, “right”, wai 外, “exterior”, and nei 內, “interior”, are toponymics referring to the location of the mints relative to the Yan metropolis. Besides these inscriptions we often find numerals corresponding probably to a foundry or a furnace identification. Yi knives are the most common of the knife-money and were found in a vast geographical area stretching from Mongolia to Korea and to the ancient territory of Zhao in the south. It is likely that the Yi knives played in the north a similar role to that of the square-feet spades in the Central Plain. This currency was issued by the Yan royal house, and its abundance is the reflection of the power of this family, but there was no casting coin monopoly in Yan. Some fiefs, and particularly those of the north Hebei plain and the Liaoning plain (e.g. Yang’an, Fenghua, Pingyin and Yichang), had a different tradition and issued square-feet spades, the reverse of which bear

Figure 16.7  Yi knife of Yan, or Mingdao (fourth–third c. bc). Bronze, 138 mm, 16,05 g Collection Wanxuanzhai 300

346

Currency

inscriptions like that of the Yi knives, zuo and you. Likewise, although most of the vassal fiefs or cities of Zhao issued flat-head spades, some issued knives; some others issued firstly spades and then knives, or vice versa, according to the influence of neighbouring fiefs or the suzerain: Lin, vassal of Zhao, issued pointed-feet spades, square-feet spades and round-feet spades, and knives and round coins as well. Handan 邯鄲, the Zhao metropolis, issued both spades and knives. Spades and knives of that period (fourth-third c. bc) are very abundant, because in the Great Plain and in the Liaohe valley, feudal lords are very numerous and enjoyed a certain autonomy in relation to the suzerain. One of the consequences of their autonomy was the great variety of monetary inscriptions clearly identifiable with the name of feudal cities or clans. There did not exist a royal coinage in the same sense as the term is used in Western numismatics: fiefs whose traditional practice was the use of spade-money did not necessarily accept knife-money, even the knives issued by the royal city. All these coins, spades or knives, are mainly of a fiduciary nature. The Yan Yi dao knives, for example, are quite regular in size, around 140 millimeters in length, whereas their weights vary from 12,2 to 19,6 grams, i.e. a difference of more than a third, without a clearly established medium weight: most of them range “regularly” between 14 and 17 g (Shanghai 1994–95: I, nos 1095–1242). The metallic composition of these knives is very special and irregular too: the copper content in the metal alloy reaches something like 35 to 50% and the lead content often exceeds 40%; among the 133 Yi dao analyzed by Pr. Zhou Weirong et Dai Zhiqiang, 66% contain more lead than copper (Zhou-Dai 2002 : 309, 315; Zhou 2004: 12–17). Monetary deposits show that knives of other states were in circulation in the Yan kingdom along with the Yi dao: although their weight is lighter (from 10 to 13 g), “straight-knives” of Handan, the capital of Zhao, and knives of Chengbo 成白, a fief of the former kingdom of Zhongshan annexed by Zhao, were readily accepted because their length corresponds to that of the Yi dao, i.e. 140 millimeters. With regard to the pingshoubu, neither their weight nor their metallic composition determine their exchange value. When we consider, for example, the forty-one pointed-feet spades issued by the city of Cishi housed in the collection of Shanghai Museum, we can observe that their sizes are quite regular, but their weights range from 7 to 3,9 grams, with a medium weight of 5,6 g; only 22, about a half of the total, have a weight between 5 and 6 grammes (Shanghai 1994–95: I, nos 325–366). Likewise, analysis of the metal alloy of the pointed-feet spades demonstrates that they are, on an average, less rich in copper (47% copper, 45% lead and 4% tin) than the squarefeet spades (69% copper, 18% lead and 9% tin) (Zhou 2004: 6–8); nevertheless, all these spades were circulating together in the Great Plain, as proved by many monetary deposits (Guo 1965a; Guo 1965b; Wang 1984; Ban 1985; Li 2006: 50–73). The kingdom of Qi is a special case in Chinese monetary history: it is the only one to have set up a “national” monetary system based on the face value of the currency; compared to the royal house, the local feudal lords were too weak to be able to play any real role in the field of money.The most famous monetary instruments of Qi are the large knives known as Qi fahua 齊 法化; the majority of the scholars consider that the two last characters are fahua and mean “legal money”;7 this mention could be considered a confirmation that some control was exercised by the state authorities over the issue of money by means of a tax established on manufacturing (Shiji: xxxii: 1487). There are six main types of Qi knives, classified according to their inscriptions: Qi zhi fahua, Jimo zhi fahua, Jimo fahua, Anyang zhi fahua, Qi zao bang zhang fahua and Qi fahua (Figure 16.8). Some scholars dated these knives from the founding of the Qi dukedom in the eleventh century bc, but now it is well agreed that they were cast during the Tian 田 dynasty of Qi, in the late Warring States period (fourth–third centuries bc). The knives Qi zao bang zhang fahua, “Everlasting legal money of Qi at the establishment of State”, were issued to commemorate the inauguration of the Tian dynasty, which usurped the throne in 386 bc; this 347

François Thierry

Figure 16.8  Qi knife, Qi fahua (fourth–third c. bc). Bronze, 185 mm, 47,86 g Courtesy Compagnie Générale de Bourse

date corresponds to both archaeological evidence and comparative elements in terms of casting technology and epigraphy, and to historical sources. Thanks to historical and epigraphical data, the knives of the two cities of Anyang and Jimo could be dated from the middle of the fourth century to the end of Qi (221 bc). The Qi zhi fahua and Qi fahua, “legal money of Qi”, are difficult to date because they do not bear any information such as place name or historical event; according to archaeological results and to metallographic analysis, Qi zhi fahua are considered the earlier type and Qi fahua the later one (Thierry 1997: 125–129). All finds of Qi knives have been made within the present Shandong province, i.e. the former territory of the ancient Qi, this appears to indicate that the circulation area of this currency did not go beyond the kingdom’s borders, probably because Qi knives were a particular kind of currency which did not integrate with the monetary systems of the neighbouring states. Indeed, reverse inscriptions indicate that Qi had a consistent monetary system based on the peng as a unit: most of these knives bear the mention sanshi 三十, “30”, i.e. the face value equal to thirty double-strings of cowries. In the same way, the round coins of Qi, issued a little bit later, likewise bear their face value: yi liu hua 賹六化, “money of 6 yi”, yi si hua 賹四化, “money of 4 yi” (Figure 16.9), or yi hua 賹化, “money of a yi”. The archaic shape of the character yi consists of two parts, peng 朋, “double-string”, and bei 貝, “cowrie”: thus, the yi corresponds to a double-string of cowrie.The face value of both the knives and round coins is unrelated to their respective weights: the coins of a yi weigh around 2.5g, that of 4 yi around 5.5 g and that of 6 yi around 9 g; the knives of 30 yi weigh between 45 and 55 g. The so-called bronze used in casting contains a high level of 348

Currency

Figure 16.9  Yi si hua of Qi (fourth–third c. bc). Bronze, 29,3 mm, 6,55 g Collection Wanxuanzhai 1743

lead, on average 35% out of 56% of copper; but there is a big gap between the different copper percentages, from 42% to 63% (Zhou 2004: 20–21, 28; Shandong Numismatic Society 1996: 62, 96–97). Quite obviously these different coins are clearly fiat money.

Intrinsic value and its limits In the southwest of the Central Plain (now southwest Shanxi and northwest Henan), the vassal fiefs of Han, as well Wei or Zheng, had submitted to the Chu kingdom influence. Chu was the most southern of the Chinese kingdoms; it was also the state where the distinctive characteristics of the non-Chinese peoples (i.e. Man, Yi and Yue) strongly persisted and were politically assumed. In the monetary field, Chu had a unique system of its own: people used both bronze cowries and gold ingots as circulating currency. The bronze cowries, known as “ant noses”, yibi 蟻鼻, are small bulging oval plates bearing an inscription; the most common is bei, “cowrie”.This type of money circulated throughout the Chu Kingdom and its vassal states, such as Cai, Chen and especially Lu, where bronze moulds for yibi have been found. It is not known how and according to which system these coins were used; it seems that they were used as single units or according to capacity measures. An enormous quantity of ant-nose money was issued, and hoards of thousands of pieces are not uncommon (Thierry 1997: 143–146). The Chu kingdom later issued, only in some regions of the north, spades with the inscription 10 huo 十貨, “10 monetary units”. The distribution of finds shows that the circulation area of ant-nose coins did not go beyond the borders of the kingdom. This is also true for the other currency of Chu: the gold squares. The monetary use of gold is unusual in ancient China; this practice is exclusive to the Chu kingdom and probably comes from a local tradition that could be observed in việt-yue peoples living in territories stretching from the southern banks of the Blue River to the Thanh Hóa valley in Vietnam.8 The gold currency appears in the form of gold plates struck with square inscriptions giving the name of the unit, yuan 爰, and the name of the mint, the Chu capital: first Ying 郢 and later Chen 陳. The finds of gold yuan are quite numerous, mainly concentrated in the modern Hubei, Jiangsu and Anhui provinces (Thierry 1997: 146–152). From the late fifth century to the middle of the fourth century bc, some fiefs close to Chu kingdom issued “arch-feet spades” with weight indications; that means that the exchange value of these spades was related to their metal weight, i.e. their intrinsic value. Archaeological excavations over the last sixty years have produced evidence that this currency circulated in the regions in which it was issued, demonstrating that this form of money could not play a role outside those areas tied to the monetary system based on the weight unit, called jin 釿, of ca. 14 grams. 349

François Thierry

Therefore, the large arch-feet spades with the inscription er jin 二釿, “2 jin” (Figure 16.10) weigh 28–30 g., medium-size spades yi jin 一釿, “1 jin”, ca. 14 g., and rare small spades ban jin 半釿, “half jin” 7–8 g. These arch-feet spades were issued by about twenty fiefs of Han and Wei; the main issuers were Anyi 安邑 (modern Xiaxian, Shanxi), the capital of the Wei kingdom,Yao 殽, Jinyang 晉陽 and Yu 虞, vassal fiefs of that kingdom, and Ying 穎 and Fufan 甫反, vassals of Han (Wang 1988: nos 1245–1456). Although the metal value theoretically determines the exchange value of the monetary instrument, the metallic composition of these spades is not higher than that of the other spades and is clearly heterogeneous: the bronze alloy of Anyi spades, for example, contains 85 to 62% copper, 25 to 6% lead and 10 to 0% tin (Zhou-Dai 2002: 49; Zhou 2004: 9–10).These data clearly show the limits of the so called “intrinsic value” of the Anyi coinage, which still was the currency of the Wei royal clan, cast in the mints of the state capital. These spades are dated from before 365 bc, the year of the removal of the capital from Anyi to Daliang, and could be considered the royal currency during the first half of the fourth century. When the royal house settled in Daliang 大 梁 (present Kaifeng, Henan), spades of the same type were issued for a time, but the value no longer appeared to be related to the jin weight unit, because the local tradition was completely different.9 In the first half of the fourth century bc, all the fiefs of this region used this currency, which theoretically had an intrinsic value; but it seems that this concept of money, completely antagonistic to the fiduciary money tradition which was prevailing in the Great Plain, was progressively abandoned. Only one state was going to take it up again: Qin. In the beginning of the fourth century, the Qin 秦 dukedom was considered a semi-barbarian state; but this principality was going to be changed in a powerful kingdom by a series of deep reforms. After some attempts during Duke Xian (384–362), Duke Xiao (362–338) was calling for a political advisor able to make the dukedom powerful, and Shang Yang, a petty advisor coming from Wei kingdom, was appointed counsellor. Shang Yang proposed the “Way of Hegemony”, a series of reforms based on the legalist school philosophy: by the modification of the laws, by making them absolute and applicable to everybody without social distinction, by the unification of weights and measures, by establishing a strong control of the population, by awarding meritorious actions, by punishing breaches and faults, by stimulating farming and taking care to build a strong and implacable army, the state could become powerful. It is a fact that, in some years, this policy bears fruit; from Duke Xiao onward the central administration issued a true state currency based on the reform of the weights and measures. Shang Yang was a native

Figure 16.10  Early spade of Wei, city of Anyi (fourth c. bc). Bronze, 68 mm, 26,68 g Collection Wanxuanzhai 2399

350

Currency

of Wei, where he was in charge before he moved to Qin; for this reason, he probably introduced the conception of the metal value of the money. It could be probable, according to historical sources, that the closeness of Wei and Chu had been a cause of “contamination” of Qin by the monetary ideas of Chu, even before Shang Yang’s reforms, and that issuing of round coins (qian 錢) began during the reign of Duke Xian .10 The monetary instrument of Qin is a round coin with square hole bearing the mention of its legal weight, banliang 半兩 “half liang”, i.e. around 8 grams.11 In actual fact, the weights are not homogenous: the oldest coins weigh generally more than a half liang, some 12, or even 15 and up to 20 grams; conversely, in several deposits, coins weighing less than 5 grams were found. Banliang coins of the first period were issued immediately prior to the legalist reforms of Shang Yang; they are usually heavy and thick coins, with a quite small hole and with irregular characters in high relief (Figure 16.11). Those issued by the official mints usually have a diameter higher than 30 millimeters. After Shang Yang’s reforms, during the reign of Duke Xiao (360–338) and during the reign of King Huiwen (337–325), the coins became more regular, but thinner and lighter too: the average weight was around 8 grams (Figure 16.12). In addition to the official issues, there was a simpler and smaller local coinage more appropriate for daily life local exchange. Thanks to archaeological evidence, it is clear that Qin had a monetarized economy from the end of the fourth century onwards. Banliang coins have been found in archaeological sites dating from this period, in both Shaanxi and Gansu, which correspond to the hearth of the Qin kingdom during the Warring States period, and in Sichuan province, which was

Figure 16.11 Early banliang of Qin (ca. 378–360 bc). Bronze, 32,6 mm, 14,70 g Collection Wanxuanzhai 1176

Figure 16.12  Banliang of Duke Xiao of Qin (360–338 bc). Bronze, 31 mm, 8,08 g Collection Wanxuanzhai 3359

351

François Thierry

the first great conquest by Qin in 316 bc. The main discoveries took place on the excavations of Yong 雍, the former metropolis of Qin (Shaanxi), at Qingyang 慶陽 (Gansu), at Zhangbao 張堡 (Shaanxi), at Yaoxian 耀縣 (Sichuan) and especially at Qingchuan 青 川 (Sichuan), in the N°50 grave dated from 309/306 bc by a wooden inscription (Wu Zhenfeng 1985: 165–167, 171; Chen-Qian 1987: 21–27; Chen-Lu 1987: 3; Doo 1992; Doo 2000: 18–21). The conditions of use and circulation of the Qin banliang of the late Warring States period are well known thanks to the discoveries made in the N°11 grave at Shuihudi 睡虎地 (Yunmeng county 雲夢縣, Hubei); from that tomb, a great quantity of administrative and legislative documents written on bamboo slips has been unearthed (Shuihudi 1990; Hulsewé 1985; Thierry 2013: 141–155). Among these bamboo documents there are an important legislative text written prior to 247 bc, consisting of several chapters of which one, entitled Jinbulu 金 布錄, “Regulations on metallic and cloth [money]”, is completely devoted to the subject of the economy and money. The Jinbulu states that in the markets, round coins (qian 錢), no matter whether they are “beautiful” (mei 美) or “ugly” (e 惡), are to be used without distinction, and that it is strictly forbidden to select them during transactions: whichever the metal value (weight) they have, all the coins must be accepted for the same exchange value. The same text mentions another form of money, a length of cloth measuring 188 × 58,5 centimeters, called bu 布; in order for a cloth to be accepted as legal tender, it must have these exact dimensions. The exchange value of the bu is fixed at eleven coins, which, according to the law, were not selectable (Shuihudi 1990: 35; Hulsewé 1985: 52). There is clearly a contradiction between the legal form of money and the legal practice. The same text says “in the public storehouses which receive cash coins, 1000 coins [form] a ben” (Shuihudi 1990: 35; Hulsewé 1985: 52); thus the unit of account is a big wicker-basket (ben 畚) or a clay jar (xiang 缿) containing 1000 cash coins.These containers should be filled, verified and sealed by the market administration.12 In 1962, an intact xiang was found at Zhangbao (Shaanxi); its lid bore the official seal of the market of Du 杜; this clay vessel contained exactly 1 000 coins, 997 banliang, one liangzi 兩甾 and two yi hua of Qi (Chen-Qian 1987; Chen-Lu 1987; Doo 1992; Thierry 2015: 437–438). The banliang coins found in this xiang could be sorted into three types: (1) heavy and thick coins generally with crude calligraphy, weighing 7 grams plus, reaching 11 grams for some; (2) coins with more regular calligraphy, weighing around 6 grams; and (3) many small heterogeneous coins of private and/or illicit manufacture, the lightest of which weigh less than 2 grams. According to the Jinbulu, all the 1 000 coins in this xiang, no matter whether they are light or heavy, have the same exchange value: despite its inscription, the banliang clearly did not function as an ingot of metal impressed with its weight, but as fiduciary money. On the other hand, metallic analysis shows that the bronze alloy is irregular: the study of thirteen banliang of Qin by Professor Zhou Weirong gives a copper content ranging from 61,35% to 86,89%, with an average content at 74,72% (Zhou 2004: 29). Despite its inscription, it is clear that the exchange value of the banliang is – up to a certain limit – bound neither to its weight nor to its metallic composition: this money has no intrinsic value and is clearly a fiduciary money.13 In practice, despite their legalist political ideas, the Qin authorities are forced to take the people’s financial habits into account and to yield – to some extent – to the tradition and to the Mohist-Confucianist conception of money. In fact, at that time, even prior to the monetary unification, what would become the standard form of the Chinese currency during almost 2 000 years was beginning to take root: a round piece of metal with a square hole, of which the size is of more importance than the weight. 352

Currency

The time of hesitation Historical sources (Shiji and Hanshu) claim that the banliang was created in 221 bc by Qin Shihuangdi, the First Emperor, after the creation of China as a unified empire, but archaeological evidence, provided by both the excavation results and the primary source studies, shows this assertion is not true. The monetary unification, i.e. the forced use of a unique currency, was only a consequence of the annexation of the former Warring States: in every conquered territory, the Qin imposed its own currency, the banliang. On the other hand, the result of the Qin political hegemony on the Chinese world prior to his victory is that, progressively, the round shape with square hole was spreading in the neighbouring states, such as the three yi 賹 coins of Qi, the yi hua 一化 and the Yi hua 易化 of Yan and the liangzi 兩甾 of feudal fiefs in Henan. . . . To some extent, the typological unification came before the monetary unification of China by Qin in 221 bc, and, therefore, the so-called unification of currency is not entirely the impressive revolution it is often described as being (Thierry 2008;Thierry 2013: 165–170).When King Zheng, the future Qin Shihuangdi, became of age in 238 bc, his kingdom already covered three-quarters of the actual China; his opponents were already vassalized or paralyzed, or even weakened. Archaeological discoveries clearly show that whenever the round coins of other states had quite the same size as the banliang coins, they were circulating in the Qin territories for the value of a banliang: in the Zhangbao hoard, there are one liangzi and two coins of Qi; in at least two tombs of Dongsunba necropolis (Sichuan), banliang and liangzi have been found together; and recently, a round coin of Xiangyin 襄陰 has been found in a banliang deposit (Zhu 1991: I, 92; Dang 2010: 36). Some aristocrats of Qin, such as Lü Buwei, marquess of Wenxin 文信 and prime minister from 250 to 237, and Cheng Qiao, the king’s brother and prince of Chang’an 長安 (256–239), issued coins with the name of their fiefs, but respecting the square-holed round coin type (Thierry 1997: 175–177). From the political unification (221 bc) onward, cowries, spade- and knife-money were demonetized and withdrawn from circulation.The round coin with square hole was the unique monetary instrument in the new empire. But as unique as it was, this money was by no means uniform. In brief, the monetary situation at the beginning of the imperial period is first characterized by a contradiction between a coin that was legally considered an ingot of metal impressed with its weight, and a legal use that made this monetary instrument a fiduciary money. Second, the circulating currency was not homogeneous, principally because of a lack of state monopoly of casting coins: there is no information about the legal status of the casting in the Qin state before and after the unification; the coins were not only heterogeneous (weight, size and metallic composition), but private casting is not forbidden; although the central authorities cast coins in official mints established in the capital and in the seats of the main provinces, it is highly probable that, in absence of a state monopoly, the government gave big entrepreneurs the right of casting coins (Thierry 2007: 230–235). There was also a problem related to the face value of the monetary instrument in some transactions: the Jinbulu states that “for what is to buy or for sale, a ticket with the price must be attached; when small objects [are] not worth one coin (qian), it must not be attached” (Shuihudi 1990: 36; Hulsewé 1985: 53). It is clear that the exchange value of the monetary instrument was sometimes too high for some daily life exchanges, but there was no fractional currency. At last, the authorities were progressively being faced with another problem: the cost of the monetary instrument. As long as the labour was mainly a form of slavery and as long as copper was cheap, the cost was not a problem; but little by little, the massive use of copper in China was going to provoke the raising of its price, and the end of slavery posed the question of the cost of labour. Just as the price of a bowl or of an article of clothing, the price of the monetary instrument, i.e. the price of labour and materials necessary to cast coins, has to be taken in account. 353

François Thierry

At first, since the authorities were looking for a monetary instrument more suited to the economic necessities and to the traditional practices, what was simpler than letting the people choose their own money? After the fall of the Qin dynasty (209–206 bc), “because the Qin coins were too heavy and difficult to use, [the law] was modified and the order was given [for] the people to cast [their] own currency” (Shiji: xxx, 1417). This passage hints that previously, even if there was no state monopoly, casting coins was not free for the population. By this decree, the first emperor of the Han 漢 dynasty was starting a Confucian policy of unrestricted casting that would deeply influence the monetary history of China. The result was complete chaos, because every village, every region, every province cast its coins corresponding to its local needs; the imperial administration was without a currency that could be used everywhere in the empire for trade and fiscal requirements; the coins became so light as it is said that they floated on water. In the following years, the Han authorities backtracked on the unrestricted minting for the people14 and were in search of the ideal monetary instrument, switching from too heavy coins (bazhu qian 八銖錢, “eight zhu coins”, i.e. ca. 5 g, issued in 186 bc) to too light ones (wufen qian 五分錢, “coins of one fifth [of banliang], i.e. c 1,6 g, issued in 182 bc, Figure 16.13) (Thierry 2003: 28–29; Shi 2009: 80–87); during Wendi’s reign (179–157 bc), coins were issued bearing the inscription banliang (i.e. “half liang”, Figure 16.14), but their legal weight is four zhu (ca. 2,5 g), one-third of the inscribed weight. The coin inscription was semantically slipping from a weight indication to a coin name. Banliang of four zhu were not uniform because at the same time, the right of unrestricted minting was restored: both aristocrats and commoners cast coins in different types, good and nice coins as well as very light banliang. And although Emperor Jingdi forbade the people from minting in 144 bc, the monetary situation was no better during his reign.

Figure 16.13  Wufen qian of Empress Gaohou of the Western Han (182 bc). Bronze, 23,8 mm, 1,36 g Collection Wanxuanzhai 2935

Figure 16.14 Four zhu banliang of Han Wendi (175 bc). Bronze, 23,5 mm, 2,54 g Collection Wanxuanzhai1035

354

Currency

As soon as Wudi 武帝 (140–87 bc), the new emperor, ascended the throne, the authorities decided on making the coin inscription corresponding to the weight: new coins weighing three zhu (ca. 2 g) bearing the inscription sanzhu 三銖, “3 zhu”, were issued (Shanghai 1994–95: II, nos 691; Thierry 2003: no 137). This “truth operation” did not produce the expected results, because this coin was considered a little bit too light and hence not well accepted; for this reason, in 136 bc, the sanzhu were demonetized and the banliang of four zhu issued again. The monetary situation was by no means satisfactory: there were all types of coins in circulation, from the old Qin banliang to the recent “four zhu banliang”, passing through the “eight zhu coins”, the “wufen coins” and the different types of “four zhu banliang”, not to mention the private and illicit productions.

An ideal currency? Changing the material The main characteristics of Wudi’s reign (140–87 bc) were the administrative stabilization and the military expansionism, which needed more and more financial means. Cash needs were so important that the authorities were exploring different ways to alleviate the problem and thought they found the miracle solution: changing the material. In the year 119 bc, one issued a leather currency, called bailupibi 白鹿皮幣, “money in white deer skin”, with a legal tender of 400 000 cash; then the “three coins in white metal”, baijin sanpin 白金三品, were issued: a round coin bearing the design of a dragon (Figure 16.15), weighing eight liang with the value of 3000 cash; a square coin bearing the design of a horse, “lighter”, worth 500 cash; an oval coin bearing a tortoise-shell pattern, “thinner”, worth 300 cash (Shiji: xxx, 1427; Hanshu: vi, 178, xxivb, 1164). According to the sources, these coins were cast in an alloy of silver and tin, but archaeological evidence shows that the metallic composition is generally mainly lead, or lead-wolframsilicium alloy, or lead-bronze alloy; there is no, or very little, silver (Wang-Xie 2003; Zhao 2003; Jiang-Ai 2003). The three coins bear the seal of the “Lesser Treasure”, shaofu 少府, which must be considered the issuer. The reverse inscription of the round coin is particularly interesting because it is an illustration of the new ties between Western Han China and Western countries: we can see a crude inversed imitation of the legend of an Indo-Greek or Indo-Parthian coin. Interesting though it was in a theoretical point of view, this attempt to transfer the value of the currency on another material than bronze was soon a failure, and the authorities had to find another means to solve the monetary issue.

Figure 16.15  Baijin sanpin of Wudi, dragon design coin (119 bc). Lead, 55 mm, 116,10 g Courtesy China Numismatic Museum, Peking

355

François Thierry

The adjustment of wuzhu In 118 bc, the authorities created a new bronze round coin heavier than the former “four zhu banliang” and bearing the inscription of its real weight, wuzhu 五銖, “five grain”, that replaced the two characters banliang. With the aim of controlling the issues and ensuring a homogeneous production, the state monopoly of casting was imposed by successive technical reforms and innovations. The first issue of wuzhu from 118–115 bc, known as the junguo wuzhu 郡國五銖, “wuzhu of provinces and fiefs” (Figure 16.16), showed links with the coins of the previous period; but, although the banliang had a plain reverse, the new coins, by ordinance, had to have a raised rim around the edge on the reverse “in order that it would be made impossible to trim them to get filings” (Shiji: xxx, 1429): this innovative obligation implied that the founders had to mastered the technique of the double mould (hefan 合范). That was a first technical obstacle; and the technological problems posed by this innovation were not always surmounted by the provincial or fiefs mints. For example, a mint of the Wudi period composed of four furnaces was excavated at Potoucun 坡頭村 (near Dengcheng, Shaanxi province), in which forty-one bronze moulds, more than 100 clay reverse moulds and the complete equipment necessary for casting were found (Cui 1982); but, as noticed by the Chinese archaeologist Wu Zhenfeng, “what is particularly interesting is the fact that more than a hundred reverse moulds found in the Potocun site are all plain and flat, without inner or outer rim hollows on the would be reverse” (Wu Zhenfeng 1989: 22); clearly some wuzhu continued to be cast in the same way as the banliang. In the year 115 bc, it was stipulated that only coins of which ragged edges have been trimmed off by the central workshop at Chang’an would be acceptable in transactions with the state; these coins, known as chice wuzhu 赤側五銖, “red edge wuzhu” (Figure 16.17), would be exchangeable with the junguo wuzhu at the rate of 1:5 (Shiji: xxx, 1434). With the good coins being trimmed off to be used in the state finances later, an immense process of control of the coinage took place in the central treasury. But this process was a profitable operation as well, because the trimmed coin had the face value of five ordinary wuzhu. Obviously, this change in the coinage corresponded to an increasing of the tax pressure for the population: the reform raised the actual poll tax from 120 to 600 junguo wuzhu and constrained the farmers to pay five times the previous amount of the monetary taxes or to get 120 chice wuzhu, probably at a usurious rate. Everything happened as if the only goal of the technological innovations of 118 and 115 bc was to make private casting increasingly difficult. Finally, in the year 113 bc, because both princes and forgers had penetrated the technique of trimming coins, and a currency not controlled by the official treasury started to circulate, a state

Figure 16.16  Junguo wuzhu (118–113 bc). Bronze, 29,1 mm, 5,21 g Collection Wanxuanzhai 652

356

Currency

Figure 16.17  Chice wuzhu (115–113 bc). Bronze, 24,6 mm, 4,18 g Collection Wanxuanzhai 1005

Figure 16.18  Sanguan wuzhu (from 113 bc). Bronze, 25,2 mm, 4,01 g Collection Wanxuanzhai 1389

monopoly of coinage was established: only the central mint of Chang’an, settled in the Shanglin imperial park, was authorized to cast coins; these coins were called Shanglin sanguan wuzhu 上林 三官五銖, “wuzhu of the three offices of the Shanglin” (Figure  16.18). The wuzhu, with a size of around 26/27 millimeters and weighing between 3 and 4 grams, ideally corresponded to the demands of both the people and the traders, not too heavy, not too light; it is well suited to the tradition and the needs of exchange: these qualities ensured its success for almost 800 years.15 It will be the pattern of the Chinese coin, known as “cash” in current usage.

Wuzhu, a weight or a coin? Basically, the wuzhu remained a fiduciary money: its exchange value did not depend on either its weight or its metallic composition. In the tomb of the king of Zhongshan, Prince Liu Sheng, who died in 113 bc, around 2000 wuzhu were found; but these coins, coming from the prince’s administration, have an average weight between 3 and 5 grams, but the gap between the lightest and heaviest is from 2,6 to 7 grams; on the other hand, their metallic composition is far from being uniform: the results of the analysis of twenty-six coins from the tomb show a copper content ranging from 88,04% to 67,23% and a lead content from 22,49% to 3,81%; but the size, on the other hand, is almost always the same (Mancheng 1980: 332–334; DaiZhou-Fan 1991: 12–13; Zhou-Dai 2002: 62, 320; Zhou 2004: 36). Above all, what is important is the look of the coin and its eye appraisal by the economic players. It is from Wudi’s reign that the string, guan 貫, a theoretical amount of 1000 cash threaded on a rattan cord, start to be in use (Shiji: xxx, 357

François Thierry

1420; Hanshu: vi, 178); but the string allowed, much better than the ben, to judge the quality of the coins: neither the inscription nor the weight were checked – the appearance (color) and the size were the only acceptance criteria. The drastic centralization of coin casting in the Chang’an mint was not a warrant for a homogeneous and good quality production, and frauds began very early. The Yantielun, a book written about thirty years after the state monopoly was established, says: after that, all the coinage in the empire was demonetized and the authority gives the monopoly of casting to the Three Offices in charge of the regulation of economy. Recently, officials were then gaining fraudulent benefits in casting coins not conform to the standard, so that there are thin and thick [coins] and heavy and light [coins].The farmers being not experienced in comparing the different coins, they preferred the old coins and suspected the new. (Yantielun: iv, 53) On the other hand, the casting did not remain centralized in the Chang’an mint for long: clay mother-moulds were found in other parts of Shaanxi province (at Fufeng, at Xingping and in the south of the present Xianyang town). Other mother-moulds were discovered in Hebei province, near modern Peking, in Shandong province and in Inner Mongolia. These findings show that the centralization was more or less abandoned during the reign of Xuandi (73–49 bc): a bronze mother-mould for Xuandi wuzhu was unearthed in Shandong province, and others of later periods were found in Henan, Shanxi and Sichuan. And, de facto, the currency became progressively more and more heterogeneous, especially from Yuandi (48–33 bc) onward. During the last decades of the Western Han, the weight of the wuzhu had a trend toward reducing in weight: for example, in a group of forty-eight wuzhu found in Peking, the weights range from 4,2 to 1,7 grams, and the copper content ranges from 95,68% to 55,60% (Zhou 2004: 37–39). This situation provoked some trouble in the circulation and had, above all, the ideological consequence to call into question the social contract answered for by the emperor: when the coins’ weights and metallic composition were irregular, but ranging in an acceptable gap, the heavy coins compensating for the light ones, the social contract was respected; but if all the coins were thin, the moral contract between the authority and the society was broken: and from a Confucian point of view that was a fundamental matter. The Confucian literati did not delay starting to claim that to name a coin using a weight designation was opposite to the philosophical principle of zhengming 正名, “the correctness of designations” (Lunyu: vii-13, 209–210). Xun Kuang, a great Confucian philosopher of the third century bc, synthesized his master’s thinking: when objects of different nature have obscure ties with the words that distinguish them, the noble and the vile will not clearly appear, the similar and the dissimilar will not be distinguished. [. . .] Intelligent people take care to establish differences and to distinguish, they use words that put one’s finger on the realities. [. . .] Consequently names are imposed: what is similar is similarly called, what is dissimilar is dissimilarly called. (Xunzi: xxii, 260–262)

Wang Mang’s reforms In 8 bc, when Wang Mang 王莽 became regent, the literati took the power and made Confucianism the theoretical ground of the monetary policies: the use of a weight unit name as 358

Currency

currency designation was not correct, because coins were not weights but means of exchange; designations such as banliang or wuzhu must be banned and replaced by the former Zhou currency designations, as bu 布, “spade”, dao 刀, “knife”, or quan 泉, “round coin”, and the coins must have the shape corresponding to their designations. On the other hand, in the framework of a return to the strict Confucian tradition,Wang Mang intended to abandon any link between metal value and exchange value in currency: the exchange value of the monetary instrument was fixed by a moral contract between the emperor and his people, a contract in which each one had reciprocal duties, such as a father toward his sons, and sons toward their father. In the year 7 ad, three new monetary instruments were issued: the jincuodao 金錯刀, “gold inlaid knife” (Figure 16.19), worth 5000 cash, the qidao 契刀, “fiduciary knife”, worth 500, and the daquan wushi 大泉五十, “great cash worth 50” (Figure 16.20). Wang Mang became emperor in ad 9,

Figure 16.19  Gold inlaid knife of Wang Mang (ad 9–10). Bronze and gold, 74,5 mm, 33,65 g Collection Wanxuanzhai 439

Figure 16.20  Great cash of 50, daquan wushi of Wang Mang (ad 7–14). Bronze, 28,6 mm, 8,80 g Collection Wanxuanzhai 595

359

François Thierry

and he started to launch a set of monetary reforms which were organized, in ad 10, in a coherent system of purely fiduciary currency, a system called baohuozhi 寶貨制, “system of the precious money”: this system mainly consisted of five metallic round coins of which the face values ranged from 10 to 50 and ten metallic spade-shaped coins of which the face values ranged from 100 to 1000; the base coin bore the inscription xiaoquan zhi yi 小泉直一, “small cash worth 1” (Figure 16.21). The reciprocal face value of these coins was not linked to their intrinsic value (metal value): the small coin of 1 weighed around 1,3 grams and the spade of 500, chabu wubai 差布五百, for example, weighed around 10 grams, which is far from 500 times the weight of the small base coin. Contrary to what is said in the Han historiography, violently hostile to Wang Mang, which was later repeated, modern archaeology has shown, as evidenced by the number of moulds and coins – mainly coins of 1, 50 and 1000 (dabu heng qian 大布橫千, Figure 16.22) – that this fiduciary system operated relatively well, as long as the confidence was prevailing. In the year 14, noticing that the population had de facto made a selection in the different coins of the baohuzhi and continued to use the wuzhu, Wang Mang launched his fifth reform.

Figure 16.21  Small cash worth 1, xiaoquan zhiyi of Wang Mang (ad 10–14). Bronze, 16 mm, 1,57 g Collection Wanxuanzhai 3275

Figure 16.22  Spade of 1000, dabu heng qian of Wang Mang (ad 10–14). Bronze, 59 mm, 16,16 g Collection Wanxuanzhai 763

360

Currency

He decided that the base unit would be a wuzhu-like coin, which could not of course bear a weight name. That was the huoquan 貨泉, “cash money” (Figure 16.23), created along with a high-value coin, the huobu 貨布, the “spade-money” (Figure 16.24), of which the face value is twenty-five huoquan. On the other hand, it was necessary to solve the problem of the amount of daquan and wuzhu; the first ones, issued without a break from the year ad 7 onward, made up a large proportion of the amount of money in circulation, and if they were demonetized there would be that much less money. Then, it was decided that daquan would circulate alongside the huoquan, and for the same face value, during the next six years; after that they would be demonetized. The circulation of the wuzhu was tolerated. Once again, contrary to what is said by the Han historiography, archaeology clearly shows that this system was running quite correctly (Bai 1990; Lu 1990; Li 1990). It was the fall of the Wang Mang’s dynasty, due to the breaking of the

Figure 16.23  Huoquan of Wang Mang (ad 14–23). Bronze, 23 mm, 3,18 g Collection Wanxuanzhai 587

Figure 16.24  Huobu of Wang Mang (ad 14–23). Bronze, 58 mm, 16 g Collection Wanxuanzhai 1386

361

François Thierry

Figure 16.25  Wuzhu of Eastern Han (Guangwudi, ca. ad 40–58). Bronze, 26,3 mm, 2,96 g Collection Wanxuanzhai 1331

Yellow River dykes and their consequences, that put an end to the monetary system, and not the monetary crisis that provoked the fall of the dynasty. In the end, Wang Mang’s reforms paradoxically led to the confirmation of the suitability of the currency created by Wudi to the economic and social development of actual China. After Wang Mang’s fall in ad 23, the wuzhu was not cast immediately, and according to local needs, every coin somewhat similar in size to the wuzhu (banliang of four zhu, sanzhu, daquan wushi, huoquan), as well as the wuzhu, was accepted both in daily transactions and for tax purposes. Sometimes, such as in Sichuan province or on the western limes, local or military authorities cast token coins, often small and thin, even in iron (Hou-Han shu: xiii, 537–538; Wang: 49). Not until the year ad 40 were the wuzhu cast again (Figure 16.25), and for two centuries, this coin remained the sole currency of the Eastern Han dynasty (ad 25–220); until the last quarter of the second century ad, the size remained remarkably stable, but the analysis of both the weight and the metallic composition show significant differences (Thierry 2003: nos 541–740; Zhou 2004: 44–45; Deng 2003: 160–162).The changes that were locally appearing from the end of the second century onward were temporary palliatives to crises linked to the collapse of the central power and to the rise of powerful regional warlords (Thierry 2003: 62–64). During the six centuries following the fall of Wang Mang, the contradiction between the mention of the legal weight on the coins and their fiduciary use were going to poison the monetary circulation and the monetary policies of the different powers and dynasties (Thierry 2001). While the authorities vainly tried to solve this contradiction by changing the inscription or the coin type, or giving permission for unrestricted casting, all the powers of Confucian obedience attempted to cut the exchange value of the coin from its metal value.

Notes 1 About the discussion concerning the number of cowries in a peng, and about the assumption of a variation in value of the cowries according to their sizes, see Creel 1937: 87, Wang 1951: 85–87, Thierry 1997: 40–41. 2 I am following the chronology given by Edward Shaughnessy (“Calendar and Chronology”, in, Michael Loewe & Edward Shaughnessy (dir.), The Cambridge History of Ancient China: From the Origins of Civilization to 221 bc, Cambridge, Cambridge University Press, 1999, pp. 19–29). 3 A small number of cypraea annulus were also found, alongside very small numbers of mauritia arabica, erronea erronnes, erosaria miliaris and large cypraea tigris; there are also some finds of such freshwater molluscshells as oliva mustelina. 4 Around 60% copper, 30% lead and 10% tin (Dai-Zhou 1993: 311; Zhou 2004: 5–6). 5 For the discussion on the character yi, see Wang 1988: 1121–1122; Zhang 1997: 31–39 Thierry 1997: 106–108; Huang 2001: 245–247.

362

Currency 6 Yi was the name of the capital, but, as often in the texts of that period, it is likewise a means to designate the kingdom (Thierry 1997: 107): we find Liang 梁 for the kingdom of Wei, the capital of which is Daliang, or Ying 郢 for the kingdom of Chu, etc. 7 Shang-Wang-Tan 1983: 144–145; Li 1995: 388–407; Thierry 1997: 124–125. Some researchers read da hua 大化, “large money” (Wang 1988:1114; Shanghai 1994–95: I, n° 914–1017) and others da dao 大 刀, “large knife” (Huang 2001: 277). 8 In other parts of China, gold was considered the high-degree money (shangbi) or middle-degree (zhongbi) money, according to different texts; it was not a circulating currency; it was used as medium of payment in state finance and as medium of exchange in interstate trade, because the monetary systems of the different kingdoms were not compatible (Guanzi: xxiv-qingzhong-2, 404, xxiv-qingzhong-5, 415). 9 On these spades, the character jin 釿 is not a weight unit, but means “coin”: Liang xin jin 梁新釿, “new coin of Liang” (Thierry 1997: 81, 215). 10 According to several scholars, the quotation in the Shiji saying that in the seventh year of Duke Xian (378) “one starts to issue [coins] for markets”, would mark the first putting in circulation of banliang in Qin (Wang-Liu 2005: 1–4; Doo 2000: 6–7; Doo 2006: 3). 11 By analyses of Qin period weights, it has been possible to know that the actual weight of a liang was 15.88 g; as one liang consists of 24 zhu 銖, the zhu is 0,66 g; 16 liang make a catty (jin 斤) of 254 g (Qiu 1984: 79–80; National Board of Measures 1984: 112–138; Shi 2009: 307). 12 In the Guanshi 關市 (“On the customs and markets”) chapter, it is quoted that “in the private workshops and in the government store houses, when coins (qian) are received, these must be immediately put in a jar (xiang); the market chiefs have to control this filling; those who do not follow the rules will be fined one armour” (Shuihudi 1990: 42; Hulsewé 1985: 56–57). 13 Despite the obvious, disregarding sources, archaeological data and analyses, Walter Scheidel defends the idea that the Qin currency is not a fiduciary money, but a money with intrinsic value (Scheidel 2009: 139–143). 14 The bamboo-slip document entitled Qian lü 錢律 (“Laws on currency”) in the Ernian lüling 二年律 令 (“Statutes and Ordinances of the year two”, dated from 186 bc) excavated from the tomb no 247 at Zhangjiashan 張家山 (Jiangling, Hubei) provides clear evidence that the laws against illicit/free minting were strongly enforced in the beginning of the Gaohou’s reign (Shi 2009: 183–188). 15 On the Han monetary history from 206 to 113 bc, see Peng 2007: 80–84; Jiang 1997: 73–100; Thierry 2003: 28–36.

Works cited Bai, Gonghe 白共和 (1990), ‘Shaanxi sheng Qingjianxian chutu Wang Mang shiqi jiaocang qianbi’ 陕西省 清涧县出土王莽时期窖藏钱币, Shaanxi jinrong 1990–154(XIV):106–108. Ban, Kaiming 班开明 (1985), ‘Hebei Weixian chutu yipi gu dao bubi’ 河北蔚县出土古刀布币, Wenwu ziliao congkan IX-1985: 210–212. Chen, Zhenzhong 陳振中 (1982), ‘Yin Zhou de qian, bo – qingtong chan he chu’ 殷周的錢，鎛 – 青銅 鏟和鋤, Kaogu 1982-III: 289–299/256. Chen, Zunxiang 陈尊祥 and Lu,Yuan 路袁 (1987), ‘Shoupa Zhangpu jiaozang Qin qian qingli baogao’ 首 帕 张堡窖藏秦钱清理报告, Zhongguo qianbi 1987-III: 3–12/31. Chen, Zunxiang 陈尊祥 and Qian,Yu 钱屿 (1987), ‘Shaanxi Chang’an Zhangbao Qin qian jiaocang’ 陕西 长安张堡秦钱窖藏, Kaogu yu Wenwu 1987-V: 21–27. Creel, Herrlee Glessner (1937), La naissance de la Chine, transl. by Clerc Salles, Paris: Payot. Cui, Hanlin 崔漢林 (1982), ‘Shaanxi Potoucun Xi-Han zhuqian yizhi fajue jianbao’ 陝西坡頭村西漢鑄 錢遺址發掘簡報, Kaogu 1982-I: 23–30, pl.VI–VII. Dai, Zhiqiang 戴志強 (1983), ‘Anyang Yinxu chutu beihua chutan’ 安阳殷墟出土贝化初谈, Wenwu 1983-III: 72–77. Dai, Zhiqiang 戴志強 and Dai, Yue 戴越 (2014), ‘Cong qian bo dao kongshoubu’ 從錢鎛到空首布, Zhongguo qianbi 2014-I: 3–8. Dai, Zhiqiang and Zhou, Weirong (1993), ‘Studies on the alloy composition of past Dynasties copper coins in China’, in Proceedings of the 11th International Congress of Numismatic, Bruxelles 1991, 4 volumes, Louvain: vol. III, 311–324. Dai, Zhiqiang 戴志強, Zhou, Weirong 周衛榮 and Fan, Xiangxi 樊祥禧, ‘Mancheng Han mu chutu wuzhu qian de chengfen jiance ji you guan wenti de sisuo’ 滿城漢墓出土 五銖錢的成分檢測及有關 問題思索, Zhongguo qianbi 1991-II: 12–17.

363

François Thierry Dang, Shunmin 党順民 (2010), ‘Shaanxi chutu faxian er ze’ 陝西出土發現二則, Zhongguo qianbi 2010-I: 36–37. Deng, Zhuanzhong 鄧傳忠 (2003), ‘Hubei Xiangyang chutu Dong Han jiaocang qianbi’ 湖北襄陽出土 東漢窖藏錢幣, Shaanxi qianbi yanjiu wenji IV-2003: 157–168. Doo,Wai-San 杜維善 (1992), ‘Shoupa Zhangbao jiaocang Zhanguo banliang xingshi fenxi chutan’ 首帕張 堡窖藏戰國半兩型式分析初探, Shanghai bowuguan jikan IV-1992: 214–243. Doo, Wai-San 杜維善 (2000), Banliang kao 半兩攷, 2 volumes, Shanghai: Shanghai shuhua. Doo, Wai-San 杜維善 (2006), Banliang tuxi 半兩圖系, Shanghai: Shanghai shuhua. Erlitou CSSA Archaeological Team, ‘Yanshi Erlitou yizhi xin faxian de tongqi he yuqi’ 偃师二里头遗址 新发现的铜器和玉器, Kaogu 1976-IV: 259–263. Guanzi 管子, (ed. by Liu Xiang, ca. 10 ad), in Zhuzi Jicheng 諸子集成, 8 vol., Shanghai: Shanghai shudian, vol. 5. Guo,Yong 郭勇 (1965a), ‘Shanxisheng Yuanping xian chutu de Zhanguo huobi’ 山西省原平县出土的战 国 货币, Wenwu 1965-I: 46–55. Guo,Yong 郭勇 (1965b), ‘Shanxi Yanggao Tianqiao chutu de Zhanguo huobi’ 山西阳高天桥出土的战国 货币, Kaogu 1965-IV: 166–170. Hanshu 漢書 (by Ban Gu 班固, ca. 90 ad), Peking: Zhonghua shuju, 12 vol., 1983. Hou-Han shu 後漢書 (by Fan Ye 范曄, ca. 450), Peking: Zhonghua shuju, 12 vol., 1987. Houma Archaeological Team (1962), ‘Houma Niucun gucheng nan Dong-Zhou yizhi fajue jianbao’ 侯马 牛村 古城 南东周遗址发掘简报, Kaogu, 1962-II: 55–62. Huang, Xiquan 黄锡全 (2001), Xian Qin huobi tonglun 先秦货币通论, Peking: Zijincheng publ. Hulsewé, A. F. P. (1985), Remnants of Ch’in Law, an annotated translation of the Ch’in legal and administrative rules of the 3rd century bc, discovered in Yün-meng Prefecture, Hu-pei Province in 1975, Leyden: Brill. Jiang, Baolian 姜宝莲 and Ai, Lin 哀林 (2003), ‘Han dai “baijin sanpin” zaitan’ 漢代白金三品在探, Shaanxi qianbi yanjiu wenji, 4: 24–29. Jiang, Ruoshi 蔣若是 (1997), Qin Han qianbi yanjiu 秦汉钱币研究, Peking: Zhonghua shuju. Li, Baohua 李葆華 (ed.) (1995), Zhongguo qianbi dacidian 中國錢幣大辭典, vol. I, Xian Qin Han bian 先秦 編, Peking: Zhonghua shuju. Li, Baohua 李葆華 (ed.) (2006), Zhongguo qianbi dacidian 中國錢幣大辭典, vol. VIII, Kaogu ziliao bian 考 古資料編, Peking: Zhonghua shuju. Li, Jinping 李金平 (1990), ‘Henan Lingbao chutu Mang qian jiaocang’ 河南灵宝出土莽钱窖藏, Shaanxi jinrong 1990–154(XIV): 97–105. Liu, Xian 刘翔, Chen, Kang 陈抗, Chen, Chusheng 陈初生 and Dong, Kun 董琨 (1989), Shang Zhou gu wenzi duben 商周古文字读本, Peking:Yuwan chubanshe. Lu, Quanyou 卢泉友 (1990), ‘Gansu Ningxian chutu Wang Mang shiqi jiaocang qianbi he Qingyang diqu faxian de ping wuqian qianhuai’ 甘肃宁县出土王莽时期窖藏钱币和庆阳地区发现的 平五千钱坯, Shaanxi jinrong 1990–154(XIV): 85–91. Lunyu 論語 (by Confucius), Les Entretiens de Confucius, Chinese text edited and transl. by Couvreur S. J., in Les Quatre Livres, Hokienfou 1895, 69–296. Luo, Xizhang 罗西章 (1985), ‘Cong faxian yi mei Xi Zhou tong bei zailun Xi Zhou huobi’ 从发现一枚 西周铜贝再论西周货币, Shaanxi jinrong 1985-IV: 1–12. Ma, Chengyuan (1986), Ancient Chinese Bronzes, Oxford, Hong Kong: Oxford University Press. Mancheng (1980), Institute of Archaeology of the China Social Sciences Academy, Mancheng Han mu fazhan baokao 滿城漢墓發展報考, 2 vol., Peking: Wenwu chubanshe. Mei, Lihua 美力華 (2010), ‘Liaoxi faxian yuanshi jianshoudao’ 辽西发现原始尖首刀, Zhongguo qianbi 2010-I: 39. National Board of Measures, China History Museum and Palace Museum (1984), Zhongguo gudai duliangheng tuji 中国古代度量衡图集, Peking: Wenwu chubanshe. Peng, Xinwei 彭信威 (2007), Zhongguo huobi shi 中國貨幣史, Shanghai: Shanghai renmin chubanshe (1958), 4th ed. 2007. Qinghai Archaeological Team (1976), ‘Qinghai Ledu Liuwan yuanshishehui mudi fanying chu de zhuyao wenti’ 青海樂都柳灣原始社會墓地反映出的主要問題, Kaogu 1976-VI: 365–377. Qiu, Guangming 丘光明 (1984), ‘Woguo gudai quanhengqi jianlun’ 我國古代權衡器簡論, Wenwu 1984X: 77–83. Shandong Numismatic Society 山西省錢幣學會 (1996), Qi bi tushi 齊幣圖釋, Dezhou: Qi Lu shushe. Shang, Chengzuo 商承祚, Wang, Guichen 王贵忱 and Tan, Lihua 谭棣华 (1983), Xian Qin huobi wenbian 先秦货币文编, Peking: Shumu wenxian publ.

364

Currency Shanghai Museum Bronzes Study Team (1994–95), Shanghai bowuguan cang qianbi 上海博物館藏錢幣, 8 vol., Shanghai: Shanghai shuhua publ. Shanxi Numismatic Society 山西錢幣學會 (1989), Zhongguo Shanxi lidai huobi 中國山西歷代貨幣, Taiyuan: Shanxi Renmin publ. Shao, Zhujiu 邵铸久 (1993), ‘Jiangsu beibi shitan’ 江苏贝币试探, Suzhou qianbi IX-1993: 4–6. Scheidel, Walter (2009), ‘The Monetary Systems of the Han and Roman Empire’, in Scheidel Walter (ed.), Rome and China, Comparative Perspectives on Ancient World Empires, Oxford: Oxford University Press, 137–207. Shi, Junzhi 石俊志 (2009), Banliang qian zhidu yanjiu 半兩錢制度研究, Peking: Zhongguo jinrong chubanshe. Shiji 史記 (by Sima Qian 司馬遷, 91 bc), Peking: Zhonghua shuju, 10 vol., 1985. Shuihudi Tomb Slips Study Team (1990), Shuihudi Qin mu zhujian 睡虎地秦墓竹简, Peking: Wenwu chubanshe. Tang, Lan 唐兰 (1976), ‘Shaanxi sheng Qishanxian Dongjiacun xin chu Xi Zhou zhongyao tongqi mingci de yiwen he zhushi’ 陕西省岐山县董家村新出西周重要铜器铭辞的译文和注释, Wenwu 1976-V: 55–59/63. Thierry, François (1997), Monnaies chinoises, I- L’Antiquité préimpériale, Paris: Bibliothèque nationale de France. Thierry, François (2001), ‘La fiduciarité idéale à l’épreuve des coûts de production: quelques éléments sur la contradiction fondamentale de la monnaie en Chine’, Revue Numismatique 2001: 131–152. Thierry, François (2002), ‘Sur les spécificités fondamentales de la monnaie chinoise’, in Testart, A. (ed), Aux origines de la monnaie, Paris: Errance, 109–144. Thierry, François (2003), Monnaies chinoises, II- Des Qin aux Cinq Dynasties, Paris: Bibliothèque nationale de France. Thierry, François (2007), ‘Peut-il y avoir faux-monnayage en l’absence de monopole de fonte monétaire ? L’exemple du royaume de Qin (IVe-IIIe siècles bc)’, in G. Béaur, H. Bonin, C. Lemercier (eds.), Fraude, contrefaçon et contrebande de l’Antiquité à nos jours, Droz, Geneva: Droz, 221–235. Thierry, François (2008), ‘L’Unification monétaire de Qin Shihuangdi: aboutissement d’un processus ou rupture radicale ?’ in Catalogue of the Exhibition Les Soldats de l’Éternité, l’Armée de Xi’an, Paris 2008 April 15th–September 4th, Pinacothèque de Paris, 217–227. Thierry, François (2013), La Ruine du Qin, Ascension, triomphe et mort du premier empereur de Chine, Paris: La Librairie Vuibert. Thierry, François (2015), “Archéologie et numismatique, Les cinq découvertes qui ont bouleversé l’histoire monétaire du Qin ”, in Wolfgang Szaivert, Nikolaus Schindel, Lichael Beckers and Klaus Vondrovec (éd.), TOYTO APECH TH XWPA Festschrift für Wolfgang Hahn zum 70. Geburtstag,Veröffentlichungen des Instituts für Numismatik und Geldgeschichte 16,Vienne, 433–451. Wang, Bin 汪斌 and Xie, Shaoshi 謝少石 (2003), ‘Dui Anhui faxian baijin sanpin de chubu renzhi’ 对安 徽发现白金三品的初步认䛊, Shaanxi qianbi yanjiu wenji, 4: 55–61. Wang, Helen (2004), Money on the Silk Road, The Evidence from Eastern Central Asia to c. ad 800, including a catalogue of the coins collected by Sir Aurel Stein, London: The British Museum Press. Wang, Jingsheng 王进生 (1984), ‘Shanxi Tunliu chutu yipi Zhanguo biqian’ 山西屯留出土一批战国币 钱, Wenwu 1984-XII: 91. Wang, Qingzheng 汪慶正 (1965), ‘Shiwu nian yilai gudai huobi ziliao de faxian he yanjiu zhong de ruogan wenti’ 十五年以來古代貨幣資料的發現和研究中的若干問題, Wenwu 1965-I: 26–36. Wang, Qingzheng 汪慶正 (ed.) (1988), ‘Xian Qin huobi 先秦貨幣’, in Ma Feihai 馬飛海 (ed.), Zhongguo lidai huobi daxi 中國曆代貨幣大系, vol. I, Shanghai: Shanghai Renmin publ. Wang, Xuenong 王雪农 and Liu, Jianmin 刘建民 (2005), Banliang qian yanjiu yu fazhan 半两钱研究与发 现, Peking: Zhonghua shuju. Wang, Yü-Ch’üan (1951), Early Chinese Coinage, New York: American Numismatic Society, Museum notes. Wei, Juxian (1936) 衛聚賢, ‘Zhongguo zui gu de huobi-Bei ke’ 中國最古的貨幣－貝殼, Guquanxue II-1936: 18–33. Wu, Zhenfeng 吳鎮烽 (1985), ‘Banliang qian ji qi xiangguan de wenti’ 半兩錢及其相關的問題, Zhongguo qianbi lunwenji I-1985: 166–185. Wu, Zhenfeng 吳鎮烽 (1987), Xi Zhou jinwen xieying 西周金文擷英, Xi’an: San Qin publ. Wu, Zhenfeng 吳鎮烽 (1989), ‘Dengcheng Potou Xi Han zhuqian yizhi zhi wo jian’ 澄城坡头西汉铸遗 址之我见, Shaanxi jinrong 1989–142(XII): 19–25.

365

François Thierry Wu, Zhenlu 吴振禄 (1987), ‘Jinguo shigui zuofang yizhi fajue jianbao’ 晋国石圭作坊遗址发掘简报, WW 1987-VI: 73–81. Xunzi 荀子 (by Xun Kuang 荀況), transl. by Kamenarovic, I, Xunzi (Sinn Tseu), Paris: Le Cerf, 1987. Yantielun 鹽鐵論 (by Huan Kuan 桓寬, 81 bc), edited by Wang Liqi, Yantielun jiaozhu 鹽鐵論校注,Tianjin: Tianjin guji chubanshe, 2 vol., 1983. Zhang, Chi 張弛 (1992), ‘Sankongbu kaobian’ 三孔布考辨, Zhongguo qianbi lunwenji II-1992: 74–84. Zhang, Chi 張弛 (1997), Zhongguo daobi huikao 中国刀币汇考, Shijiazhuang: Hebei renmin chubanshe. Zhang, Shouzhong 張守中 (1960), ’1959 nian Houma Niucun gucheng nan Dong-Zhou yizhi fajue jianbao’ 1959 年侯马牛村古城南东周遗址发掘简报, Wenwu 1960-VIII-IX: 11–14. Zhang, Tianen 张天恩 (1991), ‘Dong Zhou lieguo beihua de kaocha’ 东周列国贝化的考察, Zhongguo qianbi 1991-II: 27–33/37. Zhao, Xiaoming 赵晓明 (2003), ‘Xi Han baijin sanpin xiangguan wenti de tantao’ 西汉白金三品相关问 题的 探讨, Shaanxi qianbi yanjiu wenji 4: 41–54. Zhou, Weirong 周卫荣 (2004), Zhongguo gudai qianbi hejin chengfen yanjiu 中国古代钱币 合金成分研究, Peking: Zhonghua shuju. Zhou, Weirong 周卫荣 and Dai, Zhiqiang 戴志强 (2002), Qianbixue yu yezhu shi luncong 钱币学与冶铸 论丛, Peking: Zhonghua shuju. Zhu, Huo 朱活 (1983), ‘Gu bi san tan (1)’ 古幣三談, Zhongguo qianbi 1983-I: 10–17. Zhu, Huo 朱活 (1984), Gu qian xin tan 古錢新探, Dezhou: Qi Lu shushe. Zhu, Huo 朱活 (1985), ‘Gu bi xu tan’ 古幣續談, Zhongguo qianbi 1985-II: 5–13. Zhu, Huo 朱活 (1991), Gu qian xin dian 古錢新典, 2 vol., Xi’an: San Qin publ.

366

17 WOMEN IN EARLY CHINA

ANNE BEHNKE KINNEYWOMEN IN EARLY CHINA

Views from the archaeological record Anne Behnke Kinney

Scholars studying women in early China have produced many fine works that concentrate on a wide array of sources from traditionally transmitted texts (for example, Goldin 2000; Raphals 1998). These studies depend on works from the received tradition, specifically on texts transmitted from the early and mid-Eastern Zhou through the Former and Latter Han and which provide a wide range of information about women and gender in this period. But for decades now, Chinese scholars have published a staggering array of reports on archaeological finds dating from the Neolithic period through the early Eastern Zhou that offer new insights. Evidence drawn from the material remains of earlier periods is not subject to the kinds of anachronisms that riddle texts from the received tradition. It is only in recent times that scholars have begun publishing a large number of studies in English concerning the material culture of early China so that these resources are now available to the general reader. This chapter will draw upon this research to explore how material culture from early China illuminates women’s roles in the opening phases of Chinese civilization. The random nature of what has been preserved over the past seven millennia does not, however, present an even or comprehensive record. The earliest strata of evidence come from the period before the advent of writing, so that the meaning and significance of these materials necessarily involves a great deal of scholarly speculation and controversy. Even after the adoption of writing, our understanding of very ancient texts is still confounded by archaic forms of writing and fragmentary texts that scholars still struggle to understand. At stake as well is the deciphering of historical “truth” presented by ancient texts that possess the same kinds of biases, exaggerations and outright falsehoods that we find in any form of writing. Still, it is very much worth our while to familiarize ourselves with the astonishing range of artefacts that have not seen the light of day for thousands of years, relics that provide new narratives about women in the earliest phases of China’s development.

The Neolithic period With no written records from China’s Neolithic period (ca. 6500–1500 bce), scholars must rely on discoveries of Neolithic settlements, tools, ritual objects, graves and human remains for clues about women at the dawn of Chinese civilization. Over the course of the fifth and fourth millennium bce, a climatic event called the Mid-Holocene Climatic Optimum ushered 367

Anne Behnke Kinney

in a warmer and wetter climate. Under these favorable weather conditions, agriculture became the primary means of producing food. As populations grew in size, Neolithic settlements across China increased and spread. These conditions changed, however, and were followed by cooler and drier weather, forcing communities to expand beyond the margins of their settlements and to experiment with new forms of social and economic organization to cope with changes affecting food supplies caused by fluctuations in the ecosystem (Li and Chen 2012: 169–171). As these settlements expanded and as once-isolated areas began to interact, they began to exhibit increased social, political and economic complexity and stratification. Given the vast area under consideration, i.e., the contemporary political borders of China, artifacts, burial sites and other mute remains from the cultures of China’s early Neolithic period scattered across the Ordos and the northwest, in the northeast, along the Central Plain and in the south reveal distinctive regional and interregional variation. Given the random nature of the evidence and the extensive area under examination, it is no surprise that scholars continue to debate the precise significance of the archaeological record for our understanding of the various cultures that flourished during this very early period of China’s development and the ways in which each of them expressed gender. Earlier and now largely rejected Marxist theory posited the trajectory of social evolution from an egalitarian matriarchal/matrilineal society to a stratified patriarchal/patrilineal structure – an idea that formed the starting point for many early studies of gender in the Chinese Neolithic (Pearson 1988: 25–26; Shelach 2004; Wang 1985–1987). In contrast, archaeologists in China have recently focused first on mortuary evidence to draw conclusions about women and gender in this early period. But, as Ian Hodder cautions, burial practices are not a mirror image of social practices (Hodder 1982: 10). The symbolism of mortuary culture may thus distort, aggrandize or mystify relations of inequality between groups or classes (Pearson 1982: 112). In his study of Chinese Neolithic sites, Richard Pearson notes a further consideration: “Differences in numbers of objects, in burial arrangements, or in the specific combinations of objects may have different meanings in different contexts” (Pearson 1988: 16). Questionable criteria used in earlier studies to sex skeletons has also prompted questions about whether human remains cataloged in various reports have been correctly identified as male or female. Archaeologists have nevertheless identified the following characteristics that might tell us something about women’s social status in Neolithic times: the relatively high numbers of male burials in some areas in contrast to smaller numbers of female burials, which might signal the practice of female infanticide or that women were less likely to receive formal burial; the patterns of contiguous burials and spatial subgroups within cemeteries that might represent social groups among the living; the grave size, elaboration of coffins and labor expended that may signal social status; the burial position of the deceased (prone, supine, flexed, etc.) in graves with more than one body, which in certain configurations might suggest the suttee-like practice of “following in death,” or other manifestations of social power or subordination according to gender. Scholars have suggested that in cases where women’s bodies face away from the settlement and men’s bodies are positioned to face it, natal ties may have been used to establish social identity. Other conjectures include the practice of secondary burials (later reburial of remains that have already been interred), which might indicate group membership or continued influence of the deceased on the living; and the presence and number of grave goods, where the proportion of ornaments and tools, such as axes in men’s graves and spindle whorls buried with women, might indicate the gendered division of labor (Keightley 1999: 20; Pearson 1981: 1086; Pearson 1988: 2–14; Sun and Yang 2004: 33–40; Wu 2004: 87–88). David Keightley cautions against global generalizations, because “All burials in early China were local, and the treatment of women in mortuary ritual often varied by region.” Still, recent 368

Women in early China

scholarship suggests that by late Neolithic times, women in China may have suffered a decline in status (Pearson 1981: 1086; DeBaine-Francfort 1995: 13). The greater number of graves and grave goods for males suggests that social structures were becoming increasingly male dominated, and increasing differences between rich and poor burials indicate that discrepancies between a powerful, wealthy elite and more ordinary folk were widening. Scholars who study women in other civilizations have also noted the relationship between the introduction of intensive agriculture, often associated with the appearance of the plow, and the decline of women’s social status (Lerner 1986: 30; 49–54). It may therefore be significant that the apparent decline in women’s social standing in late Neolithic China (ca. 3000–2000 bce) occurs at a time when intensive farming began to spread in certain areas of China. Other scholars have suggested that the change was not due to the transition to agriculture but to sedentism. “The initial accumulation of property (artifacts), the need for social norms to regulate the community’s life . . . are all related to people’s ability and willingness to reside in permanent year-round settlements” (Shelach 2004: 125). Finally, as an example of how the manifestation of gender varied, sometimes dramatically, from region to region, we should note the discovery of an early Neolithic site that was part of what is now referred to as the Hongshan culture, centered in western Liaoning province and eastern Inner Mongolia, dating to ca. 4000–2500 bce. The site is unique for its nude female figurines as well as its failure to conform to theories about the formation of complex societies as occurring in one place over a long period of time. In contrast to other early sites, Hongshan’s antecedents seem to lay hundreds of kilometers to the east (Nelson 2014: 79–80). The remains of monumental tombs, evidence of craft specialization and a large array of tools found at Hongshan, including what may be a stone plow, all point to social and economic complexity (Nelson 2000: 5–8). Most notable for the study of gender in early China, a temple complex excavated there, radiocarbon-dated to ca. 3500 bce, revealed fragments of a life-sized terracotta statue of a young woman or goddess, as well as small “Venus statues” – clay figurines portraying rotund, sometimes pregnant, women (Nelson 2000: 3). While little else resembling these figurines has been found elsewhere in China, similar statues of great antiquity were found throughout Europe dating from the Upper Paleolithic to the Bronze Age (ca. 25,000–2000 bce) (Xibei daxue wenbo xueyuan kaogu zhuanye 2000: 109–111). The precise function and significance of the Chinese specimens remains unknown, though scholars have conjectured that such figures may represent emblems of fertility cults, “mother goddesses,” or priestesses (Nelson 2014: 81–82). Concerning the European examples, Margaret Ehrenberg writes that “Whichever interpretation is preferred, the dominance of female representation over male, even where the forms are not uniquely female, must be significant” (Ehrenberg 1989: 76). While the precise role of women in Hongshan culture remains enigmatic, with all due caution we can make several generalizations about gender across China in the Neolithic period. In contrast to the general egalitarianism that is suggested by evidence from early Neolithic burials, in terms of both wealth and gender, much mid- and late Neolithic mortuary evidence points to (1) the sexual segregation of labor; (2) the inferior social status of women in relation to men; and (3) the inclusion of women in ritual activity and ancestral cults. Thus, in some Neolithic sites, spindle whorls (a disc or spherical object fitted onto the spindle to increase and maintain the speed of the spin) tend to be buried with women, whereas other tools, such as stone chisels and adzes, are found in the graves of men. Some scholars interpret this evidence as demonstrating that, at least in some places, weaving was designated as women’s work. In a significant number of Neolithic sites, male skeletons greatly outnumber female skeletons. This evidence may point to female infanticide, to better care provided for male rather than female children, or at least to a cultural context in which the burial of women did not warrant the 369

Anne Behnke Kinney

same ritual attention given to men. Nevertheless, because a significant number of female skeletons have been found buried with objects such as beads, jade bracelets, pots, ivory combs and spindle whorls, it seems that even if women were less frequently the recipients of such ritual treatment at death, they were clearly included in religious observances. Furthermore, remains that show women receiving secondary burial shows that their memory was being preserved and, perhaps, that the worship of ancestresses as well as ancestors was already developing in the Neolithic period.

The earliest historical records: Shang paleographic texts With the discovery of written records dating to ca. 1200 bce, from China’s bronze age, the historian who seeks to understand the role of women in early China fares only slightly better. The primary source of information comes from oracle inscriptions found on bones discovered little more than a century ago and which date to the end of China’s first historically documented dynasty – the Shang (ca. 1600–1050 bce). Shang oracle-bone inscriptions are often enigmatic; the script is archaic, at times utilizing graphs with no modern equivalents; and many of the recovered texts are fragmentary. Though scholars have made tremendous progress in deciphering these difficult texts, current understanding of the inscriptions is far from complete. Nevertheless, oracle bones include many details of a fairly straightforward nature about the royal women of the polygamous Shang kings. After the discovery of oracle bones, scholars were surprised to learn that the king list of the Shang as found in the received tradition corresponded closely to the names of kings mentioned in the oracle bones. In contrast, the names of Shang queens were not transmitted, and it is only with the discovery of Shang bone-texts that we learn some of their names. Several hundreds of personal names are found in oracle-bone texts, and 170 belong to women. In some cases, the names document marital relations, most often between a king and his numerous consorts (Chou 1970–1971: 373). Keightley conjectures that polygamy “encouraged the development of values that stressed the husband’s authority and conferred sufficient authority on the royal patriarch for the maintenance of family order” (Keightley 1999: 31). Inscriptions concerning divinations made about the pregnancy and childbirth of aristocratic women also suggest the depressed status of women through a preference for male progeny: “Crack-making on jiashen (day 21), Que divined: ‘Fu Hao’s childbearing will be good.’ . . . After thirty-one days, on jiayin (day 51), she gave birth; it was not good; it was a girl’ ” (Keightley 1999: 34). Notably, both polygyny (a variation on polygamy that allows for only one wife but numerous concubines) and the preference for male offspring are social practices that shaped Chinese families through the end of the imperial period in the early twentieth century. Nevertheless, the oracle bones also record information about the role of elite women in positions of leadership. These texts show women concerned with the harvest; the presentation of tribute items such as turtle shells, ivory, horses, dogs and slaves to the royal court; the supervision of sacrifices to the ancestors, and most interestingly, their participation in military campaigns (Chou 1970–1971: 368–369). Women’s political power in the Shang state may have been derived in part from their roles in the forging of political alliances through marriages that joined Shang nobles with the daughters of rulers of neighboring polities across the north China plain. For example, a court diviner named Xi can be tentatively linked with a Lady Xi (Fu Xi), and both personages may be linked to a place called Xi (Keightley 1999: 33). Nevertheless, Campbell shows that the names of the majority of women mentioned in oracle-bone inscriptions do not include place names and therefore seem to represent women from within the royal domain (Campbell 2007: 156–158). 370

Women in early China

The most striking figure among the royal women is Fu Hao, the consort of Shang King Wu Ding (r. ca. 1200–1181 bce). The wealth of precious objects found in her tomb at Anyang stunned the world. There were larger and almost certainly richer burials, but Fu Hao’s tomb remained undisturbed until archaeologists excavated the site in 1976.The contents include some 195 bronze vessels, 271 weapons, tools and other small bronze objects, 755 jades and 564 carved bone objects, as well as items in marble, turquoise and other stones. Fu Hao was also equipped with a supply of 499 bone hairpins for her use in the other world (Wang 2004: 102). Because Fu Hao can be identified as a person mentioned in the oracle-bone inscriptions, her burial is distinguished as a historical resource of great value. According to the inscriptions, she was a leader of military campaigns, entrusted with court duties and took charge of specific rituals. One passage also describes Fu Hao as the recipient of an exorcism: “Crack-making on jimao (day 16), Que divined: ‘In performing exorcism for Fu Hao to Father Yi (Wu Ding’s father), we cleave a sheep and a pig and promise ten penned sheep.’ ” Another mentions her role in warfare: “Crack-making on xinsi day, Zheng divined: ‘In the present season, if the king raises men and calls upon Fu Hao to attack against the Tufang, we will receive assistance in this case’ ” (Keightley 1999: 32). The inscriptions also suggest that Fu Hao had authority to raise troops herself: “Crack-making on yiyou (day 22), Que divined, ‘Call upon Fu Hao to first raise men at Pang.’ ” The inscriptions also include numerous fretful inquiries made by King Wu Ding concerning periods in which Fu Hao was absent from the royal court at Anyang. Apart from her role in military campaigns and collection of tribute items, it is not clear what activities or duties took Fu Hao (and other royal consorts) away from Anyang. The apparent freedom of movement granted to them, which seems linked to various state responsibilities, is nevertheless striking. While some scholars have conjectured that some of Wu Ding’s consorts possessed and perhaps ruled over their own settlements, existing evidence is insufficient to make such a claim (Campbell 2007: 155). Still, we can conjecture that the peregrinations of the royal consorts must have contributed to the strengthening of social interactions between Anyang and outlying areas. Fu Hao was also the subject of King Wu Ding’s mantic inquiries concerning her illnesses. The oracle-bone inscriptions indicate that after her death, sacrifices were made to her spirit on a regular schedule, an honor apparently reserved for the formally recognized consorts of kings on the main line of descent or as spouses of ancestors who received cult and who had furthermore produced surviving sons. (Keightley 1999: 35; Campbell 2007: 154–157). In her postmortem existence, Fu Hao was given the title Bi Xin (Ancestress Xin) or Mu Xin (Mother Xin). The oracle-bone inscriptions recording the king’s prayers for progeny suggest that fertility was the province of the Shang ancestresses. One states, “Divined on xinsi (day 18): ‘We will pray for progeny to Ancestress Geng and Ancestress Bing and offer a bull, a ram, and a white boar’ ” (Keightley 1999: 42). Still, the ritual observances placed the royal consorts in a subordinate position by situating sacrifices to them at the very end of the ten-day ritual cycle, by utilizing “inauspicious” days, by offering them only four of the five rituals received by kings, and at least in one instance, offering prayers to a female ancestor only after divining about it first in the temple of her royal husband. Keightley also observes that while the kings made frequent reports to the ancestors, few were directed to an ancestress, and most ancestresses appear not to have been granted their own temples (Keightley 1999: 43–44). Archaeological remains reveal Shang society as highly stratified, with a king at its apex. Women of Fu Hao’s social status therefore formed part of a small and highly privileged elite. Monumental and lavish tombs point not only to a great concentration of wealth controlled by a select few but also to a large pool of ordinary people who supplied the labor and resources for elite consumption. The fate of women who occupied less-privileged positions than Fu Hao 371

Anne Behnke Kinney

in the Shang social hierarchy can be illustrated by the two female “accompaniers-in-death” or “death/grave attendants” (among four men and ten others of indeterminate sex) found adjacent to Fu Hao’s coffin, who had presumably been killed to accompany Fu Hao to the other world. As one scholar has described the process: Death attendants . . . are the most extreme instantiation of the bonds of obligation in the Late Shang. For one’s death and place in death to be so totally tied to another is surely an instantiation of extreme dependency (whether figured as transfiguring loyalty or the crushing burden of duty). At the same time, while a place in the entourage of an exalted ancestor may have been a better choice than the obscurity of a low status burial and though death attendants may have made willing, even joyous self-sacrifices, the practice is nonetheless a stark testament to the fixity of social place within the relationships of kinship, dependency and patronage determining life, death and after life. . . . The elite were thus, not simply due a disproportionate share of the common wealth . . . but also to unequal claims on the lives and deaths of those of lesser worth. (Campbell 2007: 263–264) In contrast to death attendants, sacrificial victims, often decapitated, and some with hands still bound, tended to be male, though by the late Shang, they were mainly adult females and children (Bagley1999: 192; Keightley 2012: 70). While early texts sporadically record instances of human sacrifice and the use of death attendants, confronting an interred victim’s remains shocks the viewer with an immediacy and palpable force that vague and infrequent textual references can never achieve. No text from the received tradition had prepared the first Chinese archaeologists to excavate the royal Shang capital for the horrific scale of Shang human sacrifice. Evidence for human sacrifice has been found for earlier periods, for example, in the remains of the Erligang civilization, which preceded the Shang. But it reached unprecedented levels in the reign of Wu Ding (Campbell and Steinke 2014a: 23; Campbell 2014b: 94; Bagley 1999: 192–194). Though inscriptions suggest that most often it was men rather than women who were used as sacrificial victims, some scholars have interpreted that rituals aimed at eliciting rain or preventing floods may have used women as the preferred victims in these rites. Others, however, argue that although the rituals involved women, their specific roles in these procedures are not clear (Wang and Kubin 2007: 104). Still, the inscriptions record that Ancestress Geng, the mother of King Wu Ding, received some seventy human victims after her death and that women were included among those sacrificed (Wang and Kubin 2007: 174–175). The evidence, as narrowly focused and cryptic as it is, suggests that while elite women formed an integral part of the ritual system of the Shang, their status was lower than that of their male counterparts. Royal marriage was polygamous, and male progeny was preferred over female, even as some women, such as Fu Hao and others, enjoyed what appears to have been considerable freedom of movement, active participation in military affairs and collaborative roles in matters of state. Kinship organization in the Shang appears to have been patrilocal, patrilineal and patriarchal, while marriage was likely exogamous. It is also worth remembering that Shang was only one of a number of large important polities in China at this time, but as the sole source of written records, it provides an important though partial account of gender roles in this period of China’s development.

Western Zhou dynasty While the Zhou overthrew the Shang, the Zhou maintained some Shang practices such as oracle-bone divination. Scholars identify one inscription as a record of a sacrifice made by 372

Women in early China

King Wu, the Zhou founder, to Shang ancestors, suggesting that the pre-conquest Zhou kings were related to the Shang by marriage. The inscription is generally understood to document the king’s sacrifice of two “surrendered women” as part of a “cauldron exorcism” (Shaughnessy 1985–1987: 157). This enigmatic inscription documents various kinds of continuity between the Shang and the Zhou, demonstrating marital links between the two dynastic houses, the Zhou’s continued use of human sacrifice (though the practice survived in very much diminished form) and the habit of recording divinations on bone. Few bone inscriptions from the Zhou survive, however, and the bulk of paleographic evidence from the Zhou utilizes the medium of bronze. Western Zhou bronze texts are valued as the only reliable contemporary source of information for the period. Scholars past and present have questioned the authenticity of most traditionally transmitted documents that purport to date from the early Western Zhou. While a very few sources – several chapters from the Shangshu, for example – appear to be consistent with paleographic sources that can be confidently dated to the period, we must still rely on epigraphic evidence – inscriptions on bronze vessels – to verify information such as names, dates and locations. In contrast to the Shang period, Zhou bronze inscriptions tend to be longer and more plentiful. Still, as highly specialized texts used by the elite to address and honor ancestors, to commemorate the honor of an audience with the king, the marriage of a daughter, or the conclusion and settlement of legal agreements, the view of women afforded by these texts is still highly limited. And, as Martin Kern cautions, “Couched in relentlessly eulogistic diction and, if necessary, undisturbed by historical facts that contradicted their own account, these are the texts with which an ancient community created its common narrative of memory and agreed upon common identity” (Kern 2009: 152–153). Women’s names and activities are recorded far less frequently than men’s in bronze texts but appear most often in what have come to be called “dowry inscriptions” (Shaughnessy2004). David Sena, however, suggests that while bronze vessels were objects of great value, their function was not so much a dowry representing an exchange of wealth from the bride’s family to the groom but rather as a means for the bride to maintain a religious and social connection with her natal family through worship of her own family’s ancestors (Sena 2005: 299; von Falkenhausen 2006: 119). The presence of women in the ancestral cult and their contributions towards dynastic strength is also documented in other kinds of bronze inscriptions (Khayutina 2014; Cao 2004; Cook 2006: 74). In the Western Zhou, in addition to bronze vessels cast to commemorate the marriage of a daughter, inscribed vessels might also be used to honor a wife, a mother, or a queen. Maria Khayutina, for example, has shown that elite wives, as representatives of their natal families, were also key in achieving political stability throughout the Western Zhou and Spring and Autumn times: “On the one hand, marital alliances helped to consolidate the radial network of Zhou states centered on the Zhou king. On the other hand, they facilitated the construction of decentralized regional and inter-regional inter-state networks” (Khayutina 2014: 39; Pulleyblank 2000: 1–12). Khayutina also notes that epigraphic evidence shows that the rule of surname exogamy, that is, the requirement to marry outside of one’s own clan, was generally observed and moreover “created the basic precondition for alliance building across geographic space” (Khayutina 2014: 43).Von Falkenhausen, in his discussion of clan exogamy, notes the important distinction between clans (xing) and lineages (shi): Though lineages constituted the basic units of social organization, clans were a higher unit of descent reckoning, each clan comprising a large number of lineages. . . . Each marriage represented lineage alliances across clan boundaries (not, as is sometimes 373

Anne Behnke Kinney

stated, inter-clan alliances, for clans themselves were not units of political or economic organization). (von Falkenhausen 2006: 117–118) These marriage-based alliances also contributed to the prestige of offspring and could be utilized for support in the face of invasion or other diplomatic crises that crossed political boundaries. Bronze inscriptions also confirm that by late Western Zhou times, male elites expressed filial piety not only toward their own parents but toward their parents-in-law as well, and that after marriage, women continued to perform sacrifices to members of their own lineages.The importance of a wife’s lineage is also documented in transmitted texts. The Zuo zhuan, for example, states: “When a prince comes to the rule of a State, he shows his affection for his maternal uncles, cultivates all relationships by marriage, and takes a head wife, to attend to the grain-vessels of the temple. This is filial piety, and filial piety is the beginning of propriety” (Legge 1970: 235; Wen 2). By demonstrating the importance of the maternal line, this practice must have bolstered the status of mothers and wives.Thus the importance of daughters is reflected in inscriptions that, in contrast to the Shang examples, state wishes for both male and female offspring (Khayutina 2014: 71). Although inscriptions that focus on women and their kin are far outnumbered by those that were made by and for men and their own lineages, they demonstrate the very practical ways in which elite women promoted dynastic well-being (Khayutina 2014: 69). In addition to inscriptional evidence, scholars have studied the role of gender in elite Western Zhou burial practices. In his study of objects buried with elite members of society, Lothar von Falkenhausen has described Western Zhou tombs as containing “the kinds of paraphernalia the occupant would have needed to perform the ritual duties corresponding to his or her social rank,” because “in his/her new capacity as a revered ancestor, a deceased person was thought to continue in the performance of his ritual duties to his/her ancestors.” Newly created ancestors thus remained “very much a part of society . . . kept alive . . . through the continuing worship by latter-day descendants” (von Falkenhausen 2006: 298–299). For the living, bronze ritual vessels were used in communal sacrificial meals held in ancestral temples for ancestors who were thought to descend from the other world to join participants and who were deemed able to provide the living descendants with supernatural assistance. One famous bronze tureen, the Dong Gui (ca. 956 bce), is inscribed with a long prayer of thanks to the donor’s deceased mother, who provided other-worldly protection to him in battle: Dong led the Master Chiefs in pursuit of the Rong to expel them from Yu forest, giving the Rong and the Hu a beating. My Accomplished Mother aggressively and persistently cleared the way, gracing his heart with openness, eternally covering his body and enabling him to meet the enemy and capture one hundred ears, two bound prisoners [of higher status], 135 military items, including shields and spears, dagger-axes and bows, quivers and arrows, and armor and helmets, and 114 Rong captives. . . . At the completion of the attack, no harm had come to Dong’s body.Your son, Dong, claps his hands and knocks his head on the ground in gratitude and in response extols the Accomplished Mother’s Auspicious Blazing presence and takes the opportunity to make a precious tureen for expressing reverence to [his] Accomplished Mother . . . so that you can cause your son, Dong, to live ten thousand years and to use this vessel to express reverence by presenting mortuary and sacrificial offerings from dawn to dusk to his Accomplished Mother. May his sons of sons and grandsons of grandsons use and treasure it. (Based on Cook and Goldin 2016: 69) 374

Women in early China

Notably, while females ancestors were honored at sacrificial feasts, an ode from the received tradition,“Thorny Caltrop,” which narrates an idealized mid-to late Western Zhou ancestral sacrifice and concluding banquet, suggests that living women participated in feasts, but only in the presentation and removal of food from the banquet (Kern 2009: 174, 176, 182; 2000: 85; Cook 1995: 249). Other inscriptions show bronze vessels cast for daily personal use in private domestic quarters, such as Cui pan, a face- or hair-washing vessel cast by a son for his mother, inscribed with the words: “Cui makes for his royal mother Kuishi this face/hair-washing gui vessel. May Kuishi have abundant age and for ten thousand years use it” (von Falkenhausen 2011: 296). Scholars speculate that ritual reforms completed in the mid-ninth century bce imposed what appear to be strict sumptuary rules across the Zhou cultural sphere.These reforms may represent a reorganization of elite society to restore order among the eminent lineages whose ranks had swelled over time and become increasingly distant from the trunk lineages from which they sprang. Especially in the period after these reforms, we see a number of differences in burial practices that tend to divide along gender lines. Bronze weapons and musical instruments, which are present in men’s tombs, are largely absent from the tombs of women. Men are also sometimes buried with female companions who may have served them in life as concubines, while women have no such companions. While the tombs of both men and women at times included sacrificial victims, women tended to be buried with fewer victims. Further gendered evidence derived from tomb remains suggest that wives were most often buried with ritual assemblages associated with one rank below that of their husbands, reflecting a lower status at least in terms of mortuary treatment. Still, in some cases, a high-ranking woman would have been buried with vastly more items than males of the next lower-rank. Men’s tombs tend to contain more chariots and chariot fittings, whereas the tombs of women include more ceramic kitchen vessels (Sena 2005: 180, 195; von Falkenhausen 2006:111–112, 122–123). Bronze inscriptions also hint at the possible existence of hierarchies of female officials, and though the evidence requires further research, von Falkenhausen suggests that “the situation of females may have been somewhat less starkly subaltern than is suggested by the consideration of sumptuary rules and sets of ritual vessels alone” (von Falkenhausen 2006: 124–125). Some surprising details about elite women have emerged as a result of inscriptional evidence. One inscribed vessel, for example, suggests that upon the death of a woman’s husband, the role of family head was transferred to her and not to her son or other male family member (Skosey 1996: 174; Khayutina 2014: 79). A vessel known as the Ci zun (Ci goblet) shows a woman, Gong Ji, supervising men in an official capacity: “In the first period of the second month on the day dingmao, Gong Ji ordered Ci to supervise the men of the fields. Praising the accomplishments of Ci, she presented him with a horse and a fur jacket. Raising in thanks the grace of Gong Ji; thus was cast this precious vessel” (Eno 2010). Another vessel, the Zuoce Xuan you (Recorder Xuan’s pot) dated to the reign of King Cheng (1042–1006 bce) shows the queen attending to state business in the absence of the king: “In the nineteenth year when the King was at An, Queen Jiang charged Recorder Xuan to calm the Elder of the Eastern Yi. The Elder of the Yi received Xuan as a guest, bestowing upon him cowries and cloth. Raising up Queen Jiang’s grace in thanks, Xuan thereby cast this precious sacrificial vessel for his late patterned-Father Gui” (Eno 2010: 13). The following inscription, the Yin Ji li (The steamer of Lady Yin of the Ji clan), dated to the era of King Zhao (977–957) shows the queen praising the valor of a woman named Yin Ji: Mu Gong had a clan shrine for Yin Ji built in the woods of Yu. In the sixth month in the period of the waxing moon on the day yimao, the gracious Royal Consort, not forgetting how Mu Gong had served the former King with sagacity and illumination, descended to the clan shrine of Yin Ji in the woods of Yu. The Royal Consort praised 375

Anne Behnke Kinney

the valor of Yin Ji and presented her with five objects of jade and four horses. Bowing prostrate and raising in thanks the grace of the Royal Consort; thus was cast this precious qi-cauldron. (Eno 2010: 27) As Robert Eno notes, One of the most troublesome phrases in the inscriptions is the one rendered here as “praising valor.” The term . . . seems obviously connected with merit on the battlefield, yet in this instance, it appears to be applied to a woman, raising the question of whether, as with Fu Hao in the Shang, the consorts of warlike nobles did not themselves sometimes play a role in military affairs (something that would have been unthinkable a few centuries later). (Eno 2010: 27) Nevertheless, others have rendered the term Eno translates as “valor” using vaguer words of praise such as “merit” (Cao 2004: 82–85). Still, inscription on the Jin Jiang ding (tripod of Jiang of Jin), though likely cast in the early Eastern Zhou (ca. 746 bce), notes the military exploits of Lady Jiang after the death of her husband, a man identified as Lord Wen of Jin (r. 780–746 bce): Lady Jiang of Jin said, “I succeeded my former aunt as the Lordess of the principality of Jin. I do not stay in leisure and reckless tranquility. I adjusted and harmonized my illustrious virtue and propagated my plans in order to glorify and to accompany my lord. Every day I promote his glorious merits. I am pious and do not retreat. I wisely hold the Capital Garrison and rule over our ten thousand people. . . . I have not neglected Lord Wen’s shining mandate. I went through and greatly punished the Yun of Fantang. I took their auspicious metal and used it to make this treasured sacrificial tripod. I will use it to make peace: gently receive and take care of various lords far away and near.” (Khayutina 2014: 81) In contrast to the virtues attributed to women, one bronze inscription, the Qiu ding (cauldron), dated to the reign of King Xuan of Zhou (r. 827/25–782 bce), shows the king’s exemplary treatment of women in his concern for widows. This inscription foreshadows a theme common in later traditionally transmitted texts associated with the Confucian school concerning the state’s concern for vulnerable women. The inscription includes the following warning: “Do not enrich yourself, for if you do then there will be bribes and indulgences at the expense of widowers and widows resulting in blame on me” (Sena 2005: 79).

Eastern Zhou Scholars note another shift in ritual behavior from the mid-Spring and Autumn period (ca. 600 bce) onward, when the focus of ritual appears to have moved from ancestors as the main beneficiaries of the sacrifices, who in return would offer support to living descendants, to achieving harmony and status in the living community. This new emphasis is in accord with the later Confucian ideal of ritual as means to affirm the social order, to practice self-cultivation and to emphasize sincerity of feeling over expenditure of valuable resources. This shift also opened the way for the participation of more ordinary people, who previously had been excluded from ritual practices predicated on the possession of eminent ancestors in the other world. But in 376

Women in early China

spite of the inclusion of people of lower status, Eastern Zhou tomb furnishings show a greater gender disparity, with women buried in tombs that were much smaller and less well equipped with goods in comparison to the burials of their husbands, which might point to a concomitant reduction in their social status (von Falkenhausen 2006: 357). In a further alteration of earlier practice, by the fifth century bce, fine bronze vessels come to be used in temple sacrifices, while the dead were buried with inferior or miniaturized “spirit vessels” (mingqi). Tombs were no longer equipped with exquisite vessels for ancestral sacrifices but built to resemble domestic spaces furnished with replicas of all of the necessities and often luxuries of everyday life, emphasizing a new discontinuity between the living and the dead and, perhaps, the wish that the comfort of the tomb would confine the deceased, now perceived not as helpful ancestors who could aid the living from the other world to potentially harmful specters who were harmful to the living (von Falkenhausen 2006: 297–301; Cook 2011: 67–69). Still, even with the introduction of spirit vessels, in which replicas replaced actual items, women (and men) continued to be buried in the tombs of their masters as death attendants. The tomb of Marquis Yi of Zeng, dated to 433 bce, for example, includes four chambers, each representing a palace compound, including the harem and his private quarters. Twenty-three young women were found encoffined in these two chambers (von Falkenhausen 2006: 306; Wu 1999: 723, 733–734). As Campbell has noted, “To be killed and be interred in the grave of another is all the more striking in its radical subordination of being made . . . permanent beyond death” (Campbell 2014b: 101). Mortuary evidence from the first part of the Eastern Zhou shows that the practice of burying sacrificial victims or death attendants, while not common, continued throughout the period. Lothar von Falkenhausen has distinguished a hierarchy of women’s mortuary treatment in relation to the tombs of the prominent men to whom they were attached. The three gradations include women who appear to have been ordinary victims, victims in their own coffins entombed with their masters and full wives, who did not serve as victims and who died of natural causes and were then buried in separate tombs. The burial of husband and wife in adjacent tombs appears to have been a privilege of high rank. In one tomb, a deceased male was buried with a “concubine” in her own coffin along with seven sacrificial victims, whereas his wife had no encoffined companion and only two ordinary victims (von Falkenhausen 2006: 94, 117, 121–122, 358). The practice of retainer sacrifice, though later condemned in texts such as the Xunzi (ca. 265 bce), and largely discontinued toward the end of the Warring States period, persisted in modified form through the imperial period in the related practice of widow suicide (Watson 1963: 105; Campbell 2014b: 103; de Groot 1976: 721–752). In considering what might be vastly different motives that prompted women to follow their husbands in death, in each specific example, we might ask, did death attendants die willingly out of love and loyalty or were they essentially forced to die as subject to the power of the deceased? Did they assume that by dying with the deceased, they would be afforded a more elevated position in the afterworld? Was the act memorialized in a way the conferred a kind of earthly immortality? Were their families rewarded for the sacrifice? (Morris 2014: 62, 82–84). While there are surely many instances in which these questions can be answered in the affirmative, clearly there must also have been women who were coerced to die. Once again, this practice, whatever the motive, points to the extreme subordination of wives (or concubines) to husbands, in which women are deemed socially dead after the death of their husbands. Not surprisingly, a term of self-designation used by a widow in the Zuo zhuan (ca. early fifth-fourth century bce) was “the one who is not yet dead” (wei wang ren) (Legge 1970: vol. 5, 371). In context, however, the term takes on an almost ironic meaning as the widow in question went on to have a rather public affair with the head of the Shusun lineage and worked 377

Anne Behnke Kinney

with him to oust other power holders. Elsewhere, the same text also recounts the gratitude of a father whose daughter was allowed to forgo dying with her husband when it became clear that her spouse had only made the request in the throes of dementia (Legge 1970: vol. 5, 328). Elsewhere, however, the Zuo zhuan records another act of gratitude in which a man whose father’s life was spared by the king sent his two living daughters to die with the ruler upon his death (Legge 1970: vol. 5, 649). It is also worthwhile considering a story (in spite of its late date) included in the early first century ce collection Lienü zhuan (Categorized Biographies of Women) that recounts the decision of one woman to follow her husband in death (Kinney 2014: 91). The story concerns the Lady of Yue, a concubine of King Zhao of Chu (r. 496–465 bce). While on a pleasure outing with the king and another concubine called the Lady of Cai, the king asks the two women if they will die with him. The Lady of Cai readily agrees, and the king asks his scribe to record her promise – an interesting detail that suggests the binding nature of such a promise. When the king asks the Lady of Yue, she replies, My own ruler . . . did not agree that I should die with you. I have heard from my aunts that a woman dies in order to honor the benevolence of her lord and to increase devotion toward him. I have never heard that recklessly following an unenlightened lord in death will redound to his glory. I dare not obey your order. (Kinney 2014: 92) Later, when the king mends his ways, the lady of Yue tells him, How great is the king’s virtue! I am now willing to follow Your Majesty. On our earlier outing, your greatest devotion was to the pursuit of pleasure, so I dared not agree. Now that Your Majesty has returned to ritual propriety, all of the capital populace will be willing to die for you. So how much more must someone like me be willing to do the same? Please allow me to go before you and drive away the fox of the underworld! The king said, “If you really intend to die, it will only magnify my lack of virtue.” The Lady of Yue said, “I have heard that the righteous do not make empty promises. I will die for the king’s righteousness; I will not die for his pleasure.” She then killed herself. (Kinney 2014: 92–93) Her speech suggests that in some cases, the agreement to follow a husband in death was made between a woman’s father and her husband at the time of her marriage and that a woman might have no say in the matter unless she later volunteered to die even though her father might not have made such an agreement. The story also adds that the Lady of Cai was unable to die when the time came, which carries the possibility that when a woman made the offer herself she might renege on the agreement. Finally, the story, though ideologically biased and perhaps composed after the Eastern Zhou, tells us something about what the victim hoped to achieve by such an act, which in this case was to glorify the reputation of the ruler and to help clear away dangers from the deceased’s progress through the underworld.The story ends with the selection of the Lady of Yue’s son as the heir to the throne based on the premise that “If the mother is faithful, the child will be benevolent.” While this story is almost certainly apocryphal – it is not found in earlier texts, and furthermore closely resembles another tale where a loyal male official rather than a concubine offers to follow his ruler in death – it is still a striking example of the sort of behavior still lauded as exemplary in early imperial times. So far, our understanding of women in early China has, by necessity, relied upon archaeological materials. But by the time of the Eastern Zhou (770–256 bce), we finally have access to 378

Women in early China

a large array of transmitted texts that can be reliably ascribed to this period and which, though limited in scope, tell us far more about women, especially concerning the moral and behavioral standards they were expected to emulate. Some of the earliest accounts of women are found in the Book of Odes (compiled ca. sixth century bce; text stable ca. late fourth century bce). Of interest is the presence of women in Zhou dynastic legend, as represented in the Book of Odes who function as both positive and negative exemplars. The most famous example is the ode “Si Zhai,” which praises Tai Jen, the mother of the Zhou dynastic founder, King Wen (d. ca. 1050 v), and Tai Si, the mother of King Wu, victor over the Shang dynasty: Pure and reverent was Tai Ren, The mother of King Wen Loving was she to Zhou Jiang, A wife becoming the House of Zhou. Tai Si inherited her excellent fame, And from her came a hundred sons. (Legge 1970: vol. 4, 446) Tai Ren and Tai Si are thus praised for their virtue and fertility. In contrast, the Odes also identify feminine behaviors destructive to the health of the dynasty, singling out, for example, malefactors such as Bao Si, the woman credited with causing the collapse of the Western Zhou (Legge 1970: 4.318). Still, the directive for women to model themselves on the example of Tai Ren can be traced back to a tradition of ancestral model emulation seen in bronze inscriptions as early as the mid-Western Zhou (Savage 1992: 11). A set of Eastern Zhou bronze vessels dated to 524 bce include inscriptions made by a ruler of Cai in honor of her marriage to the king of Wu. Here the princess is directed to pattern herself on King Wen’s mother by performing what appears to be a ritual dance imitating the founding king’s mother in order to receive her divine aid (Cook 2009: 260–261; 2005: 15–16; Cao 2004: 220–224). Another dowry vessel, the Jingong basin, dated to ca. 578 bce and cast by a Jin patriarch, marks the occasion of his daughter’s coming-of-age ceremony and marriage to a husband from the state of Chu. In the inscription, her father orders her to exhibit proper decorum: ‘Even though now you are but a youth, manage and govern your appearance, be without blame for ten thousand years so that Jin’s reputation will soar” (Cook 2011: 321; Lei 2013: 31). As Cook explains, A study of the ritual rhetoric preserved on Western Zhou bronze inscriptions shows that the Way . . . of the former kings so popular during the Eastern Zhou developed out of the Western Zhou tradition of “following the model” (shuaixing) of ancestral lineage founders by youth. . . . By the late tenth century bce, music had clearly become integral to the ritual display of successful modeling during feasts and sacrifices . . . when all the ancestral spirits were perceived as watching and adjustments were made in rank . . . The ritual display, called “Awesome Decorum” (weiyi), involved prescribed movements of “grasping” the ancestral de (i.e., power or life force), a symbolic process perceived as “opening up the heart (or body)” of the “youth.” (Cook 2011: 308–309) In addition to the perceived benefit of maintaining continuity with and eliciting aid from the other world, the practice of ancestor worship and emulation of their models facilitated social 379

Anne Behnke Kinney

reproduction and political stability; it also marked one’s status within one’s lineage and promoted a sense of shared identity. Through the performance of sacrificial rites, descendants were called upon to embody, perpetuate and enhance the virtue or merit (de) that their ancestors had accumulated across generations. But bronze texts tell us little more about the education of women. Eventually, as the erosion of political stability and the fragmentation of lineage groups accelerated through the Spring and Autumn and Warring States periods, the commemoration and emulation of role models through elite ritual performances eventually gave way (at least for men) to the inculcation of exemplary models and cultural knowledge through the memorization and recitation of the Book of Odes and the Book of Documents among a larger, more diverse sector of society in ritual, sacrificial or diplomatic contexts (Cook 2011: 334; Kern 2000: 69). With the waning of bronze culture after the mid-Warring States period, bronze ritual vessels bearing inscriptions, including those that mention women, also dwindle in number and in the length of the inscribed texts (Cao 2004: 251; So 1995: 68–71). Still, there is evidence from both the archaeological record and that of traditionally transmitted texts suggesting that elite women might have also been privy to instruction bearing on their social roles as mothers, wives, sisters, daughters and daughters-in-law that had some connection to the Odes (though far less likely, given its emphasis on rulership, with the Documents). The Zuo zhuan shows at least one woman citing an ode in a diplomatic exchange concerning the marriage of her daughter ca. 583 bce, while the Guoyu preserves a second example (Legge 1970: 5.371; Zhou 2003: 3). The much later Han dynasty Mao Commentary (mid-second century bce) also facilitates exactly this sort of application by supplying the alleged historical context of each ode, which in many cases involves discussion of the moral behavior of the women supposedly featured in the ode (Kinney 2012). An awareness of how women shape the political sphere is also demonstrated in the Xinian, a ca. 300 bce bamboo text that forms part of the collection of texts donated to Qinghua University in 2008. Although this text was almost certainly not intended for female consumption, nearly a quarter of its entries recount the vital roles played by women in the unfolding of decisive events (Milburn 2016: 53–109; periscopes 2, 5, 6, 9, 14, 15). Though evidence, such as the covenant texts of Houma and Wenxian, show that literacy began to spread as early as the fifth century bce, it is still unclear how many women were able to read at this time. One tantalizing example of a text that seems to have been written for women comes from a bamboo text dated to ca. 350–300 bce, entitled Jizi’s Instructions for Women. The text, which is in poor condition, concerns the preparation of sacrificial food (Allan 2015: 56–57). Naturally, a text written for women does not necessarily prove that women were able to read the text themselves, but the title, which indicates the text’s focus on educating a female audience, is itself significant as it represents the earliest known work directed at women. Sarah Allan has also raised questions about the possibility that the recently discovered bamboo texts, known as the Shanghai Museum Collection, came from the tomb of a woman. The texts, which had been excavated illegally, appeared for sale in Hong Kong in 1994. While their provenience is unknown, based on their similarity to the Guodian texts, scholars conjecture that they might have come from the looted tomb in nearby Guojiagang. Archaeologists generally dismiss Guojiagang as the possible source of the texts because the tomb belonged to a woman. Allan, however, cites four factors for considering the Guojiagang tomb as the source: the record of known tomb robberies in the area, the appearance of the bamboo texts on the market soon after the Guojiagang robberies were committed, the absence of evidence of other robberies in the area at the time, and the excellent preservation of the remaining artifacts in the tomb consistent with the good condition of the bamboo slips of the Shanghai Museum collection (Allan 2015: 51–56). Allan notes that there is nothing innately feminine about the contents of the collection. The texts cover an eclectic range of topics on history, philosophy, and on specific texts such 380

Women in early China

as the Odes and the Yijing. Unlike Western Zhou mortuary practice, in which tombs represent the social and official rank of the deceased, artifacts buried with the dead in the Warring States period by contrast often included far more personal items. Allan notes that “If a Warring States period woman were to have had the elite level education implicit in a collection of such manuscripts, this would very likely have been the most significant aspect of her life and character” (Allan 2015: 57–58). Unfortunately, it is currently impossible to do more than speculate about its origins. There is little evidence to suggest that women received training in reading and writing, apart from the rather specialized services provided by female scribes (nüshi), bureaucratic posts attached to the office of the queen as described by the Zhouli (Ruan 1980: 690).There are, however, occasional references to literate women.Yanzi, of the eponymous Yanzi Chunqiu (ca. third century bce), mentions that one of his concubines is able to read and write (Milburn 2016: 202). Still, as in the case of earlier examples of women leading armies or heading households, there have always been women in elite families who occasionally occupied themselves with activities that were generally the province of men. Several traditionally transmitted texts dating from the later Warring States through the early imperial period (such as the Yili, the Liji and the Zhouli) contain scattered passages that focus on the behavior and ritual duties of women. The Zhouli (Rites of Zhou), for example, a prescriptive text that purports to describe the structure of royal government in the Zhou dynasty, lays out the bureaucratic duties of officials assigned to educate the royal women through the “rites for women” as well as through a women’s curriculum (fuxue) that included the virtue, speech, bearing and work appropriate to women, though the text provides no details beyond these categories (Ruan 1980: 684, 687). Numerous accounts from the Zuo zhuan show women following (or ignoring) what appears to be a well-developed code of feminine behavior (Kinney 2013; Li 2007: 147–160). But apart from the brief and topically circumscribed Jizi’s Instructions for Women, no other texts or even titles of texts setting out this code in a sustained, detailed, and focused manner have been transmitted from the pre-Qin period. In 2010, Peking University obtained a cache of Qin (221–206 bce) bamboo slips of unknown provenance that included an untitled text that was assigned the name Jiao Nü (Educating Women) (Zhu 2015: 5–15). Educating Women, comprised of 851 characters and written on fifteen bamboo strips, is largely cast in four-character rhyming phrases that depict both good and bad behavior for women, and in tone, resembles the much later Nü Jie (Instructions for Daughters) of Ban Zhao (ca. 45–116 ce) (Swann 2001: 82–99). In form, scope and sophistication, both Jizi’s Instructions for Women and the Jiao Nü differ from the earliest received text, specifically devoted to the moral training of women, namely, the Lienü zhuan (Categorized Biographies of Women) of Liu Xiang (79–8 bce). The Lienü zhuan brings together a wide variety of ritual rules concerning women, but always through the vehicle of the exemplary biography, merging the practice seen in bronze inscriptions of modeling oneself on illustrious women of the past and (in smaller measure) the kind of specific ritual information and pointers for women’s conduct in daily life that inform Jizi’s Instructions for Women and the Jiao Nü. As archaeological work continues apace, we may well see the discovery of previously unknown documents that bridge the gap between the relatively short and easily memorized texts of Jizi’s Instructions for Women and the Jiao Nü and the lengthy and complex narratives of the Lienü zhuan.

Conclusion Archaeological information concerning women in early China, while far more limited than the vast array of writings found in the received tradition, places us squarely in the specific time we 381

Anne Behnke Kinney

seek to study.Though the accounts may be biased, as in the case of bronze inscriptions extolling the perfection of the Zhou founding matriarchs, these texts and objects afford us the privilege of seeing precisely how these people represented themselves without the distorting lens of much later accounts. And while the meaning of mute specimens of material culture may confound scholars, there is an undeniable thrill in seeing, for example, a necklace that belonged to Fu Hao, or in learning, for the first time, the name of a Shang royal consort, or experiencing the complicated emotions that arise when considering the sixteen people who were forced (or perhaps who gladly agreed) to accompany her in death. The picture that emerges from these artifacts is in itself incomplete and perplexing, yet ongoing archaeological activity promises far more information about women in early China as excavation and analysis continues apace. The remains examined in this essay span thousands of years and ranges across the enormous land mass that now comprise China’s borders. Still, we can make a few cautious conclusions about what the materials have to say about women. From Neolithic times we see the sexual segregation of labor; the inferior social status of women in relation to men; the inclusion of women in ritual activity and ancestral cults; and a cultural context in which the burial of women did not warrant the same ritual attention given to men even though they were clearly included in religious observances. By Shang times, written evidence suggests that royal marriage was polygamous and male progeny was preferred over female even as some elite women, such as Fu Hao, enjoyed active participation in military affairs and collaborative roles in matters of state. Kinship organization in the Shang appears to have been patrilocal, patrilineal and patriarchal, while marriage was likely exogamous. Western Zhou evidence conforms to these patterns but develops notions of ancestresses as role models. Bronze inscriptions document the critical importance of marriage alliances in promoting political stability and the formation of inter-state networks. Bronze texts also confirm that by late Western Zhou times, male elites expressed filial piety not only towards their own parents, but towards the parents of their wives as well, and that after marriage, women continued to perform sacrifices to members of their own lineages. As in the Shang sources, Western Zhou bronze texts portray a small number of elite women in positions of power and leadership, even from the other world, as one text records a prayer from a son, not to his father, but to his mother, acknowledging her supernatural aid that protected him in battle. Further gendered evidence derived from tomb remains suggest that wives were most often buried with ritual assemblages associated with one rank below that of their husbands, reflecting a lower status at least in terms of mortuary treatment. In the Eastern Zhou, tomb furnishings show an even greater gender disparity, with women buried in tombs that were much smaller and less well equipped with goods in comparison to the burials of their husbands. Still, archaeological remains, however, represent largely elite practices.We must not forget that, like the woman found buried in a Zhou dynasty refuse pit, not all people were granted a cemetery burial (von Falkenhausen 2003: 445). One Eastern Zhou bronze inscription provides a fascinating account of how a woman performed a sacrificial rite calling down the spirit of King Wen’s mother in an effort to embody, perpetuate and enhance the virtue or merit (de) accumulated by her ancestress across generations. This inscription supplies an important link to later texts such as the Lienü zhuan, which opened up an entire “pantheon” of female exemplars for all women to model themselves upon, not just those who claimed descent from illustrious dynastic matriarchs. Finally, while texts conveying instructions for women’s conduct, such as Jizi’s Instructions for Women and the Jiao Nü, may have been designed for oral transmission to an illiterate female audience, we see evidence, tentative as it is, that might indicate the presence of a small number of women who were trained to read. As archaeologists continue to recover and study large quantities of new evidence, we can only hope that new finds in the archaeological record will help complete our fragmentary view of women in the early phases of China’s history. 382

Women in early China

Works cited Allan, S. (2015) Buried Ideas: Legends of Abdication and Ideal Government in Early Chinese Bamboo-Slip Manuscripts. Albany: State University of New York Press. Bagley, R. (1999) ‘Shang Archaeology,’ in M. Loewe and E. Shaughnessy (eds.) The Cambridge History of Ancient China from the Origins of Civilization to 221 B.C., Cambridge: Cambridge University Press. Campbell, R. (2007) ‘Blood, Flesh and Bones: Kinship and Violence in the Social Economy of the Late Shang,’ unpublished thesis, Harvard University. Campbell, R. (2014a) ‘Erligang: A Tale of Two Civilizations,’ in K. Steinke and D. Ching (eds.) Art and Archaeology of the Erligang Civilization. Princeton: Princeton University Press. Campbell, R. (2014b) ‘Transformations of Violence: On Humanity and Inhumanity in Early China,’ in R. Campbell (ed.) Violence and Civilization: Studies of Social Violence in History and Prehistory, Providence, RI: Joukowsky Institute Publication 4. Cao, Zhaolan 曹兆蘭 (2004) Jinwen yu Yin, Zhiu, Nüxing wenhua 金文與殷周女性文化, Beijing: Beijing Daxue chubanshe. Chou, H. (1970–1971) ‘Fu-X Ladies of the Shang Dynasty,’ Monumenta Serica XXIX: 346–390. Cook, C. (2009) ‘Ancestor Worship in the Eastern Zhou,’ in J. Lagerwey and M. Kalinowski (eds.) Early Chinese Religion, Part One: Shang Through Han (1250 bc-220 ad) Leiden: Brill. ——— (2006) Death in Ancient China:The Tale of One Man’s Journey, Leiden: Brill. ——— (2011) ‘Education and the Way of the Former Kings,’ in F. Li and D. Branner, (eds.) Writing and Literacy in Early China, Seattle: University of Washington Press. ——— (2005) ‘Moonshine and Millet: Feasting and Purification Rituals in Ancient China,’ in R. Sterckx (ed.) Of Tripod and Palate: Food, Politics and Religion in Traditional China, New York: Palgrave MacMillan. --------- (1995): ‘Scribes, Cooks and Artisans: Breaking Zhou Tradition,’ Early China 20:241–277. Cook, C. and Goldin, P. (2016) A Source Book of Ancient Chinese Bronze Inscriptions, Berkeley: The Society for the Study of Early China. ——— (2006) Death in Ancient China：The Tale of One Man’s Journey, Leiden: Brill. Cook, S. (2012) The Bamboo Texts of Guodian: A Study and Complete Translation, Ithaca: Cornell University East Asia Program. Crump, J. (1979) Chan-kuo Ts’e, San Francisco: Chinese Materials Center. DeBaine-Francfort, C. (1995) Du Néolithique à l’Âge du Bronze en Chine du Nord-Ouest: La culture de Qijia et ses connexions. Mémoires de la Mission Archéologique Française en Asie Centrale, volume VI, Paris: Éditions Recherche sur les Civilizations. De Groot, J. (1976) The Religious System of China, vol. 2, book 1, Taipei: Ch’eng Wen Publishing. Ehrenberg, M. (1989) Women in Prehistory, Norman, OK and London: University of Oklahoma Press. Eno, R. (2010) ‘Inscriptional Records from the Western Zhou,’ www.indiana.edu/~g380/3.10-WZhou_ Bronzes-2010.pdf Goldin, P. (2000) ‘The View of Women in Early Confucianism,’ in L. C. Li (ed.) The Sage and the Second Sex: Confucianism, Ethics, and Gender, Chicago: Open Court. Hodder, I. (1982) Symbols in Action: Ethnoarchaeological Studies of Material Culture, Cambridge: Cambridge University Press. Hubei sheng wenwu kaogu yanjiusuo 湖北省文物考古研究所 (2000) Jiudian Chu jian 九店楚簡, Beijing: Zhonghua shuju. Keightley, D. (1999) ‘At the Beginning: The Status of Women in Neolithic and Shang China,’ Nan Nü 1.1: 1–63. ——— (2012) Working for His Majesty, Berkeley: Institute of East Asian Studies. Kern, M. (2009) ‘Bronze Inscriptions, the Shijing and the Shangshu,’ in J. Lagerwey and M. Kalinowski (eds.) Early Chinese Religion: Part One: Shang Through Han (1250 bc-220 ad), Leiden: Brill. ——— (2001) ‘Ritual Text, and the Formation of the Canon: Historical Transitions of Wen in Early China,’ T’oung Pao LXXXVII: 43–91. ——— (2000) ‘Shi Jing Songs as Performance Texts,’ Early China 25: 49–111. Khayutina, M. (2014) ‘Marital Alliances and Affinal Relatives (sheng and hungou) in the Society and Politics of Zhou China in the Light of Bronze Inscriptions’, in Early China 37: 15–38. Kinney, A. (2014) Exemplary Women of Early China: The Lienü zhuan of Liu Xiang, New York: Columbia University Press. ——— (2012) ‘The Mao Commentary to the Book of Odes as a Source for Women’s History,’ in Ho, C. (ed.) Overt and Covert Treasures: Essays on the Sources for Chinese Women’s History, Hong Kong: The Chinese University of Hong Kong.

383

Anne Behnke Kinney ——— (2013) ‘A Spring and Autumn Family,’ Chinese Historical Review, 20. 2: 113–137. Legge, J. (1970) The Chinese Classics, 5 vols., Hong Kong: Hong Kong University Press. Lei, C. ‘Antiquarianism without Antique: A Study of the Jin Gong Basin 晉公盆 and Its Methodological Significance’, Paper presented at the annual conference of the Society for the Study of Early China in San Diego, CA, March 21, 2013. Lerner, G. (1986) The Creation of Patriarchy, New York: Oxford University Press. Li, L. and Chen, X. (2012) The Archaeology of China from the Later Paleolithic to the Early Bronze Age, Cambridge: Cambridge University Press. Li, W. (2007) The Readability of the Past in Early Chinese Historiography, Cambridge: Harvard University Asia Center. Linduff, K. and Rubinson, K. (2008) Are All Warriors Male? Gender Roles on the Ancient Eurasian Steppe, Lanham, MD: AltaMira Press. Linduff, K. and Sun,Y. (2004) Gender and Chinese Archeology, Walnut Creek, CA: AltaMira Press. Milburn, O. (2016a) The Spring and Autumn Annals of Master Yan, Leiden: Brill. ——— (2016b) ‘The Xinian,’ Early China 39: 52–109. Morris, E.F. (2014) ‘(Un)Dying Loyalty: Meditations on Retainer Sacrifice in Ancient Egypt and Elsewhere,’ in R. Campbell (ed.) Violence and Civilization: Studies of Social Violence in History and Prehistory, Providence, RI: Joukowsky Institute Publication 4. Nelson, S. (2000) ‘Ritualized Pigs and the Origins of Complex Society: Hypotheses Regarding the Hongshan Culture,’ Early China 20. ——— (ed.) (2014) The Hongshan Papers, Oxford: Archaeopress. Pearson, M. (1982) ‘Mortuary Practices, Society and Ideology: An Ethnoarchaeological Study,’ in I. Hodder (ed.), Cambridge: Cambridge University Press. Pearson, R. (1988) ‘Chinese Neolithic Burial Patterns,’ Early China 13: 1–45. ——— (1981) ‘Social Complexity in Chinese Coastal Neolithic Sites,’ Science 213.4: 1078–1086. Pulleyblank, E. (2000) ‘Ji and Jiang: The Role of Exogamic Cans in the Organization of the Zhou Polity,’ Early China 25:1–27. Raphals, L. (1998) Sharing the Light: Representations of Women and Virtue in Early China, Albany, State University of New York Press. Ruan,Y. 阮元 (1980) Shisanjing Zhushu 十三經注疏, Beijing: Zhonghua shuju. Savage, W. (1992) ‘Archetypes, Model Emulation, and the Confucian Gentleman,’ Early China 17: 1–26. Sena, D. (2005) ‘Reproducing Society: Lineage and Kinship in Western Zhou China,’ Unpublished thesis, University of Chicago. Shaughnessy, E. (1991) Sources of Western Zhou History: Inscribed Bronze Vessels, Berkeley: University of California Press. ——— (2004) ‘Western Zhou Hoards and Family Histories in the Zhouyuan,’ in Yang, X. (ed.) New Perspectives on China’s Past: Chinese Archaeology in the Twentieth Century, vol. 1, New Haven:Yale University Press. ——— (1985–1987) ‘Western Zhou Oracle Bone Inscriptions,’ Early China, 11–12: 146–194. Shelach, G. (2004) ‘Marxist and Post-Marxist Paradigms for the Neolithic,’ in Linduff and Y. Sun (eds.) Gender and Chinese Archeology, Walnut Creek, CA: AltaMira Press. Skosey, L. (1996) ‘The Legal System and Legal Tradition of the Western Zhou (ca. 1045–771 b.c.e.),’ Unpublished thesis, University of Chicago. So, J. (1995) Eastern Zhou Ritual Bronzes from the Arthur M. Sackler Gallery, New York: Arthur M. Sackler Foundation in association with the Arthur M. Sackler Gallery, Smithsonian Institution. Sun, Y. and Yang, H. (2004) ‘Gender Ideology and Mortuary Practice in Northwestern China,’ in Linduff and Y. Sun (eds.) Gender and Chinese Archeology, Walnut Creek, CA: AltaMira Press. Swann, N. (2001). Pan Chao: Foremost Woman Scholar of China. Ann Arbor: Center for Chinese Studies, the University of Michigan. Wang, N. (1985–1987) ‘Yangshao Burial Customs and Social Organization: A Comment on the Theory of Yangshao Matrilineal Society and Its Methodology,’ Early China 11–12: 6–32. Wang, P. 王平 and Kubin, W 顧彬. (2007) Jiaguwen yu Yin Shang Ren Ji 甲骨文与殷商人祭, Zhengzhou: Daxiang chubanshe. Wang, Y. (2004) ‘Rank and Power among Court Ladies at Anyang,’ in K. Linduff and Y. Sun (eds.) Gender and Chinese Archeology, Walnut Creek, CA: AltaMira Press. Watson, B. (1963) Hsun Tzu: Basic Writings, New York: Columbia University Press.

384

Women in early China Wu, H. (1999) ‘Art and Architecture in the Warring States Period,’ in M. Loewe and E.L. Shaughnessy (eds.) The Cambridge History of Ancient China from the Origins of Civilization to 221 B.C., Cambridge: Cambridge University Press. Wu, J. (2004) ‘The Late Neolithic Cemetery at Dadianzi, Inner Mongolia Autonomous Region,’ in K. Linduff and Y. Sun (eds.) Gender and Chinese Archeology, Walnut Creek, CA: AltaMira Press. von Falkenhausen, L. (2006) Chinese Society in the Age of Confucius (1000–250 bc):The Archaeological Evidence, Los Angeles: Cotsen Institute of Archaeology at UCLA. ——— (2011) ‘Royal Audience and Its Reflection in Western Zhou Bronze Inscriptions,’ in F. Li and D. Branner (eds.) Writing and Literacy in Early China, Seattle: University of Washington Press. ——— (2003) ‘Social Ranking in Chu Tombs: The Mortuary Background of the Warring States Manuscript Finds,’ Monumenta Serica 51: 439–526. Xibei daxue wenbo xueyuan kaogu zhuanye 西北大學文博學院考古專業 (2000) Fufeng Anban yizhi fajue baogao 扶風案板遺址發掘報告, Beijing: Kexue chubanshe. Zhou,Y. (2003) ‘Virtue and Talent: Women and Fushi in Early China,’ Nan Nü 5.1: 1–42. Zhu, Fenghan 朱鳳瀚 (2015) ‘Beida Cang Qinjian Jiao Nü Chushi,’ 北大藏秦簡《教女》初識 Beijing Daxue Xuebao (Zhexue Shehui Kexueban) 北京大學學報 (哲學社會科學版) March, 52.2: 5–215.

385

18 AN OVERVIEW OF THE QIN-HAN LEGAL SYSTEM FROM THE PERSPECTIVE OF RECENTLY UNEARTHED DOCUMENTS KYUNG-HO KIM AND MING-CHIU LAIAN OVERVIEW OF THE QIN-HAN LEGAL SYSTEM

Kyung-ho Kim and Ming-chiu Lai Introduction* It is the best time ever for the study of early Chinese legal history. Studying the history of law before the Han dynasty used to be in relative obscurity compared to its Tang and Qing counterparts due to the scarce, fragmentary and not-always-reliable primary sources. For example, when writing his Remnants of Han Law, the late A.F.P. Hulsewé felt regret that the sources for the period preceding the Han dynasty are “usually too insecure to provide us with more than vague generalisations, uncertain moreover in point of time.”1 Fortunately, this situation has greatly improved after the publication of the Qin legal documents unearthed in Shuihudi, Hubei, in the 1970s, which has forever changed the landscape of early Chinese legal history. For the first time we can reach some undisputed legal texts that are free from later abbreviations and editions. With the help of the Shuihudi texts and the complete disclosure of the Zhangjiashan Han legal manuscripts published in 2001, the study of early Chinese legal history has been coming into its own over the last four decades.2 To better understand the intrinsic meanings and legal thought and practice embedded in these difficult texts with abundant obsolete technical jargon, generations of scholars have found ways to tackle them, resulting in countless minor textual revisions and interpretations.3 Moreover, several annotated translations concerning these legal texts had also been complied, either in modern Chinese,4 or foreign languages ranging from English,5 to Japanese6 to German.7 The Remnants of Ch’ing Law by A.F.P. Hulsewé, the first complete annotated translation of the Qin legal texts from Shuihudi tomb no. 11, for instance, is widely regarded as a classic and still serves as an important reference to anyone who is interested in early Chinese law. Also, at the time this tiny introductory chapter is written, the prolonged project of translating the legal manuscripts from Zhangjiashan tomb no. 247 to English by Anthony Barbieri-Low and Robin Yates has finally reached completion and has been published. The colossal monograph has two volumes, over 1500 pages.Volume one mainly consists of a set of introductory articles that do not merely address the principles and methodology of the translation but, rather, aim at providing an overview to the Han legal system and its practice. The second volume is the annotated translations 386

An overview of the Qin-Han legal system

of the Statutes and Ordinances of the Second Year (Ernian lüling 二年律令) and Book of Submitted Doubtful Cases (Zouyan shu 奏讞書). This voluminous monograph can somewhat be seen as a conclusion to previous studies on the Zhangjiashan texts. After the publication of the Zhangjiashan texts in 2001, however, plenty of new sources have been disclosed. Two of the most important sources are administrative documents discovered from Liye and Qin legal texts collected by the Yuelu Academy from Hong Kong. The former provides us some peerless materials about the function of Qin legal practice, in particular, the criminal proceedings in daily administration, whereas the latter comprises over 1200 slips of Qin statutes, ordinances and precedents (bi 比) dated to 221 bc, that is, after the Qin unification, and thus could be the missing puzzle connecting the Shuihudi and Zhangjiashan texts, which is to say, by using the Yuelu manuscripts, we may examine the development and changes of law codes from the Qin kingdom and the Qin Empire to the early Han period. Because of these splendid new sources, to keep our information up to date, in this chapter, not only will we provide a general introduction about the legal history in early imperial China to our readers, but we are also trying to give a brief sketch of how these new sources will shed light on the study of early Chinese law.

A brief introduction to the unearthed legal materials and their nature The Qin legal texts from Shuihudi tomb no. 11 During the winter of 1975, in the Shuihudi area of Yunmeng 雲夢 County in eastern Hubei, eleven tombs of Qin were excavated, of which the eleventh tomb unearthed 1155 slips, including 80 fragments that consist of large numbers of legal texts. According to the Chronicle (Biannianji 編年紀) of the tomb occupant, his personal name was Xi 喜, he was born in 262bc and died in 217bc. During his lifetime, Xi undertook official positions such as the Commune Scribe (xiang shi 鄉史)8 and Foreman Clerk (ling shi 令史) of Anlu 安陸 County and was later transferred as Foreman Clerk and judicial Scribe (yu shi 獄史) of Yan 鄢 County. Given that the offices that Xi held were related to government administration and criminal lawsuits, to fulfil his responsibilities, he had to acquire respective legal knowledge. Therefore, the legal manuscripts found may be duplicates that Xi had copied and used during his tenure and were buried with him after his death. The legal manuscripts from Shuihudi tomb No. 11 reveal the political situation of the late Warring States China. By the time Xi was born, the state of Chu was under attack by Qin and retreated from the Han River 漢水 region to the Huai River 淮河 basin. In fact, from the twenty-fourth year (223 bc) of King Zheng of Qin 秦王政 to Qin’s final conquest of Chu in 222 bc, Yunmeng County, which lay between the Han and Huai River basins, was Chu’s frontline against the onslaught of Qin. Due to the special political circumstance of the region, a governing system that was different from the normal one adopted among Qin’s core area, that is, the so-called Guanzhong 關中, was implemented, part of which can be seen in passages of both the Chronicle and Document of Instructions (Yushu 語書).9 The content of the legal manuscripts from Shuihudi tomb No. 11 can be summarized as follow: 1

“Eighteen Qin Statutes” (Qin lü shiba zhong 秦律十八種)

The “Eighteen Qin Statutes” were law codes stipulating various fields in society, ranging from agricultural activities, the handicraft industry, assignment of government services and duties, and 387

Kyung-ho Kim and Ming-chiu Lai

the circulation of currencies to regulations in the area of politics. These law codes signified that a centralized governing system was implemented, and those who broke the rule would be punished. Also, through a comparison between the “Statutes on Auditing” among these statues and a more detailed version attached in the Shuihudi manuscripts, it is obvious that the “Eighteen Qin Statutes” were by no means the complete law codes of Qin. 2

“Statutes on Auditing” (Xiao lü 效律)

As stated in (1), part of the “Statutes on Auditing” also appears in the “Eighteen Qin Statutes.” Given that the first slip of these statutes was written with its title “Auditing” on the verso side, we might infer that all of the first and last slips of the scroll contain a separate statute that would bear its own title. “Statutes on Auditing” were mostly administrative regulations that gave thorough stipulations on the checking of public commodities possessed by counties and capital offices (duguan 都官), especially on collations of official accounting records of military equipment such as weapons and armours and on errors concerning measuring utensils. 3

“Miscellaneous Excerpts from Qin Statutes” (Qin lü za chao 秦律雜抄)

The “Miscellaneous Excerpts from Qin Statutes” consist of a variety of different statutes, eleven of which bore their titles. Scholars suggested that the copier was extracting statutes from the Qin statutes to suit his own needs. It is also noteworthy that aside from the “Statutes on Appointing Officials,” which shared a similar title to the “Statutes on the Establishment of Officials” in the “Eighteen Qin Statutes,” other statutes in this part did not replicate those in the “Eighteen Qin Statutes,” and most of them were related to military affairs such as appointment and dismissal of military officers, army training, discipline on battlefields, war-time provision of materials from commoners and, last but not least, reward and punishment after the war. 4

“Answers to Questions Concerning Qin Statutes” (Falü da wen 法律答問)

The “Answers to Questions Concerning Qin Statutes” contain 187 articles which adopted a “question and answer” style to elucidate meanings of terminologies and legal concepts. Therefore, analysing the “Answers to Questions Concerning Qin Statutes,” we might catch sight of the legacy of Li Kui 李悝’s Classic of Law (Fajing 法經) in the subsequent Qin law. In fact, the content of the “Answers to Questions Concerning Qin Statutes” also were reminiscent of the six passages in the Classic of Law, which are Robbery, Assault, Detention, Arrest, Miscellaneous and Composition of Judgments. From the term “Sacrifice of the Duke” that appears in the texts, some of the articles might be dated to the time in which Qin’s ruler had not enthroned himself as the king, and it would not be absurd to say that several articles were established by the time of Lord Shang. In addition, the frequently used jargon “practice of the court” (ting xingshi 廷行事) suggests that usage of precedents as models in lawsuit trying had already become a regular practice. 5

“Models for Sealing and Investigating” (Feng zhen shi 封診式)

The “Models for Sealing and Investigating” were models which judicial officers were obliged to follow when transmitting their investigations into written reports, after they had completed trying or interrogating lawsuits. The models may be compiled from documents of actual cases, only changing the personal names of people involved to heavenly stems (tiangan 天干), thus reflecting the reality of Qin society. 388

An overview of the Qin-Han legal system

The Qin legal texts from Longgang 龍崗 tomb no. 6 In addition to the Shuihudi Manuscripts, other Qin legal texts have been excavated at Longgang in Yunmeng County. The legal texts, which amount to approximately 293 bamboo strips, comprising ten fragments,10 were unearthed from October to December 1989 from tomb no. 6. Among the unearthed strips, three were inscribed with calendar dates, and they are: “Since the 24th year, in the 1st month, on the day jiayin”廿四年正月甲寅以來 (slip 116);“In the 25th year, in the 4th month, on the day yihai 乙亥” (slip 98); and “In the 9th month, on the day bingshen 丙 申.” Given that terms such as “Emperor” (Huangdi 皇帝, see slips 15 and 16) and “Black-headed ones” (qianshou 黔首, see slips 6, 77, 157 and 196) had been adopted and replaced their former usages of “King” (wang 王) and “commoners” (min 民), the Longgang manuscripts likely were copied after the Qin unification in 221bc and thus are dated later than the Shuihudi one. The Longgang manuscripts were mostly statutes, but unfortunately, since they had already been heavily damaged by the time of the excavation, it is difficult to figure out their respective titles. The original editors had categorized the strips into five groups, and they are (1) Statutes on Royal Gardens, (2) Statutes on Highways, (3) Statutes on Horses, cows and lamps, (4) Statutes on farming baskets and (5) others. However, more recent scholarship has advocated that it is the “Statutes on Royal Gardens” that constituted the major part of the manuscripts, and scholars have proposed that the statutes can be correlated to several other statutes in the Shuihudi texts like Statutes on Agriculture, Statutes on Stables and Royal Gardens, Statutes on Government Service and the Major of Government Carriages’ Statutes of Hunting. Therefore, the tomb occupant who owned the manuscripts might be an officer who was responsible for administrative affairs in a royal garden.

Statutes and Ordinances of the Second Year from Zhangjiashan Tomb no. 247 The Statutes and Ordinances of the Second Year (Ernian lüling 二年律令) from the Zhangjiashan tomb no. 247 were legal texts in the early Western Han dynasty. In the early Han period, the Zhangjiashan Han cemetery was located near 1.5 kilometres northwest of Jiangling 江 陵 County, now Jingzhou District 荊州區 of Jingzhou Municipality 荊州市, Hubei. Bamboo strips have been found in tombs no. 247, 249, 258, 327 and 336 of the Han cemetery, while only the first and last tombs mentioned unearthed legal materials. Still, we do not know the name of the occupant of tomb no. 247. Nor could we ascertain his title and identity. Nevertheless, from the entry of “dismissed due to illness” (bingmian 病免) in the Calendar (lipu 曆譜), the homogeneity of archaeological findings between Zhangjiashan tomb no. 247 and Fenghuangshan 鳳凰 山 tomb no. 10 and the large number of legal manuscripts excavated from the tomb, we might surmise that he was a low-ranking judicial official. Slips unearthed from tomb no. 247 were divided into nine groups by the editors, and the Statutes and Ordinances of the Second Year were scattered mainly among group C and F, with few samples from group I and other fragments.11 Based on the example of the Yan shun 引書, which includes slips from both group E and I, it is possible that slips in group C and F originally belonged to the same scroll but later dispersed due to underground pressure after burial. Regarding the materiality of the Statutes and Ordinances of the Second Year, it starts with the slip that bears its own title (slip F14) and is followed by 526 slips, each of which is 31 centimetres in length. The texts contain twenty-seven statutes and one ordinance whose titles were written on separate slips and thus differentiated from the main body. Since the Statutes and Ordinances of the Second Year include several statutes stipulating privileges to the House of Lu 呂 (E.g. Slip 85 of the Statutes on the Composition of Judgments) and the Calendar ended at the second year of 389

Kyung-ho Kim and Ming-chiu Lai

the Empress Lu’s reign (186bc), the original editors suggested that the “Second year” in the title refers to the second year of the Empress Lu. In sum, the Statutes and Ordinances of the Second Year are crucial primary sources about various aspects of the Western Han dynasty, including but not limited to sociopolitical, military, economic and geographical disciplines. Examining these texts, we can draw a more comprehensive and lively picture of the ancient Chinese society that the rather restrictive received texts did not tell us. Studying the materials, several research perspectives might deserve our special attention: (1) security at both state and personal level (Statutes on Assault and Robbery); (2) litigation and punishments of crime (Statutes on the Composition of Judgments, Denunciation and Arrest); (3) bureaucracy, such as the appointment of officials and the salary, ranking and succession of rank (Statutes on the Establishment of Officials, Bestowals, Ranks, Salaries and Establishment of Heirs); (4) household registration and enrollment (Statutes on Households and Enrollment); (5) drafting manpower and distribution of goods (Statutes on Exemption from Taxes, Auditing, Levies, Government Service and Finance); (6) transferal of documents and transportation (Statutes on Food Rations at Conveyance Stations and the Forwarding of Documents); (7) economics (Statutes on Agriculture, Cash and Markets); (8) education (Statutes on Scribes); and (9) regulations on passing through the boundaries (Ordinances on Fords and Passes).

Major content of submitted doubtful cases The system of submitting doubtful cases in lawsuits was designated for situations under which the judges, usually magistrates and judicial scribes, were not able to match the accused’s crime with appropriate statutes, ordinances or precedents and thus could not determine his/her punishment. Encountering a doubtful case of this kind, the judges would send the case to their superior and seek his judgement. According to the received texts, the system was established in the seventh year of Emperor Gaozu of Han: 高皇帝七年，制詔御史。「獄之疑者，吏或不敢決，有罪者久而不論，無罪 者久繫不決。自今以來，縣道官獄疑者，各讞所屬二千石官，二千石官以其 罪名當報之。所不能決者，皆移廷尉，廷尉亦當報之。廷尉所不能決，謹具 為奏，傅所當比律令以聞。」 In the seventh year of Emperor Gao, an imperial instruction was directed to the Imperial Secretaries, (reading): “In doubtful judicial cases the officials sometimes do not dare to give a decision, (and so) those who have committed crimes are not sentenced for a long time, (whilst) the innocent stay in detention for a long time without any decision. From now on, when in the offices of the counties or marches, there are doubtful lawsuits, these are always to be submitted to the (official ranking at) two thousand shi to whom (the county or march) resorts. The (official ranking at) two thousand shi taking the (correct) category of the crime, matches it (with the equivalent punishment) and reports [back] a judgment. Those which (the official ranking at two thousand shi) is unable to decide, he should transfer them to the Commandant of Justice; the Commandant of Justice likewise matches (the crime to the punishment) and reports [back] a judgment. Those (cases) which the Commandant of Justice is unable to decide, he should respectfully embody it in a memorial, adding the precedents, statutes and ordinances with which (the crime) is to be matched, in order to inform (the Emperor).”12 The above imperial decision denotes two important details about the system of submitting doubtful cases. First, it narrates the historical context in which the system emerged, ascribing 390

An overview of the Qin-Han legal system

it to the delay of lawsuit due to the inability or timidity of judicial officials.13 Moreover, the principle of submitting these cases to superiors was also addressed, establishing the sequence of submitting from the county or march level jurisdiction: commandery level jurisdiction – Commandant of Justice – Emperor. Although the received texts stated that the system of submitting doubtful cases was set up during the early Han, the aforementioned “Answers to Questions Concerning Qin Statutes” from the Shuihudi manuscripts include several articles that may be relevant to the system. Because of the limitation in space, only the following example will be provided: 有投書，勿發，見輒燔之；能捕者購臣妾二人，毄(繫)投書者鞫審讞之。所謂 者，見書而投者不得，燔書，勿發；投者【得】，書不燔，鞫審讞之之謂殹 (也)。 “When there are ‘thrown letters, these are not to be opened; as soon as they are discovered, they are to be burned. Persons who are able to arrest (the person who threw the letter) are rewarded with two slaves. Detain the person who threw the letter, question him and submit (the case).” (What the statute) means is: when the letter is discovered, whereas the person who threw it is not caught, the letter is to be burned; it must not be opened.When the person who threw it is caught, the letter is not burned; he is to be questioned and the case should be submitted – that is what is meant.14 The article mentioned that after arresting the person who threw the letter, in addition to interrogating the suspect, judicial officials had to submit the case, possibly to their superiors. Even though the article is not directly related to submitting doubtful cases, a similar system had already existed before the imperial decision issued by Gaozu of Han. In short, the Book of Submitted Doubtful Cases excavated together with the Statutes and Ordinances of the Second Year from Zhangjiashan Tomb no. 247 serves as an important reference in studying the legal system of early Han.15 Likewise, Volume three of the recently appeared Qin texts in the possession of the Yuelu Academy 嶽麓書院 contains legal documents of a similar nature to those of the Book of Submitted Doubtful Cases, by which we could finally assess the Qin legal history in a more empirical way, not merely in a purely theoretical way, as before. The remaining sections will provide a brief introduction to these two texts.

The Book of Submitted Doubtful Cases from Zhangjiashan Tomb no. 247 The Book of Submitted Doubtful Cases (Zouyan shu 奏讞書) contains twenty-two cases, four of which were dated to the Qin and two to the Spring and Autumn periods, with the rest all from the early Han era.16 Even though it is uncertain whether these cases faithfully recorded the reality back then, from their detailed description of the criminal procedure involved in a lawsuit, the majority of cases should be reliable sources. Therefore, the Book of Submitted Doubtful Cases grants us an opportunity to understand how state control was actually implemented upon its people, something that cannot be generated from received texts. In other words, since these cases had direct linkage to the social life of commoners from the Spring and Autumn period to the early Han era, they could hence provide us crucial clues in understanding the functioning of governmental system of the state beyond a pure legal sense. The cases complied in the Book of Submitted Doubtful Cases always commenced with a subordinate submitting a difficult case to his superior and asking for his judgement. As such, the documents would end in phrases such as “being in doubt as to what crime (personal name of 391

Kyung-ho Kim and Ming-chiu Lai

the criminal[s]) [is guilty of]” (yi mou zui 疑某罪), or “we dare to submit this [case] to higher authorities for decision” (ganyanzhi 敢讞之). Out of the twenty-two cases compiled in the Book of Submitted Doubtful Cases, seven were about issues concerning servants; four involved lower social class people like hard labourers or ethnic minorities who broke the law; one was about an absconder who did not register his household with the government. In addition, cases 15, 16 and 20 were records of county prefects or other high-ranking officials who committed crimes like corruption or murder. All these together account for over 50 percent of the cases in the Book of Submitted Doubtful Cases and are important in understanding the society, especially that of southern regions during the Qin and early Han era. Also, cases in the Book of Submitted Doubtful Cases reveal that the majority of cases were sentenced in accord with respective statutes, ordinances or precedents but not based on the decision of rulers or certain passages from classics. As such, it seems like the society had gradually transformed from a more kinship-central one to a fairer and more objective world.

Submitted doubtful cases in the Qin manuscripts in possession of the Yuelu Academy17 The third volume of the Qin manuscripts collected by the Yuelu Academy are legal documents from King Zheng’s era. The editor divided the scripts into four parts according to their materiality, writing formats and content. The current title, Four Types of Descriptions Including That of Lawsuits (Wei yu deng zhuang si zhong 為獄等狀四種), was designated because the verso side of three slips from the second part bore titles such as “being a description of a lawsuit” (wei yu ? zhuang 為獄, 狀) or “being a description of a memorial about pleading for a trial” (wei qiju zou zhuang 為气（乞）鞫奏狀). Such nomenclature could, of course, be problematic and, in fact, has already generated some controversy.18 The first type of texts consists of 136 slips, from which seven cases dated between 215 and 222 bc had been reconstructed. Given that each of the seven cases began and ended with the phrase “we dare to submit this [case] to higher authorities for decision,” and phrases like “being in doubt as to what crime (personal name of the criminal [s]) [is guilty of],” or “being in doubt as to how much bounty (personal name of the receiver [s]) [is awarded to]” (yi mouren gou 疑某 人購) were recurrent in the main body of these cases, they seem to originally have been doubtful cases that subordinate organizations submitted to their superiors and thus share the same nature with the first thirteen cases in the Book of Submitted Doubtful Cases from Zhangjiashan,19 just with a much wider spectrum. Indeed, this part includes not only matching and punishment of crimes but also the legal responsibility of minors, treatment for negligence of duties, etc. The actors of these submitted cases were counties and after the end of the submission reports, most of the cases were attached with an “officers’ discussion” 吏議 section, two of which also were supplemented with a section of “report from the commandery.” Part two composes 73 slips, only half the size of the first part, and can be classified into three types of document by content. The first one is cases 9 and 10, in which the magistrate wrote a recommendation letter to a higher jurisdiction on behalf of judicial or foreman scribes who had solved a difficult case, and attached a detailed description of the case so as to ask the commandery to promote the capable scribe as accessory scribe (zushi 卒史) and establish them as role models for other officers. Type two contains two cases (nos. 11 and 12) related to illicit sex, both of which had been appealed by those convicted and were sent by the commandery office back down to the lower jurisdiction. Similar to the seventeenth case in the Book of Submitted Doubtful Cases from Zhangjiashan, both appeals resulted in failure.The rest of the two cases (nos. 8 and 13) were too fragmentary to analyse their dating and formats. 392

An overview of the Qin-Han legal system

Parts three and four each consist of one case. Part three contains twenty-seven slips whose content was a submitted case dated to 225 bc. However, the main difference underlying this case and those in part one is that it preserved the complete date, instead of omitting it, and its “report” section was abbreviated; therefore it is still unclear whether it was a reply from the commandery-level offices or from the Commandant of Justice. The editor advocated that the case in part four bore a different narrative style than other three parts, and it was mainly about how, in a battle against rebels, the Qin army suffered a great loss to the extent that the general who conducted the warfare was killed, and in the battlefield, several soldiers had stepped backward and thus were considered as fleeing from the battlefield because of timidity. While the plot is a bit like the eighteenth case in the Book of Submitted Doubtful Cases, its details remain a mystery.

Penal system and criminal procedure Like modern society, the penal system in the Qin and Han era consists of different levels of punishments in accordance with the severity of crime convicted. Other than just sentencing convicts by a fixed term of incarceration, however, the penal system in Qin and Han constitutes a variety of punishments ranging from the death penalty, mutilation, hard labour, exile and fines, and one of its most distinguishing features is that different punishments can be compounded with each other, thereby forming a more complex level of punishments. For example, mutilations such as tattooing (qing 黥) always served as additional punishments to hard labour, and the punishment of shaving off the beard (nai 耐), too, was closely associated with being forced to be “bond servants or women” (lichenqie 隸臣妾).20 Labour punishments such as being sentenced to be “wall builders and grain pounders” (chengdanchong 城旦舂) and “bond servants or women” were believed to be the cornerstones of the Qin and Han penal system, and punishments such as mutilation (xing 刑) or shaving off the beard (nai 耐) were simply adjunctive to them.21 Recent research, however, starts to challenge this view. For instance, Arnd Helmut Hafner 陶安, in his latest monograph, has tried to revisit the framework of the Qin and Han penal system by arguing that the main foundation of the Qin penal system was not labour punishments but a combination of capital punishment (si 死), corporal punishment (xing 刑) and shaving off the beard.These three pillars later dissolved because of the ramifications of law, and by incorporating capital, corporal and nai punishments with measures such as redemptions (shu 贖) and fines (zi 貲), the degrees of punishment was enriched and, eventually, completed the penal system.22 Although Hafner’s views are not free from problem (for example, his total rejection of hard labour as a kind of punishment does not hold under certain new sources), his effort in reconstructing the framework of the Qin and Han penal system is still worth noting. New sources from volume four of the Yuelu Academy Qin slips provide us several valuable materials in reconstructing the Qin penal system after the unification: ‧索律曰：索有脫不得者，節（即）後得及自出，‧訊索時所居，其死罪，吏 徒部索弗得者，贖耐；城旦舂到刑罪，貲1354二甲；耐罪以下貲一甲。 The Statute on Pursuing Criminals says: those [persons who were] being pursued but escaped and thus were not arrested, once [they] are later caught or surrender voluntarily, [officials should] interrogate [them] about where they lived when being pursued. Should the persons [be guilty of a crime that deserves] capital punishment and the officials and personnel of the jurisdiction [where they were staying] pursued but could not catch them, sentence them to be redeemed by shaving off their beards; 393

Kyung-ho Kim and Ming-chiu Lai

[should the persons be guilty of a crime that deserves] a punishment of hard labour up to corporal punishment, fine them two armours; [should the persons be guilty of a crime that warrants] shaving off the beard on down, fine them one armour. 其任有辠刑辠以上，任者貲二甲而癈；耐辠、贖辠，任者貲一甲；貲辠，任 者弗坐。 [Should an official] whose guaranteed person commits a crime [that warrants] corporal punishments or higher, fine the guarantor two armours and expel [him] from office; [should the guaranteed person commit a crime deserving of] shaving off the beard or redemption punishments, fine the guarantor one armour; [should the guaranteed person commit a crime deserving of] fines as punishments, the guarantor should not be liable to it. (Statute on the Establishment of Officials [Zhili lü 置吏律]) The Statute on Pursuing Criminals clearly reveals that hard labour and corporal punishment are two subsequent degrees below capital punishment. This not only shows that hard labour, which, in our definition, include punishments of both “wall builders and grain pounders” and “gatherers of firewood and sifters of white rice” (guixin baican 鬼薪白粲), was a separate degree of punishment but also demonstrates that capital – hard labour – corporal – shaving off beard punishments formulated a sequence of punishment degrees in the Qin penal system. Moreover, according to the aforementioned Statute on the Establishment of Officials, persons who had previously guaranteed someone to be an official should be fined one armour when the guaranteed person committed a crime worthy of either the punishment of shaving off the beard or redemption fee. As such, we may further infer that shaving off the beard and redemption were in a similar position in the Qin penal system, followed by fining. However, between shaving off the beard and redemption, there should be another punishment: 〼灋，耐辠以下 （遷）之，其臣史殹，輸縣鹽。能捕若詗告犯令者，刑城 旦辠以下到 （遷）辠一人，購金二兩 . . . . rule. [Should the person be guilty of a crime that matches] punishments of shaving off beard on down, exile him/her; [should] he/she be a literate servant, send him/her to [service at] the Office of Salt in the county administration. For those who can arrest or investigate a person who has broken the ordinance, and whose crime [matches] punishments from mutilating hard labours or down to exile: reward [the one who arrests the person] with two units of gold per person [arrested]. Although exile as a kind of punishment first appears in the Shuihudi text, its place within the overall penal system is still unclear. The above statute, nevertheless, clearly indicates that it was under the shaving off the beard punishment in term of punishment degree. By current sources, we can summarise at least seven punishment degrees in the penal system of the unified Qin Empire, and they were, successively, capital punishment, hard labour, corporal punishment, shaving off the beard, exile, redemption and fine punishments.23 In fact, the penal system of early Han seemed to have little difference from that of Qin, which can be shown in the following article from the “Statutes on the Composition of Judgments (Ju lü 具律): 贖死，金二斤八兩。贖城旦舂、鬼薪白粲，金一斤八兩。贖斬、府（腐）， 金一斤四兩。贖劓、黥，金一斤。贖耐，金十二兩。贖䙴（遷），金八兩。 394

An overview of the Qin-Han legal system

Redeeming the death [penalty]: two jin and eight liang of gold. Redeeming wallbuilding or grain-pounding, or gathering fuel for the spirits or white-rice sorting: one jin and eight liang of gold. Redeeming severing [of the feet] or castration: one jin and four liang of gold. Redeeming severing of the nose or tattooing [of the face]: one jin of gold. Redeeming shaving: twelve liang of gold. Redeeming banishment: eight liang of gold.24 Although the article is about the amount of gold required for redemption, its content reflects that the order of types of punishments, from exile to capital, are exile – shaving off the beard – amputation of nose or tattooing – amputation of feet or castration – hard labour – capital punishment.This sequence corresponds perfectly to the Qin punishment degrees mentioned earlier. Ultimately, the degrees of punishment could be compounded to create a variety of different punishment levels so as to construct the whole Qin penal system (Table 18.1).

Criminal procedure during the Qin and Han era Study of criminal procedure during the Qin and Han era has accumulated much scholarship. For example, the monograph by Momiyama Akira 籾山明 has nearly discussed all aspects of the criminal procedure of the time,25 and countless articles have been published concerning separate procedures in a criminal case.26 In this section, we are trying to give a brief summary of the criminal procedure as a whole and then examine how new sources will affect the study of several particular procedures. Generally speaking, although initiation of litigation can be categorized into the officially initiated “accusation” (he 劾) and privately initiated “denunciation” (gao 告), no significant evident can strongly support that the divergence between criminal and civil proceedings had already existed during the Qin and Han eras, even though lawsuits whose nature coincides with present-day civil cases did exist.27 Regarding the administration of criminal cases, the majority of cases were handled by government officials, and private jurisdiction was strictly prohibited, except those family affairs that were regarded as “denunciation beyond official jurisdiction” (fei gong shi gao 非公室告).28 According to transmitted literature, the highest chief of criminal affairs of the whole empire was the Commandant of the Court (Ting Wei 廷尉), who had two subordinates called Administrator (Zheng 正) and Supervisor (Jian 監), as well as a special group of professionally trained officers named Scribes of the Commandant of the Court (Ting Wei Shi 廷尉史), to assist him in handling difficult or doubtful cases submitted all over the empire. If the Commandant of the Court faces a difficult case, he would organize a meeting to discuss the best judgement of these cases, and it is highly possible that a similar practice had also been adopted in commandery or county-level jurisdictions. Beyond the capital, other jurisdictions that could also try a lawsuit were commanderies and counties, the latter especially, maybe the fundamental units in tackling lawsuits. Similar to later eras, local magistrates always served as chief justice of the region, but officers who actually executed accusation and interrogation may be judicial scribes (Yu Shi獄史). In fact, according to a wooden tablet excavated from Liye dated back after the Qin unification, the First Emperor had issued a provision concerning the number of officers in an interrogation: 卅年十一月庚申朔丙子。發弩守涓敢言之。廷下御史書曰：縣□治獄及覆獄 者，或一人獨訊囚，嗇夫、長、丞、正、監非能與□□殹，不參不便。書到， 尉言。‧今已到。敢言之。 8–141+8–668 十一月丙子旦食，守府定以來。ノ 連半29 萃手 8–141背+8–668背 395

Table 18.1  Overview of punishments under Qin law Degrees of punishment

Levels of punishment

1 2

Death penalty Amputation of feet and to be wall builders and grain pounders 斬趾為城旦舂 Amputation of nose or tattooing and to be wall builders and grain pounders 黥（劓）為城旦舂i Wall builders and grain pounders without mutilationii Amputation of feet and to be gatherers of firewood for the spirits (鬼薪) Amputation of nose or tattooing and to be gatherers of firewoodiii Shaving off of beard and to be gatherers of firewood and sifters of white rice 耐為鬼薪白粲iv Gatherers of firewood and sifters of white rice without mutilation Amputation of feet or castration Amputation of nose or tattooing Shaving off beard and to be bond servants or women 耐為隸臣妾 Shaving off beard and to be robber guards (Sikou 司寇) Shaving off beard and to be watchmen (Hou 候) Exile Redeeming death 贖死 Redeeming hard labours 贖城旦舂、鬼薪白粲 Redeeming amputation of feet or castration 贖斬、腐 Redeeming amputation of nose or tattooing 贖劓、黥 Redeeming shaving off beard Redeeming exile Fining two armours Fining one armour Fining one shield

Capital Hard labour

3 4 5 6 7 8 9 Corporal 10 11 Shaving off beard 12 13 14 Exile 15 Redemptions 16 17 18 19 20 21 Fines 22 23

i As seen in Qin statutes, amputation of feet or nose or tattooing gatherers of firewood can be generalised as “mutilated wall builders” (xing chengdan 刑城旦). See Shuihudi Qin mu zhu jian zheng li xiao 睡虎地 秦墓竹簡整理小組, Shuihudi Qin mu zhu jian 睡虎地秦墓竹簡 (Beijing: Wen wu chu ban she, 1978): 119; Hulsewé, Remnants of Ch’in law: 149. ii  It is noteworthy that the punishment “hard labour without mutilation” was a kind of specialty in the Qin and Han penal system. One article of Statute on Abscondence (Wang lü 亡律) from Yuelu Academy stored manuscripts says: [Should a person be guilty of crime that] matches the punishment of hard labour without mutilation on down, to shaving off beard, and those absconded government slave, robber guard, bond servants or women and absconded private servants lodge in a private house, private tavern or official tavern, and the person in charge of the tavern does not realise he/she is absconding, fine [the person in charge] two armours. 當完為城旦舂以下到耐罪及亡收、司寇、隸臣妾、奴婢闌亡者舍人室、人舍、官舍，主舍 者不智（知）其亡，貲二甲。 The aforementioned article denotes that despite “hard labour without mutilation” is within the collection of punishment degree second to capital, sometimes it has the same treatment as corporal and shaving off beard, both of which were lighter punishment degrees compared to “hard labour without mutilation. In fact, “hard labour without mutilation” is so different from the other punishments in the hard labour degree that scholars had suggested that it should be classified as a separate punishment. However, other articles in the Qin and Han statute also reveal that “hard labour without mutilation” still belongs to hard labour punishment, and the reason behind its connection with other lighter punishments might simply because it does not involve any mutilation. iii Likewise, amputation of feet or nose or tattooing gatherers of firewood can be generalised as “mutilated gatherers of firewood” (xing guixin 刑鬼薪). See Shuihudi Qin mu zhu jian: 119–120; Hulsewé, Remnants of Ch’in law: 149. iv    The punishment “shaving off beard and to be gatherers of firewood and sifters of white rice” first appeared in the slip 1114 of the Yuelu Academy materials.

An overview of the Qin-Han legal system

In the 30th year, in the 11th month, which began on a gengshen day, on the day bingzi, the chief of the Office of Shooting Crossbows presumes to report: the county office had handed down an Imperial Secretary’s letter, saying that those [officials] who try lawsuits and verify lawsuits for the county may single-handedly interrogate prisoners, and overseer, prefect, assistant, judge and supervisor cannot participate [during the process] . . . . If there is less than three [officials participating in the interrogation], it would not be convenient. When the letter is arrived, the county commandant should reply.‧Now [the letter] was arrived. Here I presume to report. Cui handled [this]. In the 11th month, on the day bingzi, at the double-hour of breakfast, Office Guard Ding went to forward [this]./Lian opened [this]. The tablet contains a letter from the chief of the Office of Shooting Crossbows to the county government, quoting a previous letter sent from the county concerning the provision issued by the imperial secretary, which is about the stipulation of trying lawsuits.The provision stated that when the county government tries or verifies lawsuits, there should be at least three officers, including the overseer, prefect or assistant, appearing. Even though we cannot assure that this provision was strictly enforced or not, it can still serve as concrete evident that the Qin Empire had attempted to govern the number of officers involved in trying lawsuits. Obviously, this is to avoid unfairness such as illicit torture that might occur in a one-man trial. After introducing some general features about the Qin and Han lawsuit, we can finally begin our discussion on the criminal procedure of the time. Basically, litigation had four main procedures. First, a person, either an officer or a commoner, would declare a formal “accusation” or “denunciation,” depending on their respective identity. Given that unearthed accusations found in Jiaqu company fort 甲渠候官 show that an accusation is an official document with certain required formats that must be strictly followed and should include personal particulars of the accused such as his/her name, rank, place of origin, etc,30 there is insufficient information to determine whether denunciations also had to bear ascertain format or were simply made orally. However, materials from Jiaqu company fort do show that denunciations may be attached with a written transcript (yuan shu 爰書) recording details of the case as a valid proof. The second procedure is interrogation (xun 訊), which most likely took place in the county’s prison and always constitutes multiple questionings. In light of the Book of Submitted Doubtful Cases, we can see that the interrogation usually needs at least three questionings before formulating the final judgement. The first one is a general inquiry about details of the case called “interrogation” (xun 訊), at which officers should minimise their interference with the accused even if they know that he/she did not speak the truth. It was not until the accused had finished his/her testimony and all details were being transmitted into a written statement that the officers could begin the second stage of the interrogation, “indictment” (jie 詰), whose main purpose is to elicit answers of unsolved problems or contradictions that arose in the testimony of the accused person, who was required to give a proper explanation to these doubts. Sometimes, the interrogators were entitled the right to torture the accused, even though overdose of torturing is prohibited by the official models. Usually, an indictment would last until the accused confessed and stated “no other explanation” (wu ta jie 毋他解). The final stage of this procedure is “inquiry” (wen 問), which consists of some supplementary questions to the accused’s home district aiming at verifying data such as personal particulars or previous conviction, if applicable, of the accused. After the verification, if any information related to the case is missing, the officers would include it as supplementary details. On the contrary, if all the details are valid and accurate, the officers would simply write “An inquiry was conducted: It was as in the statements” (wen ru ci 問如辭) to conclude the report.31 397

Kyung-ho Kim and Ming-chiu Lai

Following interrogation is the procedure of “trial” (ju 鞫). The Shuowen jiezi 說文解字 reads the word “ju” as “to thoroughly try a criminal” (qiongli zuiren 窮理罪人). As pointed out by Barbieri-Low and Yates, “ju was a process of comparing written documents,” and “it was actually a trial of the evidence, not a trial of a defendant in a courtroom.”32 After evaluating all statements made during the interrogation procedure and other evidence, the officials could finally hand down a verdict to conclude the case as “verified” (shen 審).33 The final, and by far the most important, process in the criminal procedure is “sentencing.” (lun 論). In this process, officials would weigh the crime committed by the accused and determine a suitable punishment to “match” (dang 當) his/her crime by citing an appropriate statute. After they finished trying the lawsuit, magistrates of counties had to deliver the “complete case files” (ju yu 具獄) to the respective commandery governor, who would review whether it was fairly tried. At the same time, the person convicted has the right to lodge an appeal named “to plead for a trial” (qiju 乞鞫) within a year, and if he/she did so, the associated county had to prepare a written statement of his/her appeal and dispatch it to the commandery-level governmental offices. After that, the commandery governor will appoint some of his Accessory Scribes as Metropolitan Officials (Du li都吏) to the county to directly review the case, and such a process was called a “reviewing lawsuit” (fu yu 覆獄). The seventeenth case of the Book of Submitted Doubtful Cases, for example, records that a convicted named Jiang successfully appealed his case, and since he had already been mutilated, the government had to amnesty him to work in the “hidden office” (Ying Guan 隱官). In addition, his wife, who was sold as a government slave because of the liability to Jiang, was redeemed by the government, and other confiscated properties were repaid in market price, too. However, if the appeal ended up being a failure, the convicted would have to increase one level of his existing punishment. Herein, we have given an overview to the criminal procedure of the Qin and Han eras. Fortunately, the newly unearthed sources have displayed unprecedented details of the aforementioned procedures, some of which will be examined in the remaining section.

The role of the imperial secretary in supervision (監御史) in litigations After the Qin unification, it is recorded in Shiji that the office of the Imperial Secretary in Supervision had been established alongside governor and commandant in the commanderylevel administration. Unfortunately, due to the limitation of sources, nothing other than its title was known. However, content of case 1, titled “the Case concerning Gui and Suo conspired to receive the bounty” (癸、瑣相移謀購案) from the recently published Yuelu Academy Qin slip contains new materials regarding the duties of the Imperial Secretary in Supervision and its role in criminal procedure of the time: 五月甲辰，州陵守綰、丞越、史獲論：令癸、瑣等各贖黥。癸、行戍衡山郡 各三歲，以當灋（法）；先備贖，不論沛等。監御史康劾以為不當：錢不 處，當更論，更論及論失者，言夬（決）。 In the 5th month, on the Jiachen day, Zhouling prefect Wan, assistant Yue, scribe Huo sentenced: commanding Gui, Suo and others to [the punishment of] redeeming tattooing; and Gui and Xing should service in Hengshan commandery respectively for three years, in order to match the law. [Gui and Suo and others] should first pay off their redemption. Imperial Secretary in Supervision Kang accused [the officials] and thought [the judgement] did not match [the law]: Given that the money has not been settled, [Zhouling’s officials] should change the sentence [of Gui, Suo and others]. For 398

An overview of the Qin-Han legal system

those who have changed their sentence and had sentenced the case incorrectly, report the judgement [to the commandery]. The case states that magistrates and scribe Huo of Zhouling county had sentenced the convicted Gui and Suo’s group to punishment of redeeming tattooing. When the case was sent to the commandery for possible review, however, the Imperial Secretary in Supervision of Southern Commandery (Nanjun 南郡) considered Zhouling’s judgement inappropriate and ordered a revisiting of their crime and eventual punishment. Furthermore, he required Zhouling to punish those officials who accounted for the misjudgment and submitted the results to the commandery afterward, implying that under the criminal procedure of the unified Qin Empire, once a county-level magistrate has tried lawsuits, he should send the relevant documents to the Imperial Secretary in Supervision of his commandery and let him to determine if the judgement is proper or not, and based on the aforementioned case, the Imperial Secretary was also authorised to defy judgements from counties.

Adjudication against magistrates Unlike modern society, the Qin and Han government did not maintain a separation of powers, and local magistrates served as both chief executive and chief justice within their jurisdiction. But who was responsible to judge the misconduct of magistrates? An immediate answer would be their higher authorities, that is, the central government (for commanderies) and the commandery (for counties). While such a view was attested from records of the transmitted sources, new sources do denote that under certain circumstances, the magistrate could be judged by no one but himself. For instance, an archive from Liye Qin slips no. 8–754+8–1007 records that: 卅年□月丙申。遷陵丞昌、獄史堪訊昌，辤（辭）曰：上造，居平□，侍廷為 遷陵丞。□當詣貳春鄉，鄉 渠、史 獲 誤 詣 它 鄉，□失道百六十七里。即與 史論貲渠、獲各三甲，不智劾云貲三甲不應律令。故皆毋它坐。它如官書。 8–754+8–1007 〼堪手 8–754背+8–1007背 In the 30th year, in the Xth month, on the bingshen day, the Qianling’s county assistant Chang, judicial scribe Kan interrogated Chang, whose statement says: “[Ranked] Shang zao, living in Ping. . ., serving in the county office as the Qianling county assistant. X should have been brought to Erchun commune, [but] its overseer Qu, and [Qu’s] scribe Huo had mistakenly carried [X] to other commune, and thus missing the road to 167 li. [Chang], with scribe [Kan], immediately sentenced Qu and Huo respectively to a fine of three units of armor and did not realize that suggesting a fine of three units of armours in the official accusation is not corresponding to the statute and ordinance. Both [Chang and Kan] previously did not commit other crimes. The rest is the same as [the content] of the official document. 8–754+8–1007 Kan handled this. 8–754B+8–1007B Qianling county assistant Chang had made a statement about errors, 8–754+8–1007, during the prosecution of the overseer of Erchun commune Qu and his scribe Huo. The testimony of assistant Chang states that he and his subordinate scribe had sentenced Qu and Huo to redemption of three armours and did not realise that it was not a valid punishment in the Qin penal system. Interestingly, the statement records that the officials interrogating Chang 399

Kyung-ho Kim and Ming-chiu Lai

were Chang himself and a judicial scribe named Kan 堪! In fact, in the aforementioned “Case concerning Gui and Suo conspired to receive the bounty,” the officials who were responsible to sentence the misjudged magistrates Wan and assistant Yue may also be themselves. Such phenomena may indicate that in the Qin criminal procedure, county magistrates could simply sentence and match their own crime and punishment and did not require the intervention from the third party insofar as the crime which they were committing was minor errors in daily administration.

The original of trial (ju 鞫) document excavated from Tuzi shan兔子山 Trial is the most crucial part in the Qin and Han criminal procedure because it determines the crime and punishment that a person convicted would suffer. Although important, our knowledge of trial documents is wholly built on quotations from the Book of Submitted Doubtful Cases and several received sources, which are fragmentary and always under heavy editions. Fortunately, the recently discovered Tuzi shan Han slips unearthed an original trial document dated back to the second year of Emperor Ping (平帝), in which the final countdown of Former Han was imminent: 鞫。勳，不更，坐「為守令史署金曹，八月丙午為縣輸元年池加錢萬三千臨 湘。勳匿不輸，即盜以自給。勳主守縣官錢臧二百五十以上」。守令史恭 劾，無長吏教使者。審。 J3⑤:1A 元始二年十二月辛酉。益陽守長豐，守丞臨湘右尉□，兼掾勃，守獄史勝言： 數罪以重，爵減，髡鉗勳為城旦，衣服如法，駕（加）責如所主守盜。沒入 臧縣官，令及同居會計備償少內，收入司空作。 鞫 J3⑤:1B Trial. Xun, ranked bugeng, is liable to [the crime of] “being an acting foreman clerk who was assigned to the Bureau of Currencies, in the 8th month, on the day bingwu, transferring first year’s chijia money amounting to 13000 dollars for the [Yiyang] county to the Linxiang county, but had hidden [the money] and not transferred it, instead stealing it so as to feed himself. Xun held an appointment of escorting public money [but eventually stole the money], and the illicit profit is worth over 250 dollars.” Acting foreman clerk Gong accused [Xun], and in no respect, was I instigated by a senior official [to make this accusation]. [The fact has been] verified. In the second year of yuanshi, in the 12th month, on the day xinyou, Yiyang’s acting prefect Feng, acting assistant, Linxiang [county’s] right commander X, concurrent Head Bo, acting judicial scribe Sheng said: [Under the principle of] multiple crimes only count the heaviest and a reduction [of punishment] through orders of honour, – [we decide] to shave and collar Xun to hard labour, with clothes in accordance with law, and increasing his punishment to holding an appointment and stealing [goods]. [In addition to] confiscating the illicit profit to the government, [Xun] and those who dwell together with him have to summarize their accounts and repay [the money] to shaonei, and [those responsible] should be arrested and work in sikong. J3⑤:1A Trial. J3⑤:1B 400

An overview of the Qin-Han legal system

The trial was about a corrupted government official called Xun who had stolen the fiscal budget of the previous year (the first year). Unfortunately, we still do not know much about the exact meaning of the term “chijia money,” even though it is highly possible that it is related to tax revenue generated from state-owned ponds in royal gardens. The document was written on a huge tablet of current length of 49.3 cm and 6.55 cm width, by far one of the largest tablets ever seem. In the verso of the document, a single Chinese character – “Trial” – was written so as to denote the nature of the document. In its recto, writings were divided into two paragraphs. The first part is closed with the term “verified”, and thus should be related to the trial of the case. It is worth noting that the wordings of the trial part are strongly reminiscent of the “accusation” documents discovered in northwestern China. For example, both accusation and trial start with personal particulars of the accused, and more importantly, both of which contain the clause “A accused [B], and in no respect, was I instigated by a senior official [to make this accusation]”. These similarities may imply that trial was in fact converted from accusation with few amendments. However, it is the second half that is the most interesting. The second paragraph consists of the sentence and matching of punishment of the convicted and those liable to his crime. Although there is a 200-year time difference, the principle of sentencing may have been traced back to the following Statute on the Composition of Judgments in Ernian lüling, saying that: “For one who receives bribes [to improperly decide these legal matters]: add two degrees to his crime. When the crime of the illicit profit that he has been given [as a bribe] is heavier [than the crime that he has improperly adjudicated], sentence him for the one that is heavier and also add two degrees [to his crime]. 其受賕者，駕（加）其罪二等。所予臧（贓）罪重，以重 者論之，亦駕（加）二等。”34 Here the corrupted acting foreman clerk Xun may be accused by this statute, so he was sentenced to the most serious crime which he had committed, and because Xun was stealing the fiscal budget of which he should have taken care, he was eligible for the condition of increasing punishment and thus was shaved and collared to do hard labour after a reduction through orders of honour. This paragraph shows that the trial document could contain materials from later criminal procedures like sentencing and matching, denoting that the latter two may not be separate processes but were part of the trial process. Even though it is still too early to draw any conclusion on one single document and we cannot eliminate the possibility that the second paragraph might be added after the trial process, it is still a noteworthy phenomenon that deserves future attention.

Conclusion In this chapter, we have given a brief introduction to unearthed legal manuscripts, mostly notably texts from Shuihudi tomb no. 11, Zhangjiashan Tomb no. 247 and Qin slips in procession of the Yuelu Academy, summarising their content and importance in the field of early Chinese legal study. After this, we also provide a preliminary study on the reconstruction of the penal system and a brief summary of the criminal procedure of Qin and Han eras, mainly using the latest disclosed materials of the Yuelu Academy and Liye. In sum, in light of the Qin statutes from the collection of Yuelu Academy, we believe that the Qin penal system at least had seven punishment degrees, to which twenty-three punishment levels could further be ramificated. In addition, based on existing sources, there were at least four procedures in the Qin and early Han criminal proceeding. They are: (1) accusation/ denunciation; (2) Interrogation; (3) Trial; (4) sentencing. Needless to say, it is impossible to fully elaborate the value of the new sources in this brief introduction, and their interpretations are still subject to further discoveries and discussions. 401

Kyung-ho Kim and Ming-chiu Lai

What we do hope, however, is that this tiny chapter could somewhat serve as an introduction to these new materials and provide some help to researchers who are interested in the study of early Chinese law.

Notes * This manuscript is supported partially by the Hong Kong RGC General Research Fund (CUHK444813) and the Ministry of Education of the Republic of Korea and the National Research Foundation of Korea (NRF-2018S1A6A3A01023515). The authors would like to express their gratefulness to Mr. Tong Chun Fung, the Research Assistant, who contributed a lot of time and made valuable comments on the preparation and translation of the draft. 1 A.F.P. Hulsewé, Remnants of Han Law, Vol. I: Introductory Studies and Annotated Translation of Chapter 22 and 23 of the History of the Former Han Dynasty (Leiden: Brill, 1955): 4. 2 For a detailed account of scholarships after the publication of the Shuihudi Qin manuscripts, see Xu Shihong 徐世虹 and Zhi Qiang 支強, “Qinhan falu yanjiu bainian (san) – 1970 niandai zhongqi zhijin: Yanjiu de fanrong qi” 秦漢法律研究百年（三） – 1970年代中期至今：研究的繁榮期, in Zhongguo gudai falü wenxian yanjiu 中國古代法律文獻研究 Vol. 6 (Beijing: Zhongguo zheng fa da xue chu ban she, 2012): 95–170. 3 For latest editions of the legal manuscripts discussed in this article, see Wuhan daxue jian bo yanjiu zhongxin 武漢大學簡帛研究中心, Hubei sheng bowuguan 湖北省博物館, Hubei sheng wenwu kaogu yanjiu suo 湖北省文物考古研究所 edited. Qin jian du heji 秦簡牘合集 Vol.1 and 2 (Wuhan: Wuhan da xue chu ban she, 2014); Zhu Hanmin 朱漢民 and Chen Songchang 陳松長 edited, Yuelu shu yuan cang Qin jian 嶽麓書院藏秦簡 Vol. 3 (Shanghai: Shanghai ci shu chu ban she 2013); Peng Hao 彭浩, Chen Wei陳偉, Kudō Motoo 工藤元男 edited, Ernian lüling yu Zouyan shu: Zhangjiashan er si qi hao Han mu chu tu fa lü wen xian shi du 二年律令與奏讞書：張家山二四七號漢墓出土法律文獻釋 讀 (Shanghai: Shanghai gu ji chu ban she, 2007). 4 Editors of the Shuihudi manuscripts have translated all legal texts into modern Chinese, while the Chinese translation of the Statutes and Ordinances of the Second Year (Ernian lüling 二年律令) is still in progress. For some of the results, see Huang Yijun 黃怡君 You Yifei 游逸飛 Li Chengjia 李丞家 Lin Yingjun 林盈君 and Li Xiezhan 李協展, “Zhangjiashan Han jian “Ernian lüling. Zhililǜ” yizhu” 張家 山漢簡《二年律令．置吏律》譯注, in Shi yuan 史原 Vol. 1 (2010.09): 287–337; Li Guanting 李冠 廷 and You Yifei, “Zhangjiashan Han jian “Ernian lüling. Junshulǜ” yizhu” 張家山漢簡《二年律令． 均輸律》譯注, in Shi yuan史原 Vol. 2 (2011.09): 239–256; Gao Zhenhuan 高震寰, Cai Peiling 蔡佩 玲, Zhāng Li 張莅, Lin Yide 林益德, You Yifei, Zhangjiashan Han jian “Ernian lüling. Qianlǜ” yizhu” 張家山漢簡《二年律令．錢律》譯注, in Shi yuan 史原 Vol. 3 (2012.09): 295–352; Huang Qiongyi 黃瓊儀, Liu Xiaoyun 劉曉芸 and You Yifei, “Zhangjiashan Han jian “Ernian lüling. Chuanshilǜ” yizhu” 張家山漢簡《二年律令．傳食律》譯注, in Shi yuan 史原 Vol. 4 (2013.09): 263–300. 5 For the English translation of the Shuihudi legal texts, see A.F.P. Hulsewé, Remnants of Ch'in Law: An Annotated Translation of the Ch'in Legal and Administrative Rules of the 3rd Century B.C., Discovered in Yun-Meng Prefecture, Hu-pei Province, in 1975 (Leiden: E.J. Brill, 1985); for the Zhangjiashan texts, see Anthony J. Barbieri-Low and Robin D.S. Yates, Law, State, and Society in Early Imperial China: A Study with Critical Edition and Translation of the Legal Texts from Zhangjiashan Tomb no.247 Vol. 2 (Leiden: E.J. Brill, 2015); for the Wei yu deng zhuang si zhong 為獄等狀四種 processed by the Yuelu Academy, see Ulrich Lau and Thies Staack, Legal Practice in the Formative Stages of the Chinese Empire: An Annotated Translation of the Exemplary Qin Criminal Cases from the Yuelu Academy Collection (Leiden: E.J. Brill, 2016). 6 For the Japanese translation of “Falu Dawen” 法律答問, see Matsuzaki Tsuneko 松崎つね子, Suikochi Shinkan 睡虎地秦簡 (Tokyo: Meitoku Shuppansha, 2000): 69–224, and for the complete Japanese translation of the Ernian lüling manuscripts, see Tomiya Itaru 冨谷至 edited, Kōryō Chōkasan nihyakuyonjūnana-gō bo shutsudo Kan ritsuryō no kenkyū 江陵張家山二四七號墓出土漢律令の研究 Vol. 2 (Kyoto: Hōyū Shoten, 2006), and for Zouyan shu, see Ikeda Yuichi 池田雄一, Sōgensho: Chūgoku kodai no saiban kiroku 奏2䅊書：中国古代の裁判記錄 (Tōkyō : Tōsui Shobō, 2002). 7 See Ulrich Lau und Michael Lüdke, Exemplarische Rechtsfälle vom Beginn der Han-Dynastie: Eine kommentierte Übersetzung des Zouyanshu aus Zhangjiashan (Tokyo Univ. of Foreign Studies, 2012). 8 The original editors transcribed the character “鄉” as “御”. Here we adopt the new transcription proposed by Chen Kanli陳侃理; see his “Shuihudi Qin jian “Biannianji” zhong “Xi” de huan li” 睡虎地 秦簡《編年記》中“喜”的宦歷, Guoxue xue kan 國學學刊, 2015 No. 4: 48.

402

An overview of the Qin-Han legal system 9 For the English translation of the Chronicle and Document of Instructions, see Zhang Zhenglang 張政烺, Ri Zhi 日知 edited, Inscriptions of Yumeng Bamboo Slips 雲夢竹簡 Vol. 1 (Zhangchun: Dongbei shifan daxue chuban she, 1990): 17–31; 39–42. 10 See Qinjiandu heji Vol. 2: 3. 11 See appendix one and two of the Zhangjiashan Han mu zhu jian (er si qi hao mu): 309–314, 322. 12 For the original texts, see Ban Gu 班固, Hanshu 漢書 (Beijing: Zhonghua shu ju, 1962): 1106. For English translation, see Hulsewé, Remnants of Han Law, Vol. I: 343; Barbieri-Low and Yates, Law, State, and Society in Early Imperial China: 172. 13 Ikeda Yūichi 池田雄一, Chūgoku kodai no ritsuryō to shakai 中国古代の律令と社会 (Tōkyō: Kyūko Shoin, 2008): 587. 14 Hulsewé, Remnants of Han Law,Vol. I: 142–143. 15 See Cai Wanjin 蔡萬進, Zhangjiashan Han jian "Zouyan shu" yan jiu 張家山漢簡《奏讞書》研究 (Guilin Shi: Guangxi shi fan da xue chu ban she, 2006); Ikeda Yuichi, Kandai o sakanoboru sōgen —— Chūgoku kodai no saiban kiroku 漢代を遡る奏讞——中國古代の裁判記錄 (Tōkyō: Kyūko Shoin, 2015). 16 See Peng Hao 彭浩, “Tan “Zouyan shu” zhong de Xihan anli” 談《奏讞書》中的西漢案例, Wenwu 文物, 1993.8: 32–36; “Tan “Zouyan shu” zhong Qin dai he Dongzhou shiqi de anli” 談《奏讞書》中 秦代和東周時期的案例, Wenwu, 1995.3: 43–47. 17 Shortly before the volume was formally published, a series of articles concerning the manuscripts have been launched in Vol. 27 No. 3 of the Hunan daxue xuebao (shehui kexue ban) 湖南大學學報（社會 科學版）. See Yu Zhenbo 于振波, ”Qindai lizhi guankui—yi Qin jian sifa, xingzheng wenshu wei zhongxin” 秦代吏治管窺—以秦簡司法、行政文書爲中心; Xiao Hongyong 肖洪泳, “Yuelu Qin jian suojian Qin xingshi susong chengxu de lishi jiazhi” 嶽麓秦簡所見秦刑事訴訟程序的歷史價値; Thies Staack 史達; Li Jing Rong 李婧嶸 translated, “Yuelu Qin jian “wei yu deng zhuang si zhong” juan ce yi de bian lian—yiju jian bei hua xian he beifan yin ziji fuyuan juanzhou yuanmao 嶽麓秦 簡《為獄等狀四種》》卷冊一的編聯—依據簡背劃線和背反印字迹復原卷軸原貌, in Hunan daxue xuebao(shehui kexue ban),Vol. 27 No. 3 (2013): 11–13; 14–19; 20–25. 18 See Su Junlin 蘇俊林, “Yuelu Qin jian “Wei yu deng zhuang si zhong” mingming wenti tantao” 嶽麓 秦簡《為獄等狀四種》命名問題探討, Jian du xue yanjiu 簡牘學研究Vol. 5 (Lanzhou: Gansu renmin chuban she, 2014): 9–14. 19 For a preliminary study comparing the Book of Submitted Doubtful Cases and Wei yu deng zhuang si zhong, see Ulrich Lau 勞武利; Li Jing Rong translated, “Zhangjiashan Han jian “Zouyan shu” yu Yulue shuyuan Qin jian “wei yu deng zhuang si zhong” de chubu bijiao”《張家山漢簡《奏讞書》與嶽 麓書院秦簡《為獄等狀四種》的初步比較, Hunan daxue xuebao”(shehui kexue ban), Vol. 27 No. 3 (2013): 5–9. 20 For studies of the Qin and Han penal system as a whole, see Tomiya Itaru 冨谷至; Chai Shengfang 柴生芳 and Zhu Hengye 朱恒曄 translated, Qinhan xingfa zhidu yanjiu 秦漢刑罰制度研究 (Guilin: Guangxi shi fan da xue chu ban she, 2006): 8–50; Mizuma Daisuke 水間大輔, Shin Kan keihō kenkyū 秦漢刑法研究 (Tōkyō: Chisen Shokan, 2007): 15–95; Miyake Kiyoshi 宮宅潔, Chūgoku kodai keiseishi no kenkyū 中国古代刑制史の研究 (Kyōto: Kyōto Daigaku Gakujutsu Shuppankai, 2011): 39–187. 21 See Tomiya Itaru, Qinhan xingfa zhidu yanjiu: 49–50; Li Junming 李均明, “Zhangjiashan Han jian suo jian xingfa deng xu ji xiangguan wenti” 張家山漢簡所見刑罰等序及相關問題, in Jian du fa zhi lun gao 簡牘法制論稿 (Guilin: Guangxi shi fan da xue chu ban she, 2011): 34–42. 22 See Arnd Helmut Hafner 陶安あんど, Shin Kan keibatsu taikei no kenkyū 秦漢刑罰体系の研究 (Tōkyō: Sōbunsha, 2009): 12–203. Readers may also make reference to the detailed review article by Liu Hsin-ning 劉欣寧 and Hafner’s response. See Liu Hsin-ning, “Arnd Helmut Hafner Shin Kan keibatsu taikei no kenkyū shuping” 陶安あんど《秦漢刑罰体系の研究》述評, in Fazhishi yanjiu 法制史研究 Vol. 17 (2009.12): 359–373; Arnd Helmut Hafner, “Qinhan falu shiliao shangque ji ze” 秦漢法律史料商榷幾則, in Fazhishi yanjiu Vol. 19 (2011.12): 231–244. 23 Herein we decide to exclude “imprisoned with hard labour” (xi chengdanchong 繫城旦舂) from the penal system because it is still uncertain whether it was indeed a punishment degree or just used as a kind of adjunctive to existing punishments. 24 See Barbieri-Low and Yates, Law, State, and Society in Early Imperial China: A Study with Critical Edition and Translation of the Legal Texts from Zhangjiashan Tomb no.247 Vol. 2: 510–511. 25 Momiyama Akira 籾山明; Li Li 李力 translated, Zhongguo gudai susong zhidu yanjiu 中國古代訴訟制 度研究 (Shanghai: Shanghai gu ji chu ban she, 2009): 48–109. Also, in their last monograph, BarbieriLow and Yates have given a detailed account to the criminal proceeding in the Qin and Han period;

403

Kyung-ho Kim and Ming-chiu Lai see their Law, State, and Society in Early Imperial China: A Study with Critical Edition and Translation of the Legal Texts from Zhangjiashan Tomb no.247 Vol. 1: 111–186. 26 See Xu Shihong and Zhi Qiang, “Qinhan falu yanjiu bainian (san) – 1970 niandai zhongqi zhijin: Yanjiu de fanrong qi”: 152–161. 27 See Zhang Jianguo 張建國, “Juyan xin jian “Su jun zhai Kou En” minshi susong ge'an yanjiu” 居延新 簡“粟君債寇恩”民事訴訟個案研究, in Di zhi shi dai de Zhongguo fa 帝制時代的中國法 (Beijing: Fa lu chu ban she, 1999): 315–345; Xu Shihong, “Han dai minshi susong chengxu kao shu” 漢代民事訴 訟程序考述, in Zhengfa luntan, 2001.1: 122–130. In addition, Momiyama Akira also pointed out that civil cases could sometimes turn into a criminal proceeding, insofar as officials had discovered criminal elements in the case. See Zhongguo gudai susong zhidu yanjiu: 139–141. 28 Here we adopted the translation of Paul R. Goldin but changed it from “accusation” to “denunciation” to avoid any confusion between gao and he. For studies concerning fei gong shi gao and stipulation of family affairs from the Qin and Han law, see Paul R. Goldin, “Han Law and the Regulation of Interpersonal Relations: ‘The Confucianization of the Law’ Revisited”, Asia Major, 3rd ser., 25, No. 1 (June 2012): 14–18; Ulrich Lau, “The Scope of Private Jurisdiction in Early Imperial China: The Evidence of Newly Excavated Legal Documents”, Asiatische Studien Vol. 59, No. 1 (2005): 343–345. 29 The character “ban” (to open 半) was transcribed as “shou” (to handle 手) by the original editors. Here we follow the edition by He Youzu 何有祖; see his “Du liye Qin jian zhaji (yi)” 讀里耶秦簡札記（ 一）, in Jianbowang 簡帛網, 17–06–2015, www.bsm.org.cn/show_article.php?id=2261 30 For the format of the official accusation, see Anthony J. Barbieri-Low, “Model Legal and Administrative Forms from the Qin, Han, and Tang and their Role in the Facilitation of Bureaucracy and Literacy”, Oriens Extremus Vol. 50 (2011): 132–135; Tong Chun Fung 唐俊峰,“Jiaqu hou guan di 68 hao tan fang chutu he zhuang jian ce de fuyuan yu yanjiu”甲渠候官第68號探方出土劾狀簡冊的復原與研究, in Jian du xue yanjiu 簡牘學研究 Vol. 5 (Lanzhou: Gansu renmin chuban she, 2014): 38–58. 31 Here we adopt the translation by Barbieri-Low and Yates, see Barbieri-Low and Yates, Law, State, and Society in Early Imperial China: 160. 32 Barbieri-Low and Yates, Law, State, and Society in Early Imperial China: 162. 33 For a more detailed discussion of the process of “Ju” in English, see Barbieri-Low and Yates, Law, State, and Society in Early Imperial China: 162–163. 34 See Barbieri-Low and Yates, Law, State, and Society in Early Imperial China: A Study with Critical Edition and Translation of the Legal Texts from Zhangjiashan Tomb no.247 Vol. 2: 503.

404

19 LITERATURE

STEPHEN DURRANTLITERATURE

Stephen Durrant

Preface “Literature” has no precise equivalent in early Chinese. Perhaps the closest approximate is the word wen, which originally meant “pattern” or “refinement.” This word came to be used to describe the written character and, by extension, any written text exhibiting particular refinement or what we might call “literary merit.” Such texts were highly esteemed: “Words with pattern,” wrote Liu Xie (465–522 ce), China’s first great scholar of literature, “Indeed express the mind of the universe” (Shih trans. 1983: 15). It remains to readers in China as well as in the West to sort out which texts are properly defined as “literature,” which manifests “the mind of the universe.” Such sorting cannot be based upon genre alone. Certainly poetry, the most obviously patterned form of writing, belongs to literature, as does prose narrative, chiefly found for early China in historical writing (shi). In addition, some philosophical writing, what the Chinese categorize as the work of “masters” (zi), displays highly sophisticated rhetoric, engaging anecdotes, artful characterization, and other literary features. We discuss here several early Chinese texts of poetry, narrative, and philosophy as outstanding examples of “refined writing,” but the Chinese literary tradition begins with the rhythms and rhymes of poetry, and poetry in one form or another will remain at the center of the tradition for a good many centuries to come. This is so in part because poetry was from early times considered the proper mode for emotional expression, the genre in which sentiment could be given free rein, albeit sometimes in carefully encoded form. So, with poetry we must begin and, eventually, end this survey.

The emergence of poetry The precise origin and early evolution of Chinese writing remains a subject of controversy. By around 1200 bce, writing is found as records of oracle divination inscribed on oxen bones and turtle shells. About the same time, inscriptions appear on bronze vessels used in rituals to honor ancestors.The first Chinese writings that can be appropriately categorized as literature are certain relatively lengthy bronze inscriptions from the tenth and ninth centuries bce. Literarily significant in their own right, these inscriptions display features that point forward to the earliest preserved Chinese poetry. Three features deserve mention: first, frequent use of a four-character or four-syllable line (each character in Chinese script writing a syllable, which, at that time, most 405

Stephen Durrant

usually corresponds to a word) – a stately and staccato style that we might label “the rhythm of ritual power”; second, sporadic use of rhyme; and third, quotation of liturgical or court-centered speech made up of “stock phrases from the Zhou ritual language,” an early expression of a “culture of sophisticated rhetoric” (von Falkenhausen 2011: 268). The lengthier bronze inscriptions typically contain praise for ancestors, a record of an imperial appointment, a wish for the longevity of the noble who sponsors the casting, and a concluding invocation that “sons and grandsons might forever treasure and use the vessel in offering sacrifices.” Thus, the vessel, along with its inscription, is a bridge across time, a link between ancestors, the living, and descendants yet unborn. As such, the inscribed language is elevated: Happy and helpful was Ancestor Yi! He assisted and served his ruler, Distantly planning [with] belly and heart [his] son’s acceptance. Clear-eyed and bright was Grandfather Xin of the branch lineage. Transferring [the lineage] and nurturing sons and grandsons, [He had] abundant good fortune and many blessings. Even-horned and redly gleaming, Appropriate were his sacrifices. (Trans. by Shaughnessy [1991: 189–190])1 Six of the eight lines of this eulogy to ancestors from a late tenth century bce inscription are four characters in length. It is a short step from this inscription to the following four-character lines of one of China’s earliest poems: Bright and refined are the princes, Who grant good fortune and blessings; They show us kindness unending – May sons and grandsons protect them. (Mao Ode #269) Despite the undeniable continuity between the bronze inscriptions and the earliest examples of poetry, the latter develop in ways that make them China’s first literary masterpiece and a major contribution to world literature in general. The Book of Odes (Shijing), known for most of its early history simply as the Odes, is an anthology of 311 poems, six of them preserved in title only.While it is difficult to determine the precise date of the majority of these pieces, which clearly span several centuries, the anthology was largely constituted by around 500 bce, with individual pieces reworked for several centuries thereafter. With several noteworthy exceptions, these poems are relatively brief. The fourcharacter rhythm predominates, and rhyme is typical at the end of even lines.The subject matter varies, extending from pieces in honor of ancestors and expressing reverence for sacrifices, such as the poem cited earlier, to pieces that concern such common emotions as friendship, love, sorrow, the joys of social celebrations and the horror of war. Put somewhat differently, the odes extend from the most religious to the most secular, and generally this topical shift corresponds to the passage of time – that is, they display a general shift from a world of lofty ritual to a world of more mundane sentiment as time passes or, as one scholar aptly puts it, a move from “liturgy to literature” (Shaughnessy 1997: 165–196). Some of the earliest odes have a narrative quality and honor the founders of the Zhou royal lineage, the lineage ancestor Houji, a prominent early Zhou leader named Liu the Duke and, most importantly, Kings Wen and Wu, who actually founded the dynasty. One scholar has suggested that these poems might be fragments of an epic, which he calls the “Weniad” (Wang 1988: 406

Literature

73–114), but it is a very long way from these lofty elegies in honor of royal ancestors, which may have once played a role in state ceremonies, to the Homeric epics that appear in Greece at roughly the same period of time. As noted earlier, many odes are brief. They emerge originally, many believe, from a world of oral performance and often possess a charming directness: Lush and lovely the peach tree, So very vivid its flowers! That lady goes to her bridal home, Well suited to chamber and kin. Lush and lovely the peach tree, And in such profusion its fruit! That lady goes to her bridal home, Well suited to kin and chamber. Lush and lovely the peach tree, With leaves so very abundant. That lady goes to her bridal home, Well suited to all those of her kin. (Mao Ode #6) Within a general pattern of repetition from one stanza to the next, this poem, like song lyrics elsewhere, shows incremental change. The peach tree remains “lush and lovely” throughout, but its flowers are “so very vivid” in the first stanza, its fruit is “in such profusion” in the second, and “leaves so very abundant” resounds in the third. The aesthetic of so many of the odes is constructed from a nature image, a peach tree in the example here, that both remains and transforms as the poem progresses. Such an image is typically juxtaposed with a human action and therefore points beyond itself: the luxuriance of a peach tree, with its swelling fruit and thick leaves, resonates with the potential fecundity of a lady who “goes to her bridal home” and is “well suited to chamber and kin.” A poem like this one, and so many others in the Odes, appears simple, almost naïve. But such a straightforward reading of the peach-tree ode ignores how generations of educated Chinese readers have understood it. The Odes is frequently referred to as the Mao Odes after an edition of the text compiled by two scholars surnamed Mao in the second century bce. Only one of several competing editions of the text at that time, some only recently coming to light (Kern 2005: 149–154), the Mao Odes eventually prevailed and achieved something like canonical status. In addition to the poems of the Odes anthology, the Mao version contains two introductory documents of disputed authorship, “The Great Preface” and “The Small Preface,” which have exerted profound impact upon the way the odes have been understood. In fact, the first of these prefaces has been described as “the most authoritative statement on the nature and function of poetry in traditional China” (Owen 1992: 37). The two prefaces display a tension between an affective view of poetry and a socio-political one. In what has become the classic definition of poetry in China, the Great Preface says, “That to which the heart is intensely inclined becomes poetry.” A poem is born, the preface says, when “The emotions are stirred within and take on form in words” (trans. by Owen [1992: 40–41]). Poetry, then, is the result of heartfelt intention that cannot be quieted but demands appropriate expression. Having thus advanced what appears to be an affective theory of poetry, the Great Preface proceeds to place the first poem of the anthology within a particular political context; 407

Stephen Durrant

and the Small Preface, following this example, gives a similar context and function for virtually every poem in the anthology under the presumption that intense emotion typically arises from “a specific historical experience” (Kern 2010: 30). In the case of the peach-tree poem, the Small Preface says: “This is what the Queen-Consort has brought about. She felt no jealousy, and so young men and young women were set in good order. Marriage took place in the proper season, and the state had no unmarried people.” The poem in question, according to the Mao reading, thus illustrates the astounding power of a politically well-placed moral exemplar, the queenconsort. It becomes a lesson. What the Mao interpretation does is to remove the odes from an anonymous, generalizable world and to place them in specific, known circumstances, where they are attached to members of the ruling class and are held to express a politically or morally upright message.As one critic has put it:“The prefaces and commentaries do serve a purpose, but not what we would call an artistic purpose. They provide a remedy to the moral peril of an uncontextualized poem” (Saussy 1993: 98). Readers might sense erotic potential in the swelling fruit and luxuriance of the peach tree, juxtaposed as it is with “that lady goes to her bridal home,” and there is latent eroticism in other odes as well (Goldin 2002: 8–26), but such readings are properly submerged beneath a layer of politico-moral interpretation. To put it somewhat differently, reading the odes “properly” becomes a Confucian didactic project, a project enhanced with the Han-time tradition that Confucius (551–479 bce) himself edited the anthology. And just as taming the “Song of Solomon” in Biblical exegesis required clever and ostensibly far-fetched commentary, the same has been true of the Odes. Two further comments are necessary. It is tempting to follow many modern interpreters and reject the political reading of the Odes to return them to a world of pristine folk expression. One must note, however, that even before the Mao interpretations, these poems had assumed an important role within the world of moral and political rhetoric. When Confucius, for example, is quoted in the Analects saying that all three hundred poems can be boiled down to “not going astray” (2.2), he is already countenancing such use. Second, we should not forget, as we have noted earlier, that the Mao prefaces also open a door to an affective reading of the Odes, even as they attempt to circumscribe such a reading. This points to a fundamental question in early Chinese poetics: is poetry primarily a direct expression of individual feeling without wide-ranging implications beyond our shared affective world, or is it primarily an expression of politico-moral intention, sometimes carefully veiled in ways that require the clarification of scholarly commentary? The greatness of the Odes, what has guaranteed a readership across time both in the original language and in translation, is that interpretation, whatever layers of meaning it claims to uncover, is not essential to enjoying the text, even for a reader far removed in time and space. So it is that the folk singer Joan Baez could reach back more than two millennia in her 1968 album “Baptism: A Journey Through Time” and sing a piece from this ancient Chinese classic: Minister of War, We are fierce defenders of the king. So why do you roll us into grief? We have no place to stop and rest. Minister of War, Truly you are not discerning. Why do you roll us into grief? Our mothers want for food. (Mao Odes #185; Baez follows Waley trans. 1996: 158–159) 408

Literature

Historical narrative A century or two after the Book of Odes assumed something like its present form, another genre emerged which we might label “historical narrative.” The Greek word for history, from which the English word derives, originally designates a forensic-like investigation to uncover what actually took place. Hence, we encounter truth claims in the Greek historians Herodotus and Thucydides, however doubtful parts of their accounts are, claims that do not appear in the Chinese tradition until later (Durrant 2005: 93–114). In early China the closest word for “history,” shi, is originally the name of an office, which had among its many tasks maintaining records of state-centered events. There developed among these officials a tradition of using the past for largely didactic purposes, a tradition further stimulated by the intellectual diversity and competition among thinkers of the Warring States period. Such thinkers sought to provide their own views as an antidote to an unhappy present by using the authority of the past, often an imagined past (Lewis 1999: 144–146). Consequently, “historical narrative,” as used here, refers to a type of narrative, one that draws upon some actual or imagined past event and utilizes names, dates, and other apparently real-world details to create an aura of verisimilitude. The first great work of this type is the Zuo Tradition (Zuozhuan). Authorship of the Zuo has been ascribed to Zuo Qiuming, a disciple of Confucius, but this claim is probably no more than an attempt to enhance its authority. More likely, it took form over time, a process that culminated sometime around 300 bce, with certain adjustments in the presentation of the text still later (see Durrant, Li and Schaberg 2016: XXXVIII–LIX). From at least the third century ce, the Zuo has been attached to and transmitted as a lengthy commentary to the much shorter Spring and Autumn Annals (Chunqiu), a list of events concerning the eastern state of Lu that occurred between 722 and 481 bce. Although scribes almost certainly compiled this Lu state Annals, it came to be associated with Confucius, who was believed to have edited it into its present form and imbued it with profound, often subtle meaning. The Zuo, although probably not originally a commentary, at least in the narrow sense of the word, was assuredly written against the background of the Annals, covering virtually the same period of time, quoting from it regularly, commenting upon it from time to time, and showing some stylistic influence. The Zuo is comprised of interspersed “records of events” and “records of speeches” (Liu Zhiji 1980: 3:33). Stylistic inspiration and specific source material for the records of events probably derive from the Annals and similar records, now largely lost, which were written in other states. However, the Zuo arranges the presentation of events into clear causal chains, thereby telling a story, something that does not occur in the early Annals, which simply lists events in a chronology with no connective tissue provided.The language of these narrative chains is extremely terse and typically devoid of extended description or explanation, but it also can “achieve a power and a rapidity which are unparalleled in later Chinese literature” (Watson 1962: 52). By way of contrast, the records of speeches, usually delivered in a court setting, can be surprisingly lengthy and rhetorically elaborate. By combining these two forms, the Zuo initiates a tradition for later historical narrative of mingling an ostensibly external and terse reportage of events with abundant and sometimes extensive quoted speech. Unlike the case of the roughly contemporary Greek historians Herodotus and Thucydides, the Zuo has no first-person narrator, a feature probably deriving from the older impersonal scribal style of the Annals. Despite this absence of a self-proclaimed narrative presence, there are places in the Zuo where the voice of an unidentified “noble man” (junzi) or a known, esteemed figure, most often Confucius himself, intrudes into the text to draw a moral or make an observation. The narrator, insofar as he speaks at all, almost always speaks through others. This feature imparts to the text a tone of authority: the narrator seems simply to report “what happened” 409

Stephen Durrant

and appropriates some other third-person voice when his reportage requires additional weight or explanation. The dominant narrative form in the Zuo record of events is the anecdote, well suited to the pedagogical context in which the Zuo was almost certainly compiled: “the anecdotes may often have been transmitted by teachers who carefully guided their students’ readings or may have been presented as part of persuasions that used the anecdote for didactic purposes” (Schaberg 2001: 172). Even where Zuo narratives appear at first glance longer than an anecdote, such as in the famous description of the travels of Lord Wen of Jin or one of several extended battle narratives, more careful examination reveals them to be constructed from a string of anecdotes unified only by the fact that they concern a single event or character. The following is a typical Zuo anecdote: In the fourth year of Lord Zhuang [690 bce], in spring, in the royal third month, King Wu of Chu arrayed his troops in a Chu-style formation and issued spears to the troops so that they could attack Sui. When he was about to begin a ritual fast, he entered his home and said to his wife, Deng Man, “My heart is unsteady.” Deng Man sighed and said, “Your fortune, king, is at its end! That what is full should be unsteady is the Way of Heaven. Surely the royal ancestors know this. Thus, as you approach a military affair and are about to issue a great command, they make your heart unsteady about it. But so long as the troops do not suffer a defeat, even if you expire during the march, it will prove a blessing for the state.” So the king set out, but died at the foot of Mount Manmu. (Durrant, Li and Schaberg trans. 2016: 145) In this anecdote, a situation is described, set firmly in time and circumstance. A crisis ensues as the king reports his “heart is unsteady,” his wife explains the meaning of this sign, and the king’s death and a subsequently reported Chu victory provide a sort of denouement. On the most superficial level, the story stresses the importance of sacrificing for the state; a good wife should be more concerned about that than worried about her own husband. In Deng Man’s speech one sees, furthermore, a reference to a notion common in China as well as elsewhere in the ancient world: when something reaches a high point, it becomes unsteady and can easily revert to its opposite. These lessons, however, are intrinsic and not the result of a narrator’s moralizing. Some have argued that the main message of the Zuo is that wrongdoers are eventually punished while the good find their reward (Wang 1977: 14), or that the text is really about the value of adherence to proper ritual. While such arguments have some validity, the lengthy Zuo Tradition is complex, perhaps even multi-vocal, and passages can be adduced that counter almost any attempt to reduce it to a simple message (Li 2007: 82–83). The brief story of the King of Chu and his wife Deng Man here illustrates another message one can discern in the Zuo, perhaps one more central to the text than those already offered: the world is a texture of meanings that an insightful person can perceive, allowing him or her to foresee future outcomes. Proper discernment of hidden meaning – that is, “the reading of signs” – is “a ubiquitous phenomenon in the Zuo” that “abstracts patterns and meanings from the chronological flow” (Li 2007: 173). Hidden meaning may be abstracted from a single word found in an Annals entry or from a person’s spoken words, the shape of his face, the way he dresses, his physical gestures such as walking in a certain way, or even the twitching of his index finger, etc. The skilled reader of such signs is most usually an insightful official, although it may on occasion be a commoner, a woman like Deng Man, or even a child – in short, anyone who can discern what apparently small things really imply. 410

Literature

Much of this insightful reading of the world is revealed in the Zuo’s numerous speeches, where characters of discernment so often articulate their insights. The formal power and intricacy of these speeches has been amply demonstrated and labeled “the rhetoric of good order” (Schaberg 2001: 21–56). What is curious is that successful political rhetoric in the Zuo, as elsewhere in early Chinese literature, has little to do with actually persuading the adversary. In fact, many of the cleverest speeches are either rejected outright or exert no impact on policy or behavior. The Zuo, and many other early Chinese texts, are clearly the product of men who see themselves as astute readers and skillful rhetoricians but who believe that they mostly face refractory, morally bankrupt rulers. In fact, it is tempting to conclude that in these texts the best advice is generally that which is not accepted. The Zuo Tradition, however revered it might have been, is only one collection of what must have been a rich world of manuscripts, oral traditions, and pedagogical transmissions utilizing and, as noted earlier, sometimes inventing the past. The increased ferocity of interstate conflict during the Warring States period stimulated a creative search for meaning in the intertwined worlds of historical narrative and masters’ literature. Other collections of historical materials are found in the Discourses of the States (Guoyu), which is largely a collection of speeches and basically covers the same period of time as the Zuo, albeit without the latter’s stylistic and ideological complexity, and the Strategies of the Warring States (Zhanguoce), a text particularly noteworthy from a literary perspective. The latter was actually compiled from a variety of different sources near the end of the first century bce, but most of this source material derives from the last century or so of the Warring States and reflects the heightened conflict and increasing cynicism of that period. The content, somewhat like that of the Zuo Tradition, consists of short narratives dotted with verbal exchanges and lengthy speeches. But whereas a kind of political and moral idealism is often in the background of earlier historical narrative, the Strategies goes in the opposite direction and is concerned simply with what works, realpolitik. This feature has guaranteed the text both an avid readership and also the condemnation of strict Confucians. As is typical with such “naughty books,” it has both attracted and repelled: The Dowager Xuan of Qin loved Mr. Stinky Wei. The Dowager fell ill and was on the brink of death. She issued a command, “When you bury me, you must inter Master Wei with me.” Master Wei worried about it.Yong Rui, acting on behalf of Master Wei, persuaded the Dowager, asking, “Do you think the dead are conscious?”The Dowager replied, “They are surely not conscious.” He said, “If the Dowager’s divine intelligence knows that the dead are not conscious, then why would you for no reason take someone you loved while alive and bury him among the unconscious dead? And if the dead are conscious, then the time the dead king has been accumulating anger has grown long.You, Dowager, will be hard-pressed to make up for your wrong, and what leisure will you have to carry on a private affair with Mr. Stinky Wei!” The Dowager said, “Good,” and desisted from her plan. Yong Rui’s argument in this passage is flawless. He provokes the Dowager to express a belief (“the dead are not conscious”) and then explains that her decision about her lover is inconsistent with such a belief. He then proceeds to consider the other possibility: maybe the Dowager is wrong and the dead are conscious. Even in such a case, Yong Rui argues, she is still doing the wrong thing, offending a deceased husband who awaits her in the next world. Although Master Wei is caught in a dire situation, essentially condemned to death, the Strategies here uses understatement, a characteristic of much early Chinese narrative: “Master Wei worried about it.” Notice, too, that the pejorative name “Stinky Wei,” used in the narrative voice initially, changes 411

Stephen Durrant

in the middle section to the respectful “Master Wei” when the Dowager and Yong Rui engage one another, only to revert, after the argument has essentially been won, to “Stinky Wei.” Like so many attempts at persuasion preserved from early China, this episode presents a court scene, an articulate subject speaking to a somewhat dull-witted person of political power (surely the Dowager should have reached Yong Rui’s conclusion on her own!), but this persuasion makes no pretense, like most speeches in the Zuo or the Discourses of the States, of drawing upon some lofty moral framework.The Dowager yields not because her plan is defeated by considerations of what is morally right but by nothing more than clever words. So many persuasions of this type can be found in the Strategies that some believe it was composed partly from handbooks of artful persuasion produced in a period of time when ministers were more eager actually to win arguments than to take the high road and be ignored (Crump 1964).

Masters literature A rich body of philosophical writings has been passed down from the latter half of the Zhou period. In traditional Chinese bibliography, these are classified as works of “masters,” a term that places more emphasis upon such works as products of teachers than of thinkers or of “lovers of wisdom.” The philosophy of the masters was a response to the breakdown of the moral and political order which had claimed the authority of Heaven; and the crucial question for all of them is not the Western philosopher’s ‘What is truth?’ but ‘Where is the Way?’ the way to order the state and conduct personal life. (Graham 1989: 3) A literary approach to this body of writings is particularly appropriate because “The Platonic distinction between philosophical discourses that aim at a correct representation of the Ideas and literary and rhetorical discourses that do not is entirely absent in early Chinese thought” (Puett 2001: 72). A literary discussion of masters texts might begin with two of the earliest, the Analects (Lunyu) and the Mozi (Mozi). The stylistic contrast between these two texts is useful in thinking about four subsequent works of somewhat greater literary merit: Mencius, Zhuangzi, Xunzi, and Han Feizi. The Analects is primarily a compilation of Confucius’ sayings and of short dialogues involving him and his disciples. Emerging from this compilation, which surely accrued over time (Brooks and Brooks 1998: 201–248), is a vivid picture of a complex personality, Confucius, who as much as his teachings is at the center of the stage. Since Confucius for the most part does not make arguments but simply states “truths,” whether we accept his wisdom depends largely upon how we assess his personal authority, which is of course related to our reaction to his character: The Master said, “A person of knowledge loves water; a person of benevolence loves mountains. The knowledgeable are active; the benevolent are tranquil. The knowledgeable are joyous; the benevolent live long.” (Analects 6.23) If there is any argument here at all it is in the aptness of a type of analogy, knowledge to water and benevolence to mountains, but in the end whether we accept this strange statement depends on how we feel about Confucius himself and our judgment of his wisdom. 412

Literature

The slightly later text Mozi partially arose in response to Confucius’ teachings.While this text is of greatly varied chapters, the central portion, which contains what we might call the “core doctrines” (Johnston 2010: xxxii), differs radically from the Analects both in content and style. Here, we first encounter in China topical essays, with titles such as “Opposing Offensive Warfare,” “Promoting the Worthy,” “Economizing in Burials,” and “Against Confucianism.” Unlike the case of the Analects, where the authoritative voice of a compelling master suffices, these essays are constructed of arguments. At the same time, remnants of a teaching scene do appear, with “the Master Mozi” frequently guiding the essay by provoking or responding to rhetorical questions. Nevertheless, Mozi (fifth century bce) remains a wooden figure with nowhere near the character development provided Confucius in the Analects. In the course of the Mozi’s arguments, authority is frequently provided through historical references and quotations from prestigious earlier texts, especially the Odes and a collection of political decrees and speeches known as the Book of Documents, a practice that will characterize argumentative rhetoric in later masters literature as well.Thus, the contrast we are suggesting here is, on the one side, the master primarily as a charismatic character, and, on the other side, the master as a source of relatively impersonal argument. The masters texts Mencius and Zhuangzi, as different as they are from one another, each have at their center a character who, somewhat like Confucius and unlike Mozi, interacts with disciples, political figures, and other masters, and who is described moving about in physical space. Speaking of the Mencius, one scholar notes, “The novel form of the expository essay is discarded in favor of the dialogue form, which in the Mencius can be considered a conscious return, a holding on to the rhetorical charisma of Confucius, the master staged in dialogue” (Denecke 2011: 154). Still, the return was not complete; Mencius (372?-289? bce) presents arguments like Mozi, sometimes at considerable length, rather than simply stating truths in the manner of Confucius. But unlike in the Mozi essays, other named persons typically provoke Mencius’ arguments. “What were ranks and salaries like under the Zhou?” asks Bei-gong Qi. “Dare I ask about friendship?” ventures Wan Zhang. The same types of questions occur in the Analects, but Confucius’ answers are always brief, whereas Mencius sometimes expounds at length. The teacher’s authority alone is no longer sufficient in the increasingly competitive intellectual world of late Warring States China. Two features of Mencius’s argumentation deserve comment. First, like Mozi, he repeatedly turns to the texts of the past, but when Mencius cites these sources, he frequently supplies his own interpretation: the past cannot just be referenced, it must be explained, at times even explained away (Denecke 2011: 164–167).Through such acts of textual exegesis, Mencius establishes a Confucian historiography, which is predicated upon the past “correctly” interpreted so as to provide a blueprint for the present. “Yao gave the empire to Shun. Is this so?” Wan Zhang asks, seizing on the problem that the tradition claims these early sage-emperors passed rule of the kingdom to the most worthy man and not automatically to a son. A short dialogue ensues in which Mencius argues that this did not happen. Technically it was Heaven and not Yao, he explains, who determined the succession – only Heaven can “give the empire.” Part of this new historiography is a claim, first found on the pages of the Mencius, that Confucius “produced the Spring and Autumn Annals,” with the result that “rebellious ministers and unruly sons were frightened” (6.9). As a consequence of this questionable declaration, Confucius assumes the role of a great historian, the first to have correctly understood the past and provided it as a corrective to present behavior. Second, the use of analogy is one of Mencius’ most provocative rhetorical tools. His dispute concerning human nature with his contemporary Gaozi is rightfully famous not only because it concerns an issue that remains relevant today (are we naturally good, bad, or somewhere 413

Stephen Durrant

in between?), but because it raises questions about the rhetorical function of analogy in early Chinese argument. Gaozi tries to counter Mencius’s fundamental belief that humans are by nature good through arguing that human nature is neutral and may become good or evil, just as water flows to the east or west depending upon where a channel happens to be opened. To this, Mencius replies: Water truly does not distinguish between east or west; but does it not distinguish between up and down? Human nature is good in the same way that water flows downhill. There is no person who is not good; there is no water that does not flow downward. (11.2) It appears on the surface that analogy, which typically in Western rhetoric is considered only to provide clarification, has hijacked the argument? Is Mencius only showing that he too can cleverly turn water into an analogy in support of his own position? “It is perhaps worth pointing out,” says D.C. Lau, “that the use of analogy is often the only helpful method in elucidating something which is, in its nature, obscure” (Lau 1970: 263). Human nature is surely one such obscurity, and the fact that flowing downward is more intrinsic to water than flowing east or west maybe helps us perceive some truth that applies both to water and to human nature. But this analogy might imply more: could it be that water and human nature really do somehow resonate with one another so that in some way what is true of one is also true of the other? Note in the passage quoted earlier from Confucius, a link is established between water and benevolence. Whatever the precise function of analogy in the Mencius, and this remains a perplexing problem (Chan 2002: 122), it adds literary color to a text that has long been held as a model both of clear expression and of clever argumentation: The trees on Ox Mountain were once beautiful. Because they are on the outskirts of the capital city, axemen have cut them down. So could they still be beautiful? Given the respite of days and nights, and the moistures of rain and dew, could sprouts not spring up there? But then sheep and cattle are grazed on the mountain, and it has become desolate. When people see its desolation they think no lumber was ever there. How could this be the nature of the mountain? And though a similar state should exist among men, how could it be they do not have a noble and just heart? (11.8) The human being, like Ox Mountain, is beautiful by nature, Mencius here argues. In both cases, however, opposing forces can destroy that original beauty. Perhaps for Mencius the aesthetic pleasure one derives from an analogy such as this is, in itself, persuasive. Appearing approximately at the same time as the Mencius, the Daoist text Zhuangzi both dazzles and perplexes. Experts still debate what might constitute its central, unifying message, since “Almost any part of the book may be taken literally or metaphorically, and then at any of several levels of metaphor” (Mote 1971: 81). Adding to the difficulty is the fact that the Zhuangzi is more an anthology than a unified text (Mair 1994 xlv). While this may be true of other early Chinese texts as well, the Zhuangzi is a particularly striking example of a work that accrued over time and shows not only philosophical inconsistencies but also stylistic variation. Nevertheless, one recent Chinese university textbook declares the Zhuangzi to “possess the greatest literary value” of all the pre-Qin masters texts (Yuan Xingpei, I.98). Like the Analects and the Mencius, the Zhuangzi is unified by the presence of a compelling master, who appears in a wide array of 414

Literature

dialogues and other, sometimes highly imaginative, situations. Perhaps the best summary of the Zhuangzi’s style is found in the work itself: With exaggerated explanations, wild words, and boundless expressions, Zhuangzi at times gave free reign to his ideas but was not partial and did not look at things from a single point of view. Since he believed the people of the world to be obscure and muddy, he could not engage them with stern sayings. So he used flowing words for their expansiveness, repetitions for their truth value, and extended metaphors for their breadth. (Zhuangzi jishi, 2.1204) That Zhuangzi (369?–286? bce) does not utilize “stern sayings” is probably meant to target Confucius. His work is instead appropriately described as “exaggerated,” “wild,” “boundless,” “spontaneous,” and “expansive.” But perhaps most important in this description is the term translated as “extended metaphor.” The Chinese for this, yuyan, is frequently encountered in Chinese-language discussions of the Zhuangzi and literally means “imputed words” – words filled with some other meaning beyond the surface level, including such tropes as allegory, fable, metaphor, and analogy. From the first page of the Zhuangzi, on which the great fish Kun swimming in the dark Northern Sea transforms into a Peng bird and soars across the world with wings hanging down like great clouds, we know we have entered a textual world unlike anything encountered before in China. In what follows this first page, the reader encounters a cast of characters known from other texts – legendary heroes like the Yellow Emperor, Shun, and Yu, fully historical persons such as the feudal lords of the Spring and Autumn and Warring States periods, and fellow philosophers Confucius, Yan Hui, and Laozi himself. But the general verisimilitude provided to such characters in other early texts dissolves before our eyes as they become part of a blatantly fictionalized Zhuangzian world: The Yellow Emperor was wandering north of Redwater, when he ascended the heights of K’un-lun and gazed toward the south. As he was returning home, he lost his pearl of mystery. Knowledge was sent to search for the pearl, but he couldn’t find it. Spidersight was sent to search for the pearl, but he couldn’t find it. Trenchancy was sent to search for the pearl, but he couldn’t find it, whereupon Amorphous was sent and he found it. “Extraordinary!” said the Yellow Emperor. “In the end, it was Amorphous who was able to find it.” (Trans. by Mair [1994 105]) While the Mencius and the Zhuangzi both show the influence of Analects in presenting a strong personality engaged in teaching or conversing with others, the masters text Xunzi veers toward the Mozi’s topical-essay style, albeit making it more vibrant and appealing. Rather than quoting the master Xunzi (313?–238? bce) or depicting him from a third-person perspective, the essays appear to be in Xunzi’s own voice. Stylistically his “writing is succinct and lucid, his philosophical positions original and reasoned” (Goldin 1999: xiii). One of the most important building blocks of Xunzi’s thought is found in his essay entitled “A Discussion of Heaven.” This essay concerns an issue that has been broached by other early Chinese thinkers but without Xunzi’s intense focus. His goal is to rebut the idea encountered, for example, in the Mozi that Heaven blesses and curses in some conscious, reactive fashion. To Xunzi, Heaven “plays no active role in terrestrial affairs” (Goldin 1999: 51). Man must make 415

Stephen Durrant

certain he acts in accord with Heaven’s seasons, and in accord as well with the natural endowments Heaven has provided to him, but he should not believe Heaven acts through portents or can be swayed to human will through such rituals as prayer: “When stars fall or trees make strange sounds, people of the state are afraid and ask, ‘Why is this?’ I say, ‘There is no why. These are the changes of heaven and earth, the transformations of Yin and Yang, things that occur rarely’ ” (17:333). And elsewhere, “You pray and it rains. Why is that? I say, ‘There is no why. It is the same as not praying and it rains’ ” (17.334). Several features of the Xunzi’s style can be seen here. Most obviously, the author is flying much closer to the ground than Zhuangzi. His words and examples are typically drawn from the workaday world, as we would expect from a Confucian. Like Mozi, Xunzi makes use of rhetorical questions, but they are less mechanical, more distant from a simple scene of instruction. Like other early Chinese thinkers, he refers to figures from the past and quotes from such prestigious texts as the Odes and the Documents, but historical references and quotations are typically brief and do not divert attention from his central arguments. In fact, his essays become a driving force. They have been called “unprecedentedly elaborate” (Lévy 2000: 13), but they do not build an architectonic structure of premises, examples, and conclusions so much as move sideways, with the same argument repeating in slightly different form and leading to interconnected conclusions: “Whoever can distinguish between the actions of Heaven and those of humans is worthy to be labeled ‘the highest type of man’ ”; “Only a Sage does not try to understand Heaven”; “Hence the skilled man has things he does not ponder (i.e., Heaven)”; “Heaven has its constant way; earth has its constant dimensions, the gentleman has his constant demeanor”; etc. Such conclusions do not rest one atop the other in some logical hierarchy. Instead they become much like variations on a theme – “Goldberg Variations” in prose, so to speak. Xunzi, whatever we think of his style, can be quite convincing. If the reader simply wears down, he often wears down to Xunzi’s point of view. In some ways late Zhou philosophical writing culminates in the Han Feizi, the work of one of Xunzi’s disciples. Han Feizi inherits both Zhuangzi’s creativity and his master’s argumentative skill. An irony exists at the core of this text: Han Feizi displays his genius in service of an authoritarian ideology, usually called in English “legalism” (fajia), which could hardly tolerate the very kind of stylistic abandon he deploys. In fact, Han Feizi (280?–233 bce) suffered death at the very hands of those he strove to empower, not the last person to suffer such a sorry fate. Some indication of the conflict inherent in the Hanfeizi can be discerned in his condemnation of the tendency of disputers, particularly Confucians, to use past models as a corrective to the present. After discussing the different circumstances of each past age and the unique ways rulers of those times responded to such circumstances, he concludes, “This is why the Sage does not hope to cultivate ancient ways, nor does he just follow along with that which has always been so.” To simply follow precedent, he argues, is to ensure that in the end one will become a laughingstock: There was a farmer from the state of Song. In the middle of his field was a stump. A rabbit, running along, banged into the stump, broke its neck and died. So the farmer put down his plow and waited by the stump, hoping that he would obtain yet another rabbit. But he did not get another rabbit and was laughed at by the people of Song.Yes, it is true, “Now if you desire to use the government of the former kings to rule the present world, it is exactly like the farmer waiting by the stump.” (Hanfeizi 49.1) This kind of anecdote, of which there are many in the Hanfeizi, is worthy of the Zhuangzi, but Hanfeizi makes sure we do not miss the point. He is not just entertaining us and hoping we 416

Literature

intuit the philosophical messages inherent in that entertainment; he wants his point to be crystal clear. The story of the foolish farmer from Song, sitting by a stump hoping to make a living by collecting rabbits who accidently run into the stump, as one rabbit just happened to do, teaches us that to rely entirely on precedent is stupid. The irony, though, is that no early Chinese thinker uses precedent more abundantly than Hanfeizi himself. His work is a virtual encyclopedia of stories drawn from the past. To make the point that a sane ruler might not listen to a wise advisor at first meeting, he turns to the very type of precedent he elsewhere condemns: In very ancient times there was the ruler Tang, who was a Sage of the highest type, and Yiyin was a wise man of the highest type. Now, when a wise man of the highest type attempted to persuade a Sage of the highest type, although he tried seventy different persuasions, he was not accepted. He himself had to hold the cooking vessels and act as a butcher to draw near and become familiar to Tang, until finally Tang learned of his true worthiness and made use of him. (3.2) Hanfeizi takes the disputation of Warring States period to a high point, drawing upon the rhetorical and literary features of the writers who preceded him. And he does this in the service of a power, the state of Qin, which finally united the empire and attempted to enforce not only political but also intellectual conformity, with Hanfeizi and many others as victims. The Qin dynasty did not last long, only fifteen years, but the Han dynasty that followed, while moderating the harsh policies of the Qin, witnessed a new literary eclecticism that attempted to gather the disunity of the past into a more coherent vision. We cannot leave this chronological account of the masters and the contrast that has partially ordered it without a comment about a text that has been translated into English much more frequently than any other: Laozi’s Daode jing (The Way and Its Power). It is appropriate to treat this text somewhat apart from the other masters literature precisely because it is so unlike anything else from this early period. Daode jing makes no reference to specific events, other persons, or Laozi himself. To speak of characterization or argumentation, in this case, is beside the point. We do not really know who Laozi was, or even if he was, and whether we accept the wisdom of this text is mostly a function, like certain wisdom literature or scripture elsewhere, of simply whether it inspires us. Surely it did have impact in early China, since Hanfeizi dedicates a chapter to it and several versions of this text have been excavated from early Chinese tombs. Perhaps Daode jing was considered to carry a mysterious, almost religious, power and is best read as a “series of highly elusive aphorisms and poems” (Puett 2001: 78), which makes use of imagery and symbol that can haunt the reader long after the book has been closed. It is rare to hear a modern reader say that Analects, for all its importance, has “changed my life,” but students do occasionally say this of the Daode jing. But what, really can, or should, be said of a text that begins The way that can be taken as a way, Is not the unchanging Way. The name that can be used as a name, Is not the unchanging Name. The nameless is the source of heaven and earth, The named is the mother of all things. (Ch. 1) 417

Stephen Durrant

The literature of unity Less than a century after the founding of the Han dynasty, Sima Tan (d. 110 bce), who served as grand scribe in the imperial bureaucracy, decided to collect and preserve the records of the past, perhaps with the intention to compose a comprehensive, unified history, a sort of textual correspondent of the new empire. We do not know how far his project advanced, but the task eventually fell to his son, Sima Qian (145?–87?). Sometime around 90 bce, the latter completed the monumental Records of the Historian (Shiji), which covers over two thousand years of the Chinese past in 130 chapters. Sima Qian explains that his work is more a compilation than an original creation, and in fact much of the text is either quoted or rephrased from previous works, including the Zuo Tradition, Strategies of the Warring States, and various masters writings. The structure of the Records of the Historian is, however, different from any earlier work and will exert enormous influence on subsequent Chinese historiography. Sima Qian divides his text into five sections: the first centers upon the central court and is organized by dynasty and ruler; the second is a series of chronological tables showing the temporal relationships between events and political leaders; the third contains topical essays about key political or cultural institutions; the fourth gives attention to feudal leaders and other men of hereditary power; and the fifth, by far the largest of the sections, concerns other important individuals and certain ethnic groups. From a literary point of view, the most significant of these five sections is the fifth, which Sima Qian labels the “Arrayed Traditions” (liezhuan). Here he takes the older form of the zhuan, simply “what has been passed down” or “tradition,” and organizes it largely around individual human lives, thereby creating something akin to what we might think of as “biographies.” Even outside of this particular section of the Records of the Historian, Sima Qian frequently gives prominence to influential persons, which has led some to describe his history as “humancentered.” For example, the chapters concerning Confucius or the famous anti-Qin rebel Xiang Yu (d. 202 bce), among others, are not included in the “Arrayed Traditions” section but possess the same general biographical structure. This new human-centered literary form will be highly productive, with a tremendous proliferation of biographical literature such as “Biographies of Eminent Women” (Lienu zhuan), “Biographies of Divine Transcendents” (Shenxian zhuan), “Biographies of Buddhist Monks” (Gaoseng zhuan) and many others in the centuries just after Sima Qian. In the portrayal of his individual characters, Sima Qian goes well beyond anything encountered previously. Two seemingly conflicting features of his biographies deserve comment: first, he shows the continuity in a person’s behavior across a lifetime, sometimes providing at the beginning of a biography an anecdote from the character’s childhood that reveals an essential and enduring personality trait; second, he acknowledges the essential mystery of human behavior by showing that despite this general continuity of personality, people are still not entirely predictable. When he is a child, Xiang Yu’s uncle takes him to view the imperial procession of the powerful First Qin Emperor. Rather than being impressed, Xiang Yu, Sima Qian recounts, watches the emperor and then says to his uncle, “That one can be seized and replaced!” “Don’t speak such reckless words,” his uncle replies, “Our whole clan will be exterminated!” (Shiji 7.296). This scene foreshadows all of Xiang Yu’s future rebelliousness and bold courage. Who would have thought, then, that Xiang Yu would be undone, at least in part, by one moment of hesitancy, when he has his rival under his power but simply cannot bring himself to act dramatically? Quite unlike Xiang Yu, Confucius, as a child, is portrayed in his “biography” playing with sacrificial vessels, setting them out as if engaging in ritual (47.1906), behavior completely consistent with most of what follows. Nevertheless, Confucius is ensnared in several events, one the harsh execution of entertainers, that leave the reader 418

Literature

wondering if he has always acted in a ritually correct fashion. In short, Sima Qian’s characters are consistently complex, “rounder,” to use a literary term, than we have encountered in Chinese literature before. Perhaps most significantly, Sima Qian creates himself as his text’s main character. He intrudes frequently in his history as a first-person or self-identified presence, breaking entirely with the scribal tradition of anonymous narration found in the Zuo Tradition. His comments, many of them as much emotional outbursts as evaluations, dot the text (Shankman and Durant 2000: 152fn27). Each time he reads about Mencius’ interview with King Liang of Hui, “I put down the document and sigh.” When he visits the household of Confucius, he becomes so enraptured, “I was unable to depart.” Elsewhere he groans about those “who are slandered by lesser men.” The presence and strength of Sima Qian’s personality as encountered on the pages of Records of the Historian has inclined many readers to seek ways Sima Qian’s own dramatic life experience might have shaped his account of China’s past. That experience is laid out in the last chapter of his history, in which Sima Qian provides a short autobiography and then explains how and why he came to write the Records of the Historian. This autobiography is typically supplemented with a self-revealing letter, the “Letter in Response to Ren An,” that Sima Qian supposedly wrote to a friend in prison awaiting execution (Durrant 2016: 32). Sima Qian explains to his friend that he himself had once been condemned to death for “slandering the emperor.” His sentence was eventually reduced one grade to castration, a punishment any decent man was expected to avoid by committing suicide. Sima Qian, however, suffered his dreadful punishment and explains that he chose to remain alive rather than commit suicide in order to complete his history of the past. In a maneuver to transform personal tragedy into a form of literary potency, Sima Qian goes on to argue in two different places that great literature is always the result of frustration arising from physical deformity, mutilation, imprisonment, or some other tragic circumstance, a belief that reverberates in China down through the ages (Li 2016: 97–100). Perhaps the suffering author Sima Qian admires most is the late Warring States poet Qu Yuan (ca. 340–278 bce), whose biography is one of the highpoints of the Records of the Historian. It might seem strange to consider Qu Yuan and his poetry here, when he lived a full century before Sima Qian, but the way Qu Yuan and his poetry have been understood in subsequent generations is very much a function of his portrayal in the Records of the Historian. One might go so far as to say that our knowledge of Qu Yuan, however historical he may have been, like so many other figures from China’s distant past, is largely a creation of Sima Qian. Sima Qian introduces Qu Yuan as a minister to King Huai (327–299), ruler of the powerful southern state of Chu. “His knowledge was broad, his memory strong, he clearly understood how to bring order to chaos, and he was practiced in rhetorical arts,” the historian says (84:2481). Qu Yuan has all the desirable qualities necessary to serve his lord. But immediately after he is introduced so positively, he is slandered by lesser talents and exiled from the capital. In frustration, Qu Yuan turns to poetry. Later recalled to his home state, he will be slandered and banished again, even though subsequent history vindicates the loyalty of his motives and the correctness of his advice. Despairing in the face of such rejection, Qu Yuan takes his own life, virtually transforming his poetry into an extended suicide note and inclining readers to regard him as a primary example of what has been called the theme of “the scholar’s frustration” (Wilhelm 1957). Qu Yuan’s poetry is found in an anthology entitled the Songs of Chu (Chuci).The most important and formidable of his poems is “Encountering Sorrow” (Lisao), at 372 lines, more than three times longer than the longest of the earlier odes.The poetry in the Songs of the South has a variety of line-lengths, all different from the earlier four-character rhythm and almost all characterized 419

Stephen Durrant

by the regular repetition of a caesura made up of a kind of sighing sound (xi). In the case of “Encountering Sorrow,” two lines of six syllables are joined in the middle by xi, which yields a more languorous, richer verbal flow than the staccato rhythm of the odes. The world of Qu Yuan’s “Encountering Sorrow” is intricate. Filled with difficult vocabulary and arcane mythological references, the gender-shifting narrative voice floats between worshipful pursuit of a seductive but fickle goddess and a heartfelt fear of rejection, between an astounding desire to soar to celestial heights and an inclination to fall into the depths of self-pity and despair: My years are not yet full, And my seasons have not run out. But I dread the shrike’s early warning Might cause all herbs to lose their fragrance. How my jasper belt does glitter, But commoners crowd about to hide its glory. It is the distrust of such partisan men, I fear will break it in their jealousy. (“Encountering Sorrow,” lines 297–304) “Encountering Sorrow” has been described as “an autobiographical poem, representing real persons and events allegorically, in which the poet complains of undeserved rejection by his king, consults shamans, and embarks on a shamanistic ‘flight’ in an unsuccessful quest for a divine mate” (Hawkes 1985: 38). One wonders, however, if we would arrive at precisely this description without Sima Qian’s biography, which makes Qu Yuan’s poetry an expression of personal political frustration. “Encountering Sorrow” is surely one of the world’s great works on the topic of disappointment. No matter how much we may try to transcend the limits of our own time and space and achieve some moment of supreme union that is truly worthy of what we think, in our illusions, we just might be, we are bound to fail – at least so this poem seems to say. Despite its difficulties and ambiguities, it is a poem that has resonated with centuries of Chinese readers, who glimpse something in this work that captures more than the just the experience of one man. The rich descriptive verbiage of the Songs of the South exercises an important influence on the rise of a new literary form that will come to be associated with the Han period and even considered “the essence of Han literature” (Knechtges 1997: 1): the Fu, often appropriately referred to as “the Han Fu” and translated variously as “rhapsody,” “rhyme prose,” or “poetic exposition.”The Fu can be lengthy, and its poetic line and rhythm are much closer to the Songs of the South than to the Odes. While some of the earlier Fu carry the same tone of personal sorrow and frustration encountered in Qu Yuan’s poems, the genre quickly becomes mostly descriptive, in fact a celebration of the material and verbal richness of the time, a form of expression much appreciated at court during the Han period. Such detailed, ornate, hyperbolic descriptions of the glories of a garden, an imperial progress, a hunt, the capital city or some other scene can be of greater interest to material historians and lexicographers than readers seeking aesthetic pleasure. These literary compositions may indeed have functioned “as encyclopedias or lexicographies as much as descriptions” (Connery 2001: 231). One of the most skilled authors of Fu was the deeply learned scholar and philosopher Yang Xiong (53 bce–18 ce). He eventually renounced such compositions, and when asked about Fu replied, “They always go too far. . . [they] are stunningly beautiful but excessive” (Nylan trans. 2013: 25).The Fu were obviously recited before a highly literate audience and were often meant to dazzle with erudition more than to move with genuine feeling, but some pieces, especially 420

Literature

those written outside of a court setting and expressing a sense of disappointment or ending with a moral lesson, continue to speak with power to a modern reader. As the Han progresses, a general tendency toward increasingly learned and what some have called “orthodox” literary expression proliferates in genres other than the Fu. This was a time when the government began to sponsor mastery of a group of Confucian classics (see Nylan 2001). Moreover, the court commissioned a group of scholars, Liu Xiang (77–6 bce) and his son Liu Xin (ca. 50 bce-23 ce) chief among them, to undertake a collation and bibliographic classification of texts in the imperial library, a classification that has exercised deep influence upon the way knowledge has been understood and organized in traditional China. In the course of this activity, Liu Xiang drew upon earlier works to produce a number of new texts, chief among them his Biographies of Eminent Women. One can, of course, criticize the somewhat flat, moralistic presentation of women in this volume, but one must also concede that as the earliest book to focus attention almost exclusively on women, it occupies a critical place in the history of Chinese literature. The increased formality of literary presentation is also reflected in a new work of historical narrative, Ban Gu’s massive History of the Han (Hanshu). Ban Gu drew extensively upon the work of his historian forerunner Sima Qian but limited his work to just the first half of the Han dynasty. A long scholarly debate has ensued concerning the relative strengths and weaknesses of the Han dynasty’s two great historians, Sima Qian and Ban Gu. One early commentator on Sima Qian’s Records of the Historian conceded that Ban Gu’s style was “more lucid” and his phrasing more “patterned” (Klein 2010: 76). It is probably fair to say that while Sima Qian’s prose is livelier, Ban Gu’s is statelier and exercised greater influence upon subsequent historical writing. The History of the Han also exhibits increased attention to the Confucian tradition (Clark 2008: 13–15), which by Ban Gu’s time had become more or less a state orthodoxy. As Ban Gu worked on his vast history, an acquaintance, the brilliant and curmudgeonly Wang Chong (27–ca 100) continued the earlier tradition of masters literature with an imposing philosophical work entitled Balanced Discourses (Lunheng). Wang, contrary to the conservative trends of his time, was an ardent advocate of originality. His work is a striking defense of masters literature and displays his own capacity for creative, highly disputatious thought (Denecke 2011: 86). Wang’s essays are lively and leave a modern reader simultaneously impressed with the piercing skepticism of his mind and, on occasion, his gullibility as he tries, in essay after essay, to debunk what he considers false beliefs. In his fascinating “Essay on Death,” for example, he marshals an array of proofs that the dead have no consciousness, one of them with a familiar and somewhat perplexing use of analogy: The nature of heaven and earth is such that one can light a fire a second time, but one cannot cause a fire that has burned out to ignite again. You can give birth to another person, but you cannot cause a dead man to come back to life. If you cause a burned out fire to ignite again, then I would at least suspect a dead man could again take on living form. But based upon the comparison of an extinguished fire not again being ignited, it is clear that a dead person cannot again take form as a ghost. (20) Despite the comprehensive and more expansive voices of the Han period, the lyrical voice, encountered centuries before in some of the simple, emotive poetry like that on the peach tree, had never gone away. Popular verse, retrospectively called yuefu or “Music Bureau” poetry, was supposedly collected under the auspicious of an official bureau Emperor Wu established to help government leaders understand public sentiment, although most of the works now 421

Stephen Durrant

spoken of as belonging to this genre are clearly later than the Han (Allen 1992: 50). But genuine Han poetry is encountered in both Sima Qian’s Records of the Historian and especially Ban Gu’s History of the Han. Some of this Han verse has a pentasyllabic line, an entirely new poetic form that will prove fruitful in the centuries to come. This Han lyrical tradition culminates in a remarkable effervescence of poetry in the last century of the Han that includes such prominent writers as Zhang Heng (78–139), Cai Yong (133–192), and the Jian’an poets, who were active in the last reign of the Han dynasty (196–220) (Cutter 2001: 249–251). Particularly important and influential, however, are a group of brief works entitled “The Nineteen Old Poems” (gushi shjiu shou). Preserved in a sixth century anthology but probably originating in the last century of the Han (Diény 2010: xxxvii–xxxviii), these poems are all pentasyllabic. Their two major themes of separation and death perhaps reflect well a specific time of political decline, but such universal themes as these have guaranteed an avid readership in the centuries to come: Crossing the river, I gather lotus In this orchid marsh so full of fragrant grass. But to whom would I send what I gather? For the person I love travels a distant path. I turn and look back toward my homeland, And that long road stretches on forever. Of one heart, we live apart, In sorrow, until we reach old age. (Nineteen Old Poems #6) The “Nineteen Old Poems” have been considered a link in a powerful current of poetic expression that extends from the Odes and other poetry of the early period down to the great flowering of poetry that will occur in the Six Dynasties and Tang periods: “harvesting the heritage of the past, they announce the future” (Diény, xii).

Note 1 All translations are my own unless otherwise noted.

Works cited Allen, Joseph R. (1992) In the Voice of Others: Chinese Music Bureau Poetry, Ann Arbor: Center for Chinese Studies Publications, University of Michigan. Brooks, E. Bruce and Taeko Brooks (1998) The Original Analects: Sayings of Confucius and His Successors, New York: Columbia University Press. Chan, Alan Kam-leung (2002) Mencius: Contexts and Interpretations, Honolulu: University of Hawai’i Press. Clark, Anthony E. (2008) Ban Gu’s History of Early China, Amherst: Cambria Press. Connery, Christopher Leigh (2001) “Sao, Fu, Parallel Prose, and Related Genres” in Victor Mair (ed.) The Cambridge History of Chinese Literature, 223–247, New York: Columbia University Press. Crump, James I. (1964) Intrigues: Studies of the Chan-kuo Ts’e, Ann Arbor: University of Michigan Press. Cutter, Robert Joe (2001) “Poetry from 200 b.c.e. to 600 c.e.” in Victor Mair (ed.) The Cambridge History of Chinese Literature, 248–273, New York: Columbia University Press. Denecke, Wiebke (2011) The Dynamic of Masters Literature, Harvard-Yenching Sinological Institute Monograph series no. 74, Cambridge: Harvard University Press. Diény, Jean-Pierre (2010) Les dix-neuf poèmes anciens, Paris: Les belles lettres. Durrant, Stephen (2005) “Truth Claims in Shi ji” in Helwig Schmidt-Glintzer, Achim Mittag, and Jorn Rusen (ed.) Historical Truth, Historical Criticism, Historical Ideology: Chinese Historiography and Historical Culture from a New Comparative Perspective, 93–114, Leiden : Brill.

422

Literature ________ (2016) “Seeking Answers, Finding More Questions” in Stephen Durrant, Li Wai-yee, Michael Nylan and Hans van Ess, The Letter to Ren An and Sima Qian's Legacy, 30–50, Seattle: University of Washington Press. Durrant, Stephen, Wai-yee Li and David Schaberg (trans.) (2016) Zuo Tradition/Zuozhuan: Commentary on the “Spring and Autumn Annals,” Seattle: University of Washington Press. Graham, A.C. (1989) Disputers of the Tao: Philosophical Argument in Ancient China, Chicago: Open Court. Goldin, Paul Ratika (1999) Rituals of the Way:The Philosophy of Xunzi, Chicago: Open Court. ________ (2002) The Culture of Sex in Ancient China, Honolulu: University of Hawai’i Press. Hawkes, David (trans.) (1985) The Songs of the South: An Anthology of Ancient Chinese Poems by Qu Yuan and Other Poets, Harmondsworth, Middlesex, England: Penguin. Johnston, Ian (trans.) (2010) The Mozi, New York: Columbia University Press. Kern, Martin (2005) “The Odes in Excavated Manuscripts” in Martin Kern (ed.), Text and Ritual in Early China, 149–193, Seattle and London: University of Washington Press. ________ (2010) “Early Chinese Literature, Beginnings through Western Han” in Kang-I Sun Chang and Stephen Owen (ed.) The Cambridge History of Chinese Literature, Volume 1: To 1375. Cambridge: Cambridge University Press. Klein, Esther (2010) “The History of a Historian: Perspectives on the Authorial Roles of Sima Qian,” unpublished dissertation, Princeton University. Knechtges, David R. (1997) “Introduction” to Gong Kechang, Studies on the Han Fu, trans. and ed. by David R. Knechtges, 1–50, New Haven: American Oriental Society. Lau, D.C. (1963) Lao Tzu:Tao Te Ching, Harmondsworth and New York: Penguin Books. ________ (1970) “On Mencius’ Use of Analogy in Argument” in Mencius, 235–264, Harmondsworth and New York: Penguin Books. Lévy, André (2000) Chinese Literature, Ancient and Classical, trans. by William H. Nienhauser, Bloomington and Indianapolis: Indiana University Press. Lewis, Mark Edward (1999) Writing and Authority in Early China, Albany: State University of New York. Li Wai-yee (2007) The Readability of the Past in Early Chinese Historiography, Cambridge, Massachusetts: Harvard University Asia Center. ________ (2016) "The Letter to Ren An and Authorship in the Chinese Tradition" in Stephen Durrant, Li Wai-yee, Michael Nylan and Hans van Ess, The Letter to Ren An and Sima Qian's Legacy, 96-123, Seattle: University of Washington Press. Liu Zhiji (661–721 C.E.) (1980) Shitong tongshi (A Comprehensive Explanation of “A Comprehensive Study of History”), Taibei: Qihai. Mair,Victor H. (trans.) (1994) Wandering on the Way, New York: Bantam Books. Mote, Frederick W. (1971) Intellectual Foundations of China, New York: Alfred A. Knopf. Nylan, Michael (2001) Five “Confucian” Classics, New Haven:Yale University Press. ________ (trans.) (2013) Exemplary Figures: Fa Yan, Seattle: University of Washington Press. Owen, Stephen (1992) Readings in Chinese Literary Thought, Cambridge, Massachusetts: Council on East Asian Studies Harvard University. Puett, Michael (2001) “Philosophy and Literature in Early China” in Victor Mair (ed.) The Cambridge History of Chinese Literature, 70–85, New York: Columbia University Press. Saussy, Haun (1993) The Problem of a Chinese Aesthetic, Stanford: Stanford University Press. Schaberg, David (2001) A Patterned Past: Form and Thought in Early Chinese Historiography, Cambridge, Massachusetts: Harvard University Asia Center. Shankman, Steven and Stephen Durrant (2000) The Siren and the Sage: Knowledge and Wisdom in Ancient Greece and China, London: Cassell. Shaughnessy, Edward L., (1991) Sources of Western Zhou History: Inscribed Bronze Vessels, Berkeley, Los Angeles, Oxford: University of California Press. ________ (1997) Before Confucius: Studies in the Creation of the Chinese Classics, Albany: State University of New York Press. Shih, Vincent Yu-Chung (trans.) (1983) The Literary Mind and the Carving of Dragons, Hong Kong: The Chinese University Press. von Falkenhausen, Lothar (2011) “Royal Audience and Its Reflections in Western Zhou Bronze Inscriptions” in Li Feng and David Prager Branner (eds) Writing and Literacy in Early China: Studies from the Columbia Early China Seminar, Seattle: University of Washington Press. Waley, Arthur (trans.) (1941) Translations from the Chinese, New York: Alfred A. Knopf.

423

Stephen Durrant _______ (trans.) (1996) The Book of Songs:The Ancient Chinese Classic Poetry, edited with additional translations by Joseph R. Allen, New York: Grove Press. Wang, C.H. (1988) From Ritual to Allegory: Seven Essays in Early Chinese Poetry, Hong Kong: The Chinese University Press. Wang, John C.Y. (1977) “Early Chinese Narrative: The Tso-chuan as Example” in Andrew H. Plaks (ed.) Chinese Narrative: Critical and Theoretical Essays, 3–20, Princeton: Princeton University Press. Watson, Burton (1962) Early Chinese Literature, New York: Columbia University Press. Wilhelm, Hellmut (1957) “The Scholar’s Frustration: Notes on a Type of Fu” in John King Fairbank (ed) Chinese Thought and Institutions, Chicago: Chicago University Press. Yuan, Xingpei (2011) Zhongguo wenxue shi (A History of Chinese Literature), 2 vols., Dingyue Publishers.

424

20 ART

WANG HAICHENGART

Wang Haicheng

A short chapter introducing four millennia of art cannot attempt a connected history. In this essay I will instead comment on fifteen objects, as though I were organizing a small exhibition, taking them in chronological order and grouping them according to three major phases in Chinese history. Sampling a range of art forms, from ceramics to painting to beautiful writing, draws our attention to materials and hence to the dialogue between what the artist thinks and what his or her hands do. Connections between the history of art and the history of technology emerge naturally from such a focus. Important also are the purposes of patron and artist, in other words the functions that the works were designed to serve: what audience were they aimed at, and what visual impact were they meant to have on it? The contrasts among the objects will raise other issues as well, including broad changes over time and the regional diversity natural in a land as vast as China.

The late Neolithic: fourth and third millennia bc The first group of objects, Figures 20.1–20.4, were made by farming societies with different degrees of social stratification but without metallurgy or writing, in other words, Neolithic societies. The materials represented by the four objects are clay and jade, materials that survive in abundance because they are durable. There is archaeological evidence also for perishable luxuries such as lacquer and silk in Neolithic times, but well-preserved examples survive only from the latter half of the first millennium bc. Clay, jade, lacquer, silk, and the techniques devised to work them remained important in all later periods of Chinese history. The knowledge of clays and kiln construction developed by Neolithic potters moreover laid a foundation for the metal-casting industries of the Bronze Age. Chinese art was already distinctive in Neolithic times (Bagley 2013). The painted pot illustrated in Figure 20.1 belongs to the Majiayao culture, a society of village farmers that flourished in northwest China for a few centuries before and after 3000 bc. Shards of painted jars, bowls, and bottles have been found in and around houses, but complete pots like this one were most likely unearthed from cemeteries, although no cemetery site has yet been scientifically excavated. The Majiayao people used painted pottery both in daily life and in the afterlife. Since pots in daily use sooner or later break, making pottery was probably a regular task in every household, though it is possible that all the households in a village shared a single 425

Wang Haicheng

Figure 20.1 A. Earthenware bottle painted in black, from Gansu Lanzhou Xinghetai. Height 25.4 cm. Majiayao culture, a few centuries before and after 3000 bc. B.View of bottle from the top. A. Courtesy of Guo Sike; B. photograph by Kyle Steinke.

kiln. Ethnographic studies of more recent village societies make it plausible to suppose that the potters were women. To make the bottle in Figure 20.1 the potter first built a disk-shaped base. On this she coiled noodles of loess-based clay of fine particle size. If she were making a cooking vessel that needed to withstand repeated heating, she would add temper to the clay, but the bottle required little temper. Loess is dust blown from the Mongolian deserts and deposited across north China. In Gansu, where the bottle was made, the deposit can be up to 300 meters deep. Mixed with water, loess has low plasticity and is well suited to the coiling method. The potter had to work the clay coils tightly together, since the pot would break during firing if the clay contained air bubbles. To achieve a thin wall the bottle was probably trimmed and shaved on a slow turntable, rough areas being smoothed down in the process. 426

Art

Figure 20.1 (Continued)

After the shaping was done the bottle would be left to dry and lose its plasticity. The potter then burnished the surface, except for the handles and the rim of the mouth, with polishing stones. She now had a three-dimensional blank surface on which to paint. Her paint was a watery solution of the same fine clay she used for potting, to which she added a bit of powdered iron ore or manganese ore that would turn dark brown or black on firing. The brush she used was probably very simple, not remotely as sophisticated in construction as the brushes of later calligraphers, but in an experienced hand it gave elegant results. In ink writing on paper and silk, erasure is not possible, and the same was true of the Neolithic potter’s art. Because the paint and the pot were the same clay, they bonded indelibly. The painter had to visualize her design as a whole before she began to paint, and she had to execute it without error. She must have concentrated intently and painted with slow deliberation, yet the tiny flicks of her brush everywhere in 427

Wang Haicheng

the design (e.g. just below the big circle’s southeast corner) speak of dash and buoyancy, giving dynamic movement to the lines whirling about the big circle. The painter used overshoots just as a calligrapher does, to give the lines individuality and expose process.This is the workmanship of risk: the quality of the finished object depends critically on the artist’s care, judgment, and skill (Pye 1978: 4). No doubt the painter exerted herself partly out of pride – the other women in her village, potters themselves, were an expert audience, and could recognize her hand – but we must not forget also the strictly personal satisfaction that comes from the exercise of mastery. Majiayao pottery is a good example of the art of a relatively egalitarian society, one in which many people enjoyed the free exercise of their creativity. In the complex, increasingly stratified societies that were forming on the east coast around the same time things were very different. Pottery-making became a specialized craft, and the exploration of artistic talent became the province of a few specialists. One notable sign of the new division of labor is the invention of the fast wheel around 3000 bc in the Dawenkou culture of the Shandong peninsula. This technology originated in the potter’s effort to purify and control the color of the clay used for vessels serving food and drink in ceremonies such as funerals. The high plasticity of finegrained clay with uniform particle size produced by repeated levigation and kneading allowed converting the slow turntable into the fast wheel, an innovation that opened up a world of new possibilities. The ensuing exploration of the new material and technique was dramatic. It shifted the potter’s focus from applied surface decoration to pot shape: shapes became complicated, and decoration almost disappeared. The development culminated in the Shandong Longshan culture, which succeeded Dawenkou around the middle of the third millennium. Its famous black wares with eggshell-thin walls are represented here by the fantastically impractical goblet in Figure 20.2, an object whose silhouette of smooth curves and crisp angles is articulate in a way altogether foreign to the swelling volumes of Majiayao pots. The goblet has a tectonic quality that comes from joining three parts thrown separately – cup, upper stem and midsection, and lower stem and ring foot.The rim of the cup is less than a millimeter thick and its wall not much thicker, giving it an almost metallic look. Electron microscopy of similar sherds has detected horizontal grooves spaced at 70–80 micrometer intervals on the interior wall (Figure 20.2), prompting some researchers to suggest that cups like this were shaved after drying by a notched scraper turning vertically on a lathe (Vandiver et al. 2005: 67). Slender pots and pots that change diameter rapidly do not invite painting. Longshan potters liked pure white and black. The jet black of the goblet in Figure 20.2 comes from firing in a kiln starved of oxygen and, late in the firing, producing black smoke in the kiln to carbonize the ware. The potter burnished the cup but not the stem to a fine gloss before firing, creating a contrast between shiny and matte areas. He also rhythmically perforated the midsection with minute raindrops. These visual refinements must have been sought for their effect in communal ceremonies. The finest Dawenkou and Longshan pots are unearthed not at habitation sites but from tombs, in large numbers. Drinking vessels are often found in piles on an earthen shelf around the wooden burial coffin or chamber, as though a ritual of drinking or libation that involved dozens of people was performed before the burial was sealed. The fragility of eggshell wares may mean that they were made specifically for this one-time use. Discarding the work of specialist potters in large quantity amounted to conspicuous consumption of wealth, a universal means for the rich and powerful to bolster their social status in a competitive society, but fine pots may also have been used in communal feasting. At Rizhao Liangchengzhen, a Longshan settlement on the Shandong coast with three concentric moats enclosing an area of more than 100 hectares, archaeologists found a pit containing nearly two hundred complete vessels, many

428

Figure 20.2 A. Black earthenware goblet, from Shandong Rizhao Donghaiyu. Height 22.6 cm. Longshan culture, late third millennium bc. B. This electron microscope image shows the interior of a shard from another eggshell pot. A. Courtesy of Guo Sike; B. after Vandiver et al. 2005, pl. 6.6, by permission of Fang Hui.

Wang Haicheng

with residues of a kind of rice wine (other ingredients include honey and possibly wild grape), among them an eggshell goblet. The excavators suggest that these vessels were discarded after a banquet (McGovern et al. 2005: 82). Other contemporary societies focused their artistic energies on jade, a more precious and less tractable material than clay. Human fascination with shaping stones started when our ancestors first began making tools by such techniques as knapping and pressure flaking. Shaping by grinding and polishing came later. Today we use the word “Neolithic” for the period that began when hunting and gathering gave way to farming, but the nineteenth century prehistorians who coined the word were looking at polished stone tools.The grinding and polishing of stone could have begun in several ways. Querns and handstones for grinding nuts and cereals need smooth surfaces. They may however have been preceded by the making of stone beads to ornament the body. A polished surface brings out the color of an attractive stone and makes it smooth to the touch. Some hard stones – jade is a prime example – can only be worked by polishing, not by chipping or flaking. In China several early societies showed a decisive preference for ornaments of nephrite (“true jade”) and related hard stones with similar attractions: a range of colors (owed to impurities), the ability to take a lustrous polish, toughness (that is, resistance to breaking), translucency, and beautiful patterning when sliced. Archaeologists have not so far been able to match Neolithic jade artifacts to sources of raw material, but the sources are likely to have been in mountains or in the beds of rivers that erode mountain formations. Jade pebbles found in riverbeds often have a rind that must be cut through to assess the quality of the material. Given a pebble of fine material, the jade worker’s task is to exploit its virtues – wasting as little as possible – with a very limited repertory of techniques: drilling, sawing, and grinding, all done with abrasives such as quartz sand, which is not much harder than nephrite. The lapidary could either saw the pebble into slabs, producing a flat object like the plaque in Figure 20.3, or trim it into a rectangular solid and then modify that shape to produce, for example, the strange object in Figure 20.4. Leftover bits could always be used to make smaller objects like beads.

Figure 20.3 Jade from Shaanxi Fengxiang Shangguodian. Length 11.4 cm. Hongshan culture, second half of the fourth millennium bc, but found in a Qin tomb of the sixth century bc. Photograph courtesy of Yu Cailing.

430

Art

The jade in Figure 20.3 is from the Hongshan culture (3500–3000 bc) of northeast China. It is a rectangular openwork plaque with a pair of creepy eyes, a line of barbed teeth below them, and two C-shaped elements bracketing the teeth that might be interpreted as jaws seen in profile.Two horizontal bars separate these elements from similar ones upside down on the upper edge of the plaque. The surface is not flat but shaped into ridges alternating with dished grooves and beveled edges.The fluidity of these gently curving surfaces represents a staggering investment of labor.The jade worker may have used a belt-driven drill, holding the drill fixed and moving the jade against it. Hongshan jades of this type have been found next to the head of the deceased or lying on top of the chest. Suspension holes like the one in the forehead in Figure 20.3 suggest that they were attached to clothing or headgear. Most Hongshan jades seem to have functioned as personal ornaments of some kind. The examples recovered by archaeologists all come from elite graves lined with stone slabs and located on hilltops. Jade ornaments must have had great importance in Hongshan society, for they are normally the only artifacts found in elite graves, anywhere from one to twenty in a grave. Ornaments remain a major category of jades in all later periods. Another important category is glorified versions of tools and weapons originally made in ordinary stone, baked clay, or shell. The Liangzhu culture (3300–2200 bc) of the Yangtze delta region owes something to the Hongshan jade industry, which is older, but it is much richer in jade, and the types include not only personal ornaments but also an abundance of fine axe blades. The mysterious object illustrated in Figure 20.4 is of unknown purpose and origin. Perhaps, like jade axes, it derives from a stone

Figure 20.4 A. Jade cong from Zhejiang Yuhang Fanshan. Height 8.9 cm. Liangzhu culture, first half of the third millennium bc. B. Line drawing of the figure repeated twice on each side of the cong. A. Photograph after Zhongguo wenwu jinghua, no. 13; B. drawing after Zhejiangsheng Wenwu Kaogu Yanjiusuo 2005, vol. 1, p. 56, fig. 38.

431

Wang Haicheng

Figure 20.4 (Continued)

weapon, a macehead, a token of authority in many societies. Conventionally called cong, it takes the form of a jade cylinder with decorated corners, and it is a hallmark of the Liangzhu culture. The circular top and bottom of the present example resemble in size and shape another jade type invented in the Liangzhu culture, also of unknown purpose, the disk with circular perforation. Unlike the cong, the disk is a standard jade form that has continued to be made throughout Chinese history. Cong are found in graves of elite males. They come in all sizes and proportions. The massive one illustrated here, 9 centimeters high and 6.5 kilograms in weight, was found next to the head of the deceased. Five smaller cong were found near the arms, and the grave contained another 641 jades of other types, not counting numerous beads and flakes. It is the richest of nine shaft burials in an elite cemetery near modern Hangzhou. The cemetery is on top of an artificial platform located in the center of an ancient city with an earthen wall that encloses 300 hectares. The city’s jade industry probably served mainly its own elite, but it may well have set fashions for the ruling class of other Liangzhu settlements too, establishing what kinds of assemblage and decoration were appropriate for each level of the elite (Qin Ling 2013: 593–594). The cong in Figure 20.4 was made by first sawing a rectangular block out of a pebble, then drilling a hole from top to bottom, then grinding parts of the block away to leave what looks like a cylinder with four corners projecting from it. The corners are not quite right angled, so a block that was originally square now has gently bulging sides: this subtle effect cost the jade

432

Art

worker much labor. The corners have paired-eye designs that resemble faces. These are standard on cong. Uniquely, however, the maker of the object in Figure 20.4A put a more complicated design on the sides, midway between corners. The design, repeated twice on each side, is shown in a line drawing in Figure 20.4B. A trapezoidal human face wears a feathered headdress. The man has a wide flat nose, a mouth lined with teeth, and circular eyes with horizontal prongs. His arms turn inward to touch a pair of spectacled eyes belonging to a beast. The beast seems to be resting its head on its clawed forelegs. It too has a wide nose; its mouth has four tusks, two up and two down. Parts of the design – the man’s head and headdress, the beast’s eyes, spectacles, nose, and mouth – are in high relief embellished with engraved lines. These raised parts were made by lowering the surface around them.The rest – the limbs of man and beast – are engraved on the lowered surface. The engraving was done with a sharp-pointed broken piece of abrasive stone. Circular eyes were drilled with a hollow drill, perhaps of bone. The profuse decoration of Liangzhu jades makes for a very different effect from the smooth and unctuous surfaces of Hongshan jades. The paired-eye designs on the corners of the cong in Figure 20.4A, a standard feature of the cong shape, are much simpler than the man-and-beast design on the sides. Remarkably, however, they seem to be abbreviations of the man-beast combination. On each corner, beginning at the top, we see a pair of small circular eyes with horizontal prongs, then big spectacled eyes, then the circular eyes again, then the spectacled eyes. This link with the full form of the design tempts us to interpret other features: the striated bands just above the circular eyes might for instance stand for the feathered headdress. The full form of the man-and-beast motif is found only on this cong and on a few other objects from the same burial and one other in the same cemetery.The two burials are commonly interpreted as belonging to a ruler of the city and his wife. At the time of this comparatively early cemetery, the abbreviated form was applied to many jade types, but as time passed it was gradually restricted to cong and further simplified, sometimes to little more than pairs of circles. Many interpretations have been offered for these faces, but we have no evidence by which to choose one over another. In the absence of written evidence from the Liangzhu culture, speculation about symbolic meanings seems fruitless. It is more useful to notice that, while the ideas behind the cong of Figure 20.4 and the Hongshan plaque of Figure 20.3 are likely to be entirely different, jade workers in both cultures are exploiting the attention-getting power of paired eyes. As the artists of many other times and places have realized, a motif that stares at the viewer causes the viewer to stare back. This power over the viewer’s attention can be exploited for whatever purpose the artist requires, from making a religious icon hypnotic to selling a product on a billboard (Bagley 2015: 102–119).

The Bronze Age: second and first millennia bc Long after the disappearance of the Hongshan and Liangzhu cultures, we encounter a new repertoire of paired-eye designs in the decoration of ritual bronzes, objects that by 1500 bc had become the leading art form in China’s Bronze Age societies. The second group of artifacts in our imaginary exhibition, Figures 20.5–20.12, consists mostly of vessels for making offerings of wine. The bronze industry also produced vessels for food offerings, however, and many other objects as well: musical bells, weapons (some for elite display), mirrors, fittings for architecture and for chariots, and more. Most of these are found chiefly in tombs but sometimes also in hoards or deposits of uncertain purpose. Some of the cultures that made and used them, the civilizations that flourished in the Yangzi region in the late second millennium bc for example,

433

Wang Haicheng

have left little indication of what they used them for. But in the cultures of the middle Yellow River region, the vessels held offerings made to deceased ancestors.Vessels filled with food and drink were placed in tombs as part of the ritual of burial; similar offerings were made above ground at intervals after the funeral; and, beginning in the Western Zhou period (ca. 1000–771 bc), bells played music during the offering ceremonies. As the apparatus of solemn rituals, the vessels had a visual function, like the liturgical vessels on the altar of a cathedral, and the finest ones are great works of art.They vary in quality of workmanship, however, presumably in accord with the wealth and power of their owners. Some of our information about functions comes from dedicatory inscriptions cast inside vessels or on the exteriors of bells. In the second millennium the inscriptions are brief, but by the Western Zhou period they can be very long, and they are often great works of scribal art (Steinke 2012). The ritual bronzes drove the development of the highest technology of their time, a moldmaking and casting technology that in the long term had consequences in realms as different as iron casting, printing, and factory organization. Their manufacture consumed vast resources. The quantity of metal that was buried in Bronze Age tombs rather than recycled is unparalleled in other ancient civilizations. The fifth-century tomb of Marquis Yi of Zeng, for example, contained 10 metric tons of bronze (Figure 20.10). The rituals of royal sacrifice were meant to keep the king on good terms with the most powerful spirits, and the central role played by vessels and bells in these rituals gave them associations with legitimation, turning them in the eyes of later commentators into symbols or even quasi-magical vehicles of dynastic legitimacy. This, together with inscriptions mentioning such paragons as Confucius’ hero the Duke of Zhou, brother of the first Zhou king, has made them prized throughout Chinese history. In the Han dynasty the unearthing of an ancient bronze was reported to the emperor and recorded as a good omen. The bronze vessel in Figure 20.5 is a wine bottle of a type conventionally called you. This comparatively early example, made probably a little before 1300 bc, is pear-shaped and circular in cross-section, with a lid, a ring foot, and a swing handle that imitates rope. The body, lid, handle, and a link that connects lid and handle all move freely but are permanently interlocked. In effect they form a chain of four links, and the links were all made by casting. Four molds and four casting operations were required, each newly-cast part being embedded in the mold for the next part. There are easier ways to make a bottle with a swing handle. This at first glance simple object is a tour-de-force whose owner no doubt took pride in his caster’s skill. Ignoring the lid, link, and handle, we can take the body of the vessel to illustrate the essentials of early Bronze Age mold-making. The caster began by making a clay model of the object he wished to cast. On this he would form the outer parts of a clay mold, a mold divided vertically into pieces whose fit was ensured by mortises and tenons in their edges. In the case of a simple vessel of circular cross-section like the present one there would be three identical outer pieces. The mold needed also two inner pieces, solid clay cores corresponding to the hollow interior of the vessel and the interior of the ring foot. The caster fitted the outer pieces together around the cores, inserting small chips of metal between the two cores and between them and the outer pieces to guarantee a little space between them, and then poured metal into that space. After the bronze solidified the outer pieces of the mold were broken off and the cores were dug out. The thickness of the vessel wall was the thickness of the metal chips, which remained part of it. Successful casting depended on mastery of clays, great precision in mold-building, and careful control of pouring temperature (judged by color and viscosity of the metal in the crucible).

434

Art

Figure 20.5 A. Bronze you from Hubei Huangpi Panlongcheng. Height 31 cm. Erligang culture, fourteenth century bc. B.View of the linked lid of the you from the top. Photos by Kyle Steinke.

The very first cast bronze vessels, two or three centuries older than the one in Figure 20.5, had no decoration. But decoration was soon added, and the way the casters chose to make it had lasting consequences for the decoration’s structure. They chose at first to carve all their decoration in the mold pieces and to invent pattern units that would be self-contained and complete within the boundaries of each piece. The result, on a bronze cast in a three-piece mold, was a register of decoration in which the same pattern unit repeated three times. The layout of the patterns, in other words, reflected the structure of the mold. Subdivided layouts of this kind were kept even after the casters switched to carving decoration on the model and then transferring it to the mold pieces when they were formed on the model. (The switch

435

Wang Haicheng

Figure 20.5 (Continued)

made it easier to execute complex designs and, later, high relief.) In Figure 20.5 we see both options in use: the borders of little circles and the thread relief on the shoulder and neck were executed directly in the mold pieces, but only after the main register had been carved on the model and transferred.Yet the main register is still formed of three matching units corresponding to the arcs of three mold pieces, even though it might seem more natural for a decorator working on the model to fill the circumference with a continuous pattern. From the time of the you in Figure 20.5 down to about 500 bc, the bronze decoration was with rare exceptions carved entirely on the model. But for most of that time caster and patron remained wedded to an aesthetic of subdivision. Their taste had been formed earlier, at the stage when the mold pieces were carved one by one. The decoration of any vessel shape had a distinctive fit supplied by the way its mold was divided. In Figure 20.5 the pattern unit in the main register consists of a pair of staring eyes surrounded by swirling lines in which we can detect, or imagine, a few other anatomical features (nose, jaws, horns. . .).This unit centered on a pair of eyes is by convention called a taotie.Though

436

Art

they may prompt us to imagine a face, the earliest versions are scarcely more than eyes. A little later the casters invented another unit, a shorter one containing a single eye amid swirls and hooks. The one-eyed unit invites us to imagine a creature seen from the side. Both types of unit remained more pattern than animal during the Erligang period (1500–1300 bc), but around the beginning of the Anyang period (1300–1000 bc) the animals they vaguely evoke were given unambiguous shapes and turned into vivid animal images. Thus around 1300 bc, not long after the you of Figure 20.5 was cast, the two-eyed unit was transformed into a tangible but peculiar creature, a face seen from the front with, when space allowed, a body in profile view on each side of it. At the same moment the one-eyed unit gave birth to a wide variety of creatures that we call, by convention, dragons. The altar set in Figure 20.6 shows what the animals had become by the end of the Anyang period. On the house-shaped vessel (fangyi) at the left we see a taotie in the main register, dragons in the narrow registers above and below, and on the lid an upside-down taotie. Because of the spaces to be filled, rectangles almost as tall as they are wide, the taotie are faces without bodies. In Figure 20.6 the fangyi stands on a rectangular bronze altar along with a ladle and two more vessels, one of them with a small rectangular stand of its own. The latter two vessels, as it happens, belong to the same wine-bottle type as the you in Figure 20.5, and comparing them with

Figure 20.6 Bronze altar set from Shaanxi Baoji Shigushan. Length of altar 94.5 cm, heights of fangyi 63.7 cm, larger you 50 cm, smaller you 36 cm. End of the Anyang period, ca. 1000 bc. Photograph courtesy of Zhou Ya.

437

Wang Haicheng

their predecessor gives us some feeling for the changes of design the bronze vessels underwent in the three centuries of the Anyang period. A first difference, not obvious in the illustrations, is size. The you in Figure 20.5 is 17 cm in diameter and 31 cm high. The one at the center of Figure 20.6 is 31 cm wide and 50 cm high. The early vessel, which in its time was a formidably sophisticated object, looks simple and innocent next to its massive descendant. In the later version the ring foot has been given a high molding, and the lid is so high that its vertical part has its own register of decoration. The body in Figure 20.5 is circular in cross-section; both of the you in Figure 20.6 are elliptical in cross-section. This flattening of the body locks us into the frontal view, the broad view shown in Figure 20.6, and the vessels have been designed for impact in this view. Spurred flanges that jut from the sides of the body and enormous upturned hooks on the lid (on the big you one hook is broken) make their effect from the front. The swing handle is oriented to cross the body over the short axis of the ellipse, allowing it to clear the lid hooks and preventing it from confusing the silhouette by competing with the hooks and the flanges. The handle terminals are animal heads with pronged antlers, again designed for the frontal view. In the main register we see not a taotie but a strip of vertical ribbing and beneath it a pair of large birds confronted across a flange. On these elliptical vessels the design repeats twice, on front and back, not three times as in Figure 20.5. The three vessels in Figure 20.6 clearly were designed not just to hold wine but to present it with the utmost splendor: their function was to make the rituals in some aristocrat’s ancestral temple awe-inspiring. Fragments of molds for similar vessels have been found at an Anyang foundry site (Li Yung-ti 2003: 259–261), arguing that the set in Figure 20.6 was cast at Anyang, but it was found in a rich tomb at Baoji in western Shaanxi, deep in the territory of the Western Zhou state that succeeded Anyang around 1000 bc. Perhaps it was plunder carried off by some early Western Zhou military man. Despite all differences of appearance, the late Anyang vessels in Figure 20.6 were cast in essentially the same way as the Erligang one in Figure 20.5, exploiting the same basic technology. That technology had been developed mainly at the Erligang capital, at Zhengzhou in Henan province, but well before the end of the Erligang period around 1300 bc it had spread not only north to Anyang but also south to the Yangtze River valley, as explained in Bagley’s chapter in this volume (Chapter 3). The famous southern castings mentioned in that chapter, a bronze tree from Sichuan and a zun with four rams from Hunan, were made by skillful but straightforward application of Erligang techniques. The technology seems to have been spread primarily by a military expansion of the Erligang state. Though short-lived, the Erligang empire triggered the birth of distinctive local bronze industries that in the middle and lower Yangzi region are characterized above all by the use of tuned sets of large bells. When around 1000 bc the Anyang state in the middle Yellow River valley was overthrown by the Zhou, the Zhou adopted Anyang rituals and ritual apparatus, as well as Anyang writing and bronze inscriptions, but to this northern inheritance they added bells and bell music acquired from the south. The mixing of Anyang, southern, and also northwestern artistic traditions in the early Zhou period gave birth to a few bronzes even more flamboyant than the Anyang castings in Figure  20.6. The quest for visual impact was pushed to the limit, with results that range from dramatic to willfully eccentric. In this heated atmosphere, one way for bronze designers to stand out from the crowd and impress their patrons was to change direction. In the hands of some casters the change was very abrupt, a reversal of taste that replaced crushingly architectonic vessels like those in Figure 20.6 with the simplest of gently rounded pottery shapes. The zun in Figure 20.7 is a less radical departure, made by a caster not yet quite willing to give up barbed flanges and compartmented decoration. It is exceptional in quality of design and execution, however, and in at least three

438

Figure 20.7 A. Bronze zun from Shanxi Yicheng Dahekou. Height 25.4 cm. Western Zhou, late tenth century bc. B. Close-up of zun. A. Photograph courtesy of Xie Yaoting; B. photograph by Kyle Steinke.

Wang Haicheng

Figure 20.7 (Continued)

respects very suggestive for the future. First, unlike the vessels in Figure 20.6, it was not designed with one fixed “frontal view” in mind. The zun does not fully reveal itself from one viewpoint but invites the viewer to turn it or move around it. Second, while the animals in Figure 20.6 are set out distinctly in rectangular compartments, neatly framed by vertical flanges and horizontal blank strips, the birds on the zun escape their frames. The confronted birds in the main register meet at the flange between them to share a massive beak; their legs cross down into the register on the vessel’s ring foot. Both horizontal and vertical boundaries are infringed. And third,

440

Art

the impression we have in Figure 20.6 of distinct, legible, neatly organized birds, dragons, and taotie is beginning to blur in Figure 20.7. In all three registers the birds’ anatomies have been freely reshaped into swirling curves to make them fill the whole of a compartment. In the main register the tail plumes are almost unrecognizable because they have been turned upward to repeat the curves of the wing. In the upper register, pairs of birds have been shaped to fill ogival compartments. The structured system of decoration inspired by piece-mold casting, a system that originated two centuries before the you of Figure 20.5, is here in the late tenth century bc showing signs of breaking down. This caster has taken a first small step in a direction that his successors found very inviting. What followed over the next few centuries was the almost complete elimination of compartments and imaginary animals. Static, focused schemes of decoration gave way to designs that flow. By 800 bc birds and taotie had dissolved into or been replaced by continuous curves like the wave patterns on the lid and neck of the hu in Figure 20.8A. The only animal motif to survive was the dragon, probably because it needed so little to keep it alive: any ribbon with a head is a dragon. In Figure 20.8B the main register gives an even stronger sense of fluid movement than the ones above it, and it shows a further innovation very important for the future, interlace patterns with an implied third dimension. The bodies of dragons seem to pass over and under one another. Nothing like this had existed before Western Zhou. With the collapse of the Zhou empire in the eighth century the Western Zhou period ends and Eastern Zhou begins. The Eastern Zhou decorator’s starting point was a three-part inheritance: horizontally continuous patterns that ignore mold divisions, the idea of interlace, and the dragon motif. These ingredients were combined and varied endlessly over the next half millennium, sometimes yielding patterns full of movement, sometimes executed on so small a scale as to be more texture than pattern. Most significantly, perhaps, around the seventh century they gave birth to a technique that mass-produced decoration, a technique in wide use by the early fifth century (and by the fourth century, in very tricky forms, the chief method for decorating the backs of bronze mirrors).The new technique is best known from a foundry in Shanxi province at Houma, capital of the Jin state, though it was not necessarily invented there. The hu in Figure 20.9, from about 482 bc, is an outstandingly fine Houma product. Its rich surface texture is supplied by seemingly continuous horizontal registers of dragon interlace (and a main register in which we notice a revived taotie motif). Unlike the decoration in Figures 20.5–20.8, these patterns were carved neither into the mold pieces nor onto the caster’s model. They were instead made by replication from baked clay pattern blocks somewhat like the one shown in Figure 20.9B.The Houma caster first made a plain model of the hu and on it formed a plain two-piece mold. Into the blank surfaces of the two mold pieces he then implanted soft clay slabs of decoration, clay negatives taken from pattern blocks like the one in Figure 20.9B. Having chosen a pattern block for each register from the foundry’s collection, he made a register by packing soft clay into the block, peeling out the resulting negative, bending and trimming it as necessary to fit the curvature and dimensions of the blank mold, fixing it in place, and then repeating the process until the circumference was filled (Bagley 2015: 142–155). Once he had done this for all registers, the interior of the mold had been in effect wallpapered with negatives. Since the circumference of a register seldom happened to be an integer multiple of the pattern block’s width, the last negative set into it usually had to be cut short, like wallpaper. The usefulness of the technique thus depended on a taste for patterns so complicated and fine-textured that the missing bits would not be noticed. The misfits are in fact hard to spot, and ancient patrons may never have known why their foundries were so efficient.

441

Figure 20.8 A. Bronze hu from Shaanxi Meixian Yangjiacun. Height 58.8 cm. Western Zhou, late ninthearly eighth century bc. B. Close-up of main register of hu. A. Photograph by Wu Zhenlong; B. photograph after Shaanxisheng Wenwuju and Zhonghua Shijitan Yishuguan 2003: 25, by permission of Shan Yueying.

Art

Since the pattern blocks were baked hard after carving, they could be reused many times. Moreover, because the negatives taken from them could be bent and trimmed, a pattern block could be used to decorate a mold of any shape. The blocks used to decorate the hu of Figure 20.9A were not limited to decorating hu vessels.Their patterns could be applied to any bronze. The pattern-block technique amounts to mass production of decoration. The quality of the decoration was predetermined by the quality of the pattern block, putting a premium on the master carver’s design and execution. The pattern blocks unearthed at the Houma site can be dazzlingly sharp, almost surreal, the bodies and feathers of dragons richly embellished with striation and granulation or commas and volutes spiraling into high relief. Carving of this quality had been unknown for centuries. Perhaps the new high standard of execution was inspired by encounters with fine Anyang bronzes. Revival of the taotie and other long-disused motifs argues that the carvers had indeed seen Anyang and early Western Zhou bronzes, items found in an ancient grave, perhaps, or sitting for generations on the altar of a royal temple. In jade the recovery and transmission of ancient objects and thus of ancient designs is well established (Sun Qingwei 2012). Neolithic jades have been found in Anyang tombs alongside Anyang jades that imitate or adapt them, and Anyang jades are often found in Western Zhou tombs. The fourth-millennium Hongshan jade in Figure 20.3 was found in a sixth-century tomb in Shaanxi, far from Hongshan territory. Houma designers clearly drew on ancient sources for ideas, but they had more recent sources of inspiration as well, some from outside China. The taotie in Figure 20.9 are not motionless, isolated, and self-contained; they are entangled with winged dragons that they fight and devour. Such designs combine the idea of interlace with two foreign motifs brought to China from Western Asia by the nomads of the northern steppes, the animal-combat motif and the composite animal (e.g. tigers with wings, the inspiration for Chinese dragons with wings). Both motifs have a long history in the ancient Near East. A few Houma products that seem to have been made specifically for the nomad market attest a deep familiarity with foreign designs. But the quality of the finished bronze did not depend on the quality of the pattern block alone. It depended just as critically on the skills of the person who took soft negatives from the blocks and planted them in the mold. This amazingly delicate operation, which might set a hundred or more thin slabs into a single mold, with indetectable joins between them, had to produce a mold that could be baked without distortion or cracking and that could then withstand the pouring of molten metal. A bronze like the one in Figure 20.9 is a tour-de-force of both metal and ceramic technology, with centuries of accumulated knowledge and experience behind it. Though we know it best from the Houma foundry site, the only site at which actual pattern blocks have been found, the technology was used with at least equal virtuosity in the south, where the Hubei tomb of Marquis Yi of Zeng (d. 433 bc) has yielded several hundred outstanding pattern-block castings. It yielded also the astonishing zun and pan in Figure 20.10, a matched vase and basin that combine Houma-like replication with the lost-wax process. Wild appendages and frothy cakes of openwork cast from molds made by the lost-wax process have been soldered onto a pair of simple vessels whose comparatively tame surface decoration came from pattern blocks. The additions overwhelm the objects they were added to. An inscription shows that Marquis Yi inherited the zun-pan set from an earlier marquis, and a few bronzes from other finds show that the lost-wax process was in use at least a century before Yi’s time, but how early it became known in China and by what means – local invention or import from the northern steppe – is uncertain.The process was clearly very slow to win the affections of Chinese patrons. Of the ten tons of bronze objects in Yi’s tomb, only the zun-pan set he inherited was made with

443

Figure 20.9 A. Bronze hu cast at the Houma foundry in Shanxi. Height 44.8 cm. Eastern Zhou, ca. 482 bc. B. Clay pattern block, also from Houma. Width 32.8 cm. A. Freer Gallery of Art and Arthur M. Sackler Gallery, Smithsonian Institution,Washington, D.C.: Purchase – Charles Lang Freer Endowment, F1957.22; B. after The Chinese Exhibition, no. 132.

Art

Figure 20.9 (Continued)

its help. All the bronzes he commissioned for himself are decorated either by the pattern-block method or by inlaying them with metal (gold, silver, copper) or semiprecious stone (turquoise, malachite). From the sixth century onward, in fact, it was not lost-wax but inlay that increasingly attracted the interest of casters and patrons. The pattern-block method, despite all its technical novelty, produced cast patterns in relief, decoration of a kind that by 500 bc was a millennium old. Inlay patterns are patterns in color, dependent on color contrast. They enabled the metalworker for the first time to compete with the seductions of silk and lacquer, brilliantly colored materials invented in the Neolithic but surviving in well-preserved examples only from about Marquis Yi’s time. Borrowing of ideas and subjects across media became ever more common thereafter, and while inlaid bronzes may have been the costliest luxuries, silk and lacquer began to challenge the millennium-old dominance of bronze as the leading vehicle of artistic invention and elite display. Silk is made from the cocoons spun by domesticated silkworms fed on mulberry leaves. Its production involves not only on rearing the delicate worms but also on cultivating mulberry trees, for the worms eat prodigious quantities of leaves before spinning their cocoons. Filaments reeled (unwound) from cocoons can be as much as a kilometer long. The filaments can be wound or twisted to make thread of whatever gauge is needed for use as warp or weft. Their tensile strength is high enough for silk strings to be used in musical instruments. The length and strength of silk thread encouraged warp-faced patterns, in other words patterns made by accentuating the warp threads, either by packing them densely or by making them thicker than the weft thread. The pattern repeat drawn in Figure 20.11, from a shroud covering the body of a woman of modest wealth buried about 300 bc in a cemetery of the Chu state, is warp-faced. 445

Wang Haicheng

Figure 20.10 A. Bronze zun and pan set from the tomb of Marquis Yi of Zeng at Hubei Suizhou. Heights of zun 30.1 cm, pan 23.5 cm. Eastern Zhou, a generation or two before Yi died on 433 bc. B. Close-up of openwork in the zun created by the lost-wax process. A. Photograph after Zhongguo meishu 1979/1, p. 64; B. photograph after Hubeisheng Bowuguan 1989, vol. 2, pl. 70.1.

The repeat consists of a zigzag running among eight symmetrical pairs of dancers and fantastic animals. In the warp direction it is 5.5 cm high and in the weft direction it extends the full width of the piece of cloth, 49.1 cm. The pattern was woven from yellow, red, and brown warp threads and a brown weft thread. Some textile specialists have suggested that for a complex pattern of this size the weaver needed a drawloom with devices for lifting groups of warp threads preselected by a harness tower set up according to the pattern (see Kuhn 1995: 90n32).Whether or not this is correct, the loom was certainly a complex one, and setting it up was a laborintensive process. In Figure 20.11, the zigzag’s first complete rectangle from left has a defect that is repeated in each unit throughout the length of the fabric, indicating that the pattern-setting of the loom could not be corrected during weaving. 446

Art

Figure 20.10 (Continued)

Woven patterns favor straight lines and diagonal axes, so figural elements acquire angular, stepped silhouettes. Chain-stitching by contrast makes curvilinear patterns easily, and because it is effectively drawing in thread, the pattern does not have to repeat at all. On embroideries from the Chu woman’s tomb, underdrawing in black or red ink can still be seen.Yet even in embroidery, diagonal elements (birds’ wings, long dragons’ beaks) are sometimes prominent, suggesting influence from the aesthetic of woven silk (Mackenzie 2001: 341). Interestingly, similar diagonals appear even in lacquer, for instance on the toilet box in Figure 20.12, though lacquer painting is a medium in which fluid curvilinear designs are easy and natural. Lacquer is a varnish made from the sap of the lac tree. Resistant to water, heat, and acids, it has been used to protect and beautify wooden objects since Neolithic times. It can be colored with a wide variety of pigments, but cinnabar for red and carbon for black are the most common because of their stability.To make the box in Figure 20.12A, from about 316 bc, the lacquer worker gave a piece of animal skin the shape desired for the finished box by pressing it, fur side 447

Wang Haicheng

Figure 20.11 Drawing of a pattern repeat on a piece of polychrome weave from Hubei Jiangling Mashan. Eastern Zhou, ca. 300 bc. After Hubei sheng Jingzhou Diqu Bowuguan 1985: 44, fig. 36.3.

outward, into a form. He put a piece of ramie fabric over the skin, applied a coat of varnish, and allowed the varnish to harden (cure in a damp atmosphere). Then he applied another coat of varnish and pressed another piece of skin into the form, this time with the fur side facing inward. When the second coat of varnish had hardened, he removed the three-layered box from the form and gave it a final trimming. It would be the core of the finished box (Wang Hongxing 1991: 497). A coat of red lacquer was applied to it and allowed to harden, then a coat of black lacquer. When the black coat had hardened, the designs we see were painted on it with lacquer in five colors. The lid of the box is painted in two circular zones on the top and one horizontal register on the side. The bottom is painted in a horizontal register on the side and one circular zone on the underside.With the exception of the register on the side of the lid (Figure 20.12B–20.12C), all the designs are elegant ornaments whose inspirations may well include inlay (in the use of silver and gold pigments), woven silk (in the prominent diagonals and bilateral symmetry), and dragons and birds dissolved into swirling curves. It is tempting to think of all these ornaments as borders to the register on the lid, which is of an entirely different character, a picture painted on a black background. The composition fills the circumference and is divided by trees into five sections or space cells. It was evidently meant to be read from right to left, just as a contemporary bamboo book would have been read (e.g. Figure 20.15), but inconsistent details, owed perhaps to a painter in a hurry, make interpretation speculative, and the space cells cannot quite be equated with successive episodes in a narrative. Reading should perhaps begin with the shortest section, framed by two windswept trees. Between the two trees a hound chases a boar, signaling that the setting is in the wild and also nudging us leftward. In the next section we see a sedate chariot, an armed attendant walking behind it, and two wild geese flying overhead, all again moving left. The chariot is drawn by three horses, an orange one flanked by two brown ones. Only the legs of the nearest horse are depicted, but their overlapping bodies give the scene a little depth. Three figures ride in the chariot: a driver, an official, and between them a retainer seen from behind. The retainer occludes the other two and is a little closer to us. The third section begins with a chariot with the same three passengers, but the walking attendant is missing, along with, unaccountably, one of the three horses. The horses have picked up speed, however, for the tassels behind their heads are flying. Still in the same space cell, we come next to three running attendants and then to what might be the same chariot again: the missing horse and walking attendant have reappeared, but the passengers in the chariot seem different. Though moving fast, the chariot is approaching a kneeling figure just in front of the tree that closes the space cell, perhaps someone who has come to greet the official. The next section contains five human figures, perhaps a welcoming party. Two officials march to the right while three attendants stand in rearview with their backs to us. In the final section, we come first upon the chariot, which has turned around to face the direction from which it came. The horses, one of them again missing, are stopped in front of a dog. The driver is now alone, for his passengers 448

Art

Figure 20.12 A. Lacquer box from Hubei Jingzhou Baoshan. Circumference 87 cm. Eastern Zhou, ca. 316 bc. B. and C. Detail of the register on the rim of the box’s lid. Photographs courtesy of Zhang Changping.

have dismounted to meet the greeting party: their meeting fills the remainder of the section. Then we are back to the shortest section, the hound chasing a boar. Perhaps this is not the first but the last section painted, an irrelevant bit contrived by the painter to fill a gap that remained when he finished telling the story. However we interpret the story or subject matter (travel, hunt, diplomatic meeting?), the difference between this register and the ornamental ones above and below it is unmistakable. This is representational art, art that concerns itself with real-world appearances and activities. In the Neolithic and most of the Bronze Age, art was dominated by ornament – nothing in 449

Wang Haicheng

Figures 20.1–20.11 is much concerned with images or actions of real animals or people – but by the fifth century, perhaps with some encouragement from steppe art, this began to change (Bagley 2015: 156–173). The mostly imaginary animals of the ornamental tradition were given organic vitality and a plausible space to move in. Human figures became more common in both two- and three-dimensional renderings, as the terracotta army made for the first Qin emperor’s mausoleum complex abundantly demonstrates, and though the human body for its own sake never acquired the importance in art that it had in the ancient West, human activities became a central concern, largely though not exclusively the activities of the elite. We might take chariots as symptomatic of these changes. From the time of its arrival in China at the beginning of the Anyang period, the chariot was always a vehicle of high status, and real chariots were long standard burial items in the tombs of high-ranking nobles (Wang Haicheng 2002; Wu Hsiao-yun 2009). But they would in time be supplemented by pictures, as the late fourth-century lacquer box from Baoshan shows, and finally replaced by them.

The early imperial period: Qin (221–206 bc) and Han (206 bc–ad 220) The Qin emperor’s terracotta warriors and the pictures on the lacquer box of Figure 20.12 both qualify as representational art, but while three-dimensional images were made from the Neolithic onward, two-dimensional pictures like those on the box barely existed before the fifth century bc. Representing the three-dimensional world in two dimensions posed technical and conceptual challenges to the painters who first attempted it, and the painter of the box already knows some basic tricks. He uses color contrasts to distinguish overlapping figures, bright colors to mark important figures, and natural objects of easily recognizable shapes – trees, birds – to establish outdoor settings. He uses the same trees as space dividers so that he can put side by side different episodes in which the same figures appear, but not quite consistently. Overlapping (horses, three figures in a chariot) and placement lower or higher on the picture plane (horses, three runners) imply depth, but it is hard to know whether the painter intended them to; he may only have been trying to prevent confusion by keeping closely grouped figures distinct from one another. At present the earliest paintings in which it is clear that the third dimension was consciously constructed are from tombs at Mawangdui that date from the first few decades of the second century bc. Everything depicted on the present lacquer box, including the trees that imply landscape, is silhouetted on a neutral black ground, not set in a continuous terrain or even on a ground line. The idea that a picture could be more than an assembly of discrete, isolated objects, that it could supply a continuous setting for animal and human actors to inhabit, seems to be a Western Han discovery. The inlaid bronze fitting in Figure 20.13 is a revealing example. Fittings of this kind are often found in pairs in high-ranking tombs. They served to reinforce the two joints of a chariot’s canopy shaft. The example in Figure 20.13, divided into four registers by raised bands, is set with semiprecious stones. The stones are arranged in regular columns of circles and diamonds, but the space around them is densely packed with birds, animals, humans, celestial creatures, and billowing lines that we interpret as forested crags partly because of hints of plant life but mainly because of the creatures that climb, fly, fight, and frolic on and around them. Creatures and terrain are executed in gold and silver inlay with a jeweler’s precision. Remarkably, the lines we interpret as landforms because of the animals that inhabit them are actually adapted from dragon bodies, the same dragon bodies that lie behind the abstract ornaments on the fourth-century lacquer box of Figure 20.12. On a few other chariot fittings the bodies still have heads. Landscape pictures are such familiar objects to us that it is hard to appreciate the conceptual difficulty faced by the painter who has never seen one. A man and 450

Art

a tree are easy to draw, but how do you draw the land that connects them? The Western Han painter’s answer was, you don’t; you borrow suitable shapes that you already know, like dragon bodies, adjust them a little, and make them into terrain by putting men and trees on top of them. When the subject – land, sea, fire – had no fixed, well-defined form, easy to draw and easy to recognize, the painters who first attempted them took familiar ornamental patterns and turned them into land or water or flames partly by reshaping but mainly by adding context and cues (Loehr 1980: 10, 26). At present eight such pictorial chariot fittings are known, four excavated from Han princely tombs of the first century bc (seven listed in Wang Haicheng 2005: 346n6; Henansheng Wenwu Kaogu Yanjiusuo 2010: pls. 33–35).Their consistency in subject matter and composition suggests

Figure 20.13 A. Bronze chariot canopy shaft fitting with inlays of gold, silver, and semiprecious stones from Hebei Dingxian Sanpanshan. Height 26.5 cm. Western Han, first century bc. B. Detail of inlay. After Zhonghua Renmin Gongheguo chutu wenwu xuan 1976, no. 66.

451

Figure 20.13 (Continued)

Art

that they derive from a single model, a model that was itself probably based on palace or temple mural paintings or painted silk hangings, such as are mentioned in Han texts. Conspicuous among the animals in Figure 20.13 are elephant, horse, camel, and a peacock-like bird, all of which the Han court knew as tribute from barbarians in the north, west, and south. Depictions of these exotic animals, and of riders who do not look Chinese, might have evoked the submission of barbarians to the Han court, a suitable theme for royal display (Miao Zhe 2013: 55n166). But evidence so scanty and indirect does little to help us visualize the paintings on the walls of Western Han palaces. The evidence for the palaces of Eastern Han, the two centuries that followed the Wang Mang interregnum (9–23 ad), is a little better, though still indirect. Stone reliefs from funerary shrines concentrated in Shandong province were very likely modeled on pictorial programs carried out for the first two emperors, Guangwudi and Mingdi, in the royal palace and funerary temples at Luoyang. The transmission of this art to Shandong and down the social ladder has been convincingly reconstructed in two recent articles (Miao Zhe 2009, 2013; see also Miao Zhe forthcoming). Briefly, two sons of Guangwudi enfeoffed in Shandong, Liu Qiang 劉彊 and Liu Cang 劉蒼, were granted royal prerogatives on two separate occasions by the first and second Han emperors. Liu Qiang’s residence, the famous Lingguang Palace of Lu (sometimes translated as the Hall of Numinous Brilliance), was painted following pattern books that had been used for the royal palace at Luoyang, and both his funerary temple and his brother’s were modeled on the temples of the first two emperors. Later, for use in decorating their own funerary shrines, Shandong officials sent their painters to copy the pictures on the walls of the palace and the two temples; and as time passed, their shrines in turn were copied and adapted for use still lower down the social scale. The finest surviving reflections of Han royal art are shrines of officials, one at Xiaotangshan, about 60 km from Liu Cang’s mausoleum, and the Wu family shrines, about 50 km from Liu Qiang’s palace and mausoleum. All are simple rectangular halls with a gabled roof. Table 20.1 shows the layout of subjects in the Xiaotangshan shrine. As the table indicates, a procession of chariots runs around three walls of the shrine. Part of it can be seen near the top of Figure 20.14, a rubbing of the west wall. On the back wall the central carriage has a caption saying that it is the carriage for “the great king,” obviously not an appropriate theme for a local official. Some of the other regalia shown were probably not appropriate even for an enfeoffed king unless he had been granted them by the emperor. These elements are not isolated borrowings, they are details of an imperial original that was adopted wholesale, details of an iconography of power that had been created in the service of the Eastern Han emperors. After Wang Mang, imperial rule had been reestablished under a revised interpretation of the Mandate of Heaven doctrine (Hou Xudong 2015). Imperial art was then deployed to define the empire in three ways, by relating it to Heaven, to history, and to the four quarters under Heaven (Miao Zhe 2014: 293). (1) Heaven’s bestowal of legitimacy was affirmed by declaring the Han court to be an image of the heavenly court, as it was oriented to the Dipper and the Pole Star. The Han emperor’s special relationship with Heaven was manifest in auspicious and inauspicious omens. (2) Because Heaven bestowed and withdrew its mandate according to the ruler’s conduct, history was represented (e.g. in the Wu family shrines) as a gallery of ancient kings and equated with the process of the Mandate of Heaven taking its course. To retain the mandate the emperor had to play the role assigned him in standard accounts of history: he had to hold audiences, perform royal processions, sacrifice to his ancestors, and venerate his teachers. (3) Heaven having given the empire the right to rule all under Heaven, the barbarians of the four quarters could be conquered or brought to submission. 453

Wang Haicheng Table 20.1  Distribution of pictorial themes in the Xiaotangshan shrine West Wall

Back Wall

celestial bodies /terrestrial beings chariot procession

chariot procession

attending officials

court audience

vanquishing the barbarians hunting

Confucius studying with Laozi Chariot procession

East Wall

Ceiling

celestial bodies /terrestrial beings chariot procession and the tribute bearers Duke of Zhou assisting King Cheng kitchen slaughter and hunting in/auspicious omens

celestial bodies/ terrestrial beings

Adapted from Miao Zhe 2013: 107, fig. 51.

Figure 20.14 Rubbing of the west wall of the Xiaotangshan shrine in Shandong Changqing. Height 180 cm, width 210 cm. Eastern Han, late first century ad. After Fu Xihua 1950, pl. 6

This analysis helps us to make sense of the Xiaotangshan shrine. As Table 20.2 shows, the themes it depicts form a coherent ensemble meant to symbolize the duty and power of the emperor. And as Figure 20.14 shows, they are presented with great clarity. Even the most chaotic scenes of war and the hunt are rendered in an uncluttered way. Though the building in which 454

Art

the oversized ruler sits to receive the victory report is commonly understood as a two-story pavilion, Han viewers may instead have seen it as two palatial buildings placed one in front of the other, the one in front for the emperor and the one in the back for the empress, symbolizing the balance of yin and yang powers (Miao Zhe 2013). Han artists knew how to use oblique angles and diminishing size to signal depth, but in a design like this one depth would be achieved at the cost of order. At the opposite end of the same register, barbarian archers emerge from mountains rendered as repeated overlapping curves, each with an archer silhouetted against it. The landscape is less dazzling than the one in Figure 20.13 but more legible. Perhaps because a dazzling luxury article was not his assignment, the artist who invented it took a step away from the security of age-old decorative forms toward the construction of an image of nature. Our small exhibition closes with the item in Figure 20.15, a piece of writing. Since its first appearance in the archaeological record of the late second millennium bc, writing in China has served many functions (Wang Haicheng 2014), but one constant throughout its history has been exploration of its artistic possibilities. Most of the writing that survives from the pre-imperial period is display inscriptions written on durable and costly materials, and beauty was of course a central concern in writing meant for display. Speed of execution was not. We have no examples of the cursive scripts that busy scribes employ until lucky burial conditions begin to preserve the perishable surfaces that were used for everyday writing. Everyday writing was done with brush and water-based ink on strips of bamboo tied together to form a scroll, as in Figure  20.15. The earliest bamboo documents yet found come from the fifth-century bce tomb of the Marquis of Zeng. A wealth of everyday writing done in the Han period survives in the form of administrative documents produced by low-ranking scribes stationed at the empire’s garrisons in the arid northwest. The inventory in Figure 20.15 is an example. Scribes developed cursive scripts for practical reasons; a literate Eastern Han elite saw and explored aesthetic possibilities. We read in the Hou-Han shu that Liu Mu 劉睦, king of Beihai and a cousin of the second emperor Mingdi, so excelled at cursive that his handwriting was taken as a model by his contemporaries. Upon learning that Liu Mu was seriously ill, the emperor sent fast couriers to order him to write ten letters in cursive (Fan Ye 1965: vol. 2, 557). Only the promulgation of such master copies can account for the standardized orthography of the cursive we see in Han administrative documents. Brushes excavated from Qin and Han tombs had only a small reservoir for ink. A scribe writing with such a brush on a narrow bamboo strip could connect the strokes of a character and write it in one uninterrupted movement of the brush, but he might need to dip his brush again before writing the next character.Towards the end of Eastern Han the impulse to beautify cursive writing inspired leading practitioners of cursive to make a number of technical improvements: Zhang Zhi’s 張芝 long brush, Wei Dan’s Table 20.2  Pictorial themes in the Xiaotangshan shrine and their symbolisms, as analyzed by Miao Zhe Themes

Symbolism

celestial bodies/terrestrial beings in/auspicious omens chariot procession hunting – kitchen slaughter vanquishing the barbarians – tribute Duke of Zhou assisting King Cheng – Confucius studying with Laozi

mandate-granting Heaven and its court emperor’s relationships with Heaven emperor’s majesty emperor’s duty of ancestral veneration emperor’s power to subdue the barbarians emperor’s reverence to his ministers and teachers

Adapted from Miao Zhe 2013: 64

455

Wang Haicheng

Figure 20.15 An inventory from Gansu Juyan, ink on wooden slips tied with cords, each slip ca. 13.5 cm. Eastern Han, ad 93 Photograph courtesy of the Institute of History and Philology, Academia Sinica, 128.1.

韋誕 pitch black ink, and Zuo Bo’s 佐伯 glossy paper.1 Now a gifted writer like Zhang Zhi could not only link strokes within a character but could also link successive characters, writing whole phrases without interrupting the expressive movement of the brush. The kind of cursive writing we know from the fourth century on may well have come into being in the second century, though no actual example has yet been found (Hua Rende 1999; Bai 2003: 588). Cursive writing breaks away from the old idea that all characters, whether simple or complicated, have the same size. Mixing large and small characters, relying on fine connecting lines for balance, made writers think consciously about spacing, both vertical and horizontal, and it opened up new possibilities for large-scale rhythms. In an essay entitled Denouncing Cursive (Fei caoshu 非草書), the second-century poet Zhao Yi 趙壹 lamented the calculation required by this new way of writing, complaining that devotees were sweating blood over a script that defeated the whole purpose of speedy writing.2 We might almost wonder if the development of new script types was driven by enthusiasts whose refinement of existing scripts again and again made an existing cursive too cumbersome for everyday use. In any event, the prerequisites for the elevation of beautiful writing to its later status as the most admired art form in China were all in place by the end of the Eastern Han period: (1) Writing materials that could record the tiniest movements of the brush. The way was open for a literate elite to claim that handwriting expressed a distinguished man’s personality or character.These literati practiced their handwriting and created a critical discourse about fine writing. (2) A ruling class that was interested in handwriting, practiced it, and actively promoted it by collecting and circulating it. (3) A technology for the wide dissemination of admired models made possible by paper, a Han invention. Fine pieces of writing were traced onto stone steles and then carved into them, and paper and ink were then used to take rubbings from the steles. Conceptual forerunners of the technology include seals (which may go back to Anyang times), the pattern-block method described in connection with Figure 20.9, Han stack-casting of iron tools (a descendant of pattern-block casting), and Western Han printed textiles. Its offspring include printing, first from woodblocks that could easily mix writing and images. If art after Han seems a world away from art before the imperial period, fine writing, printing, pictorial art, and the Buddhist religion introduced to China in Eastern Han are among the central reasons. 456

Art

Notes 1 Zhao Qi, ‘Sanfu juelu’ apud Yu Shinan, Beitang shuchao, chapter 104, in Xuxiu siku quanshu, Shanghai: Shanghai Guji Chubanshe, 2002, vol. 1212, p. 482. 2 Zhao Yi, ‘Fei caoshu’ apud Zhang Yanyuan, Fashu yaolu, chapter 1, in Siku quanshu, Shanghai: Shanghai Guji Chubanshe, 1987, vol. 812, pp. 105–6.

Works cited Bagley, R.W. (2009) ‘Anyang mold-making and the decorated model’, Artibus Asiae 69, no. 1: 39–90. Bagley, R.W. (2013) ‘Was China an Egyptian colony?’, in E. Frood and A. McDonald (eds.) Decorum and Experience: Essays in Ancient Culture for John Baines, Oxford: Griffith Institute. Bagley, R.W. (2014) ‘Erligang bronzes and the discovery of the Erligang culture’, in K. Steinke (ed.) Art and Archaeology of the Erligang Civilization, Princeton: Princeton University Press. Bagley, R.W. (2015) Gombrich among the Egyptians and Other Essays in the History of Art, Seattle: Marquand Books. Bai, Qianshen (2003) ‘Research notes on the calligraphy of Wu dynasty bamboo and wood slips from Zoumalou, and some related issues’, in Wu Hung (ed.) Han-Tang zhijian de shijue wenhua yu wuzhi wenhua, Beijing: Wenwu Chubanshe. The Chinese Exhibition: A Pictorial Record of the Exhibition of Archaeological Finds of the People’s Republic of China, Kansas City, MO: The Nelson-Atkins Museum of Art, 1975. Fan Ye (1965) Hou Han shu, Beijing: Zhonghua Shuju. Fengxiangxian Bowuguan (2005) ‘Shaanxi Fengxiangxian Shangguodiancun chutu de chunqiu shiqi wenwu’, Kaogu yu wenwu (1): 3–6. Fu Xihua (1950) Handai huaxiang quanji chubian (Corpus des Pierres Sculptées Han), Beijing: Centre D’études Sinologiques. Guoli Gugong Bowuyuan (2013) Hehe zongzhou, Taipei: Guoli Gugong Bowuyuan. Hou Xudong (2015) ‘Zhulu huo Tianming: Hanren yanzhong de Qin wang Han xing’, Zhongguo shehui kexue (4): 177–203. Henansheng Wenwu Kaogu Yanjiusuo (2010) Yongcheng Huangtushan yu Zancheng Hanmu, Zhengzhou: Daxiang Chubanshe. Hua Rende (1999) ‘Cong chutu jiandu kan liang Han shufa’, in Zhonghua Shudao Xiehui (ed.) Chutu wenwu yu shufa xueshu yantaohui lunwenji, Taipei: Zhonghua Shudao Xiehui. Hubeisheng Bowuguan (1989) Zeng hou yi mu, Beijing: Wenwu Chubanshe. Hubeisheng Jingsha Tielu Kaogudui (1991) Baoshan Chumu, Beijing: Wenwu Chubanshe. Hubeisheng Jingzhou Diqu Bowuguan (1985) Jiangling Mashan yihao Chumu, Beijing: Wenwu Chubanshe. Kuhn, D. (1995) ‘Silk weaving in ancient China: from geometric figures to patterns of pictorial likeness’, Chinese Science 12: 77–114. Li,Yung-ti (2003) ‘The Anyang Bronze Foundries: Archaeological Remains, Casting Technology, and Production Organization’, unpublished PhD thesis, Harvard University. Loehr, M. (1980) The Great Painters of China, Oxford: Phaidon. Mackenzie, C. (2001) ‘The influence of textile designs on bronze, lacquer and ceramic decorative styles during the Warring States period’, in Chinese Bronzes, Selected Articles from “Orientations”, Hong Kong: Orientations Magazine. McGovern, P.E. et al. (2005) ‘Shandong Rizhao Liangchengzhen yizhi Longshan wenhua jiu yicun de huaxue fenxi’, Kaogu (3): 73–85. Miao Zhe (2009) ‘Shi ‘tuhua tiandi pinlei qunsheng zawu qiguai shanshen hailing’’, in Fan Jingzhong, Zheng Yan, and Kong Lingwei (eds) Kaogu yu yishushi de jiaohui: Zhongguo Meishu Xueyuan guoji xueshu yantaohui lunwenji, Hangzhou: Zhongguo Meishu Xueyuan Chubanshe. Miao Zhe (2013) ‘Chongfang louge’, Meishushi yanjiu jikan (33): 1–111. Miao Zhe (2014) ‘Review of Anthony J. Barbieri-Low’s Artisans in Early Imperial China and Lillian Lanying Tseng’s Picturing Heaven in Early China’, Zhejiang Daxue yishu yu kaogu yanjiu 1: 292–318. Miao Zhe (forthcoming) ‘Cong Lingguangdian dao Wuliangci’, Zhejiang Daxue yishu yu kaogu yanjiu 3, forthcoming. Pye, D. (1978) The Nature and Art of Workmanship, Cambridge: Cambridge University Press. Qin Ling (2013) ‘The Liangzhu Culture’, in Anne Underhill (ed) Wiley-Blackwell Companions to Anthropology: Companion to Chinese Archaeology, London: John Wiley & Sons.

457

Wang Haicheng Shanxisheng Kaogu Yanjiusuo et al. (2014) Youyou luming, Beijing: Kexue Chubanshe. Shaanxisheng Kaogu Yanjiuyuan et al. (2014) Zhouye luming, Shanghai: Shanghai Shuhua Chubanshe. Shaanxisheng Wenwuju and Zhonghua ShijitanYishuguan (2003) Shengshi jijin, Beijing: Beijing Chubanshe. Steinke, K. (2012) ‘Script change in Bronze Age China’, in S.D. Houston (ed.) The Shape of Script: How and Why Writing Systems Change, Santa Fe: School for Advanced Research Press. Sun Qingwei (2012) ‘Fuyu yu fenqi: Zhoudai muzang zhong qiandai yuqi de laiyuan yu liuchuan’, Gugong wenwu yuekan 354: 36–41. Vandiver, P.B. et al. (2005) ‘Shandong Rizhaoshi Liangchengzhen Longshan wenhua taoqi de chubu yanjiu’, Kaogu (8): 65–73. Wang Haicheng (2002) ‘Zhongguo mache de qiyuan’, Ouya xuekan 3: 1–75. Wang Haicheng (2005) ‘Chariot canopy shaft fitting (bini)’, in Cary Liu, Michael Nylan, and Anthony Barbieri-Low (eds.) Recarving China’s Past: Art, Archaeology, and Architecture of the “Wu Family Shrines”, New Haven:Yale University Press. Wang Haicheng (2014) Writing and the Ancient State: Early China in Comparative Perspective, Cambridge: Cambridge University Press. Wang Hongxing (1991) ‘Baoshan erhao Chumu qiqiqun yanjiu’, in Hubeisheng Jingsha Tielu Kaogudui 1991. Wu Hsiao-yun (2009) Shang Zhou shiqi chema maizang yanjiu, Beijing: Kexue Chubanshe. Zhejiangsheng Wenwu Kaogu Yanjiusuo (2005) Fanshan, Beijing: Wenwu Chubanshe. Zhongguo wenwu jinghua, Beijing: Wenwu Chubanshe, 1990. Zhonghua Renmin Gongheguo chutu wenwu xuan, Beijing: Wenwu Chubanshe, 1976.

458

21 ‘MEDICINE’ IN EARLY CHINA

MIRANDA BROWN‘MEDICINE’ IN EARLY CHINA

Miranda Brown

This essay seeks to answer the question, what was medicine like in early China? The question is more complicated than it would first appear. Why? Our current usage of the term medicine hardly matches the ancient Chinese practices scholars lump under the rubric of healing. To see this, we have only to consider the term medicine. Nowadays, the English term conjures up a host of images: prescription drugs carefully measured and dispensed by pharmacists, surgical procedures performed by M.D.s wielding scalpels and directing lasers, and a host of office equipment, stethoscopes, sphygmomanometers, and ultrasound devices. To be sure, some of these features had parallels in the early Chinese context. Ancient Chinese healers availed themselves of an array of devices: scalpels, heating stones, and needles. They also devised medicaments to treat conditions recognizable to any modern patient – throat discomfort, toothaches, fevers, chills, wounds from sharp objects, back pain, and impotence, to name but a few. Such ancient people also employed diagnostic techniques that resembled, however crudely, their modern counterparts: they listened to their patients’ breathing, counted the number of heartbeats, and examined the colour and texture of the skin. The similarities end there. Much of ancient Chinese healing fell outside the scope of modern biomedicine, for it encompassed exorcisms, invocations, meditation, fasting and dietetics, sexual cultivation or abstinence, breathing exercises, gymnastics, moxibustion, and acupuncture. Early healers also administered substances anathema to modern health care, the least of which included neurotoxins (lead and mercury) and exotic substances such as “man’s slime,” “earthworm excrement,” and the incinerated remnants of a “woman’s first menstrual cloth” (Harper 1998). Such healers also concerned themselves with a host of conditions without equivalents in modern Western medicine: blockages or heterodox movements in the vessels, disorders caused by spirits and demons, wind ailments, and so forth. The disjuncture between modern Western health care and ancient Chinese medicine is to be expected. The characters comprising the classical Chinese term we often translate as “medicine” (yiyao 醫藥) were broader in scope than any modern English equivalent. Yi 醫 not only referred to healers or those “who treated illness” in ancient texts, but also the attendants who looked after the various aspects of the body’s health: the women who oversaw the nursing of imperial children and the gentlemen who took charge of the ruler’s food and drink, for example. Yao 藥 covered an assortment of medicinal preparations, including those used not merely to alleviate the ills of the body but also the elixirs taken to forestall aging and death altogether (Brown, Forthcoming). 459

Miranda Brown

The goal of this essay is not to reconcile ancient and modern conceptions of medicine. Instead, it argues that the discrepancies reflect the fact that what we call “Chinese medicine” was no single thing, being practiced in various contexts by a range of parties and often to dissimilar aims. Toward this end, this essay explores the disparate contexts of medical practice: the temple, home, market, bureaucracy, and court.Though they sometimes shared common actors, ideas, and technologies of healing, as well as sources, each of these contexts had their own logics.

Temple Our discussion of contexts begins with the temple, the earliest attested site of healing. Here, I use “temple” in a metaphorical sense, to refer to attempts to communicate with or influence the numinous realm. Although some historians regard practices like divination and sacrifice as characteristics of the earliest, most primitive stage of medical history, this characterization is misleading (Brown 2012). The early Chinese, in fact, appealed to the divine through divination and sacrifice long after the emergence of alternative frameworks for combatting human sickness. The story of Chinese medicine often begins with the archives of the late Shang 商 kings, the earliest records of healing. Archaeologists recovered from these archives nearly five hundred records of medical divination, dating to the mid-thirteenth century bc. These records documented the Shang rulers’ persistent concerns with their health and the well-being of their closest associates. A typical example of one record reads, “Divined: As to the sick tooth, it is not due to Father Yi harming him [i.e., the king]” (Keightley 2001: 153). Another record noted with respect to the same ailment, “If we pray by means of these (offerings), the sick tooth will certainly be cured.” A second record, relating the illness of a military general, reveals that the Shang elite regarded sacrifice as the solution to illness. To determine what should be done for a toothache, someone at court ordered a divination: “Divined: (We) offer a dog to Father Geng (K18) (and) split open a sheep” (Keightley 2001: 154). Like the author of the first record, this one assumed that the answer to illness lay in appeasing the ancestors. The Shang kings were not alone in attributing illness to divine interference. Divinatory records retrieved from the tombs of southern aristocrats who died in the third and fourth centuries bc indicate a comparable set of beliefs. The case of a Chu 楚 nobleman illustrates this point. His physical woes – ranging from discomfort in the heart and afflictions of the abdomen to a loss of appetite – occasioned the intervention of twelve diviners. Over the course of a year, the diviners sacrificed no less than thirty-six pigs, six dogs, twenty-three sheep, nine oxen, and a horse (Kalinowski 2009: 394). To be sure, the solution proposed by the diviners was hardly new. The Shang elite had employed precisely such a strategy. What was novel was the imagined roots of the man’s illness. Unlike their Shang forebears, Warring States diviners ascribed illness to forces both within and beyond the ancestral realm. Besides royal ancestors, diviners blamed a wide cast of characters for headaches and heart illnesses: a pantheon of nature deities hungry for their share of offerings, vengeful ghosts, and the war dead incensed by the lack of sacrifices (Lai 2005: 11). Practitioners of temple medicine also began to look beyond the numinous realm for the sources of human sickness. After the fourth century bc, some of them began to argue that both the source and solution of illness lay in the moral reformation of princes as opposed to the offering of rich sacrifices. Take the famous story recounted in a historical chronicle regarding the sickness of the lord of Qi. Suffering from a chronic condition for more than a year, the lord attempted to remedy the situation by doubling the number of offerings to the deities, but to no avail. Frustrated by his lack of relief, the lord then proposed to sacrifice his diviners. His plan met with resistance from the chief minister, who explained that the gods were angry about the 460

‘Medicine’ in early China

quality of the lord’s governance rather than the quantity of sacrifices. In the minister’s view, the lord would be better served by becoming a better shepherd of the people (Brown 2012: 381–383). What is new here is the solution to the scourge of sickness. Whereas earlier accounts presented the spirits as hungry for sacrifices and prone to lash out at rulers who were stingy with their offerings, this account instead portrayed illness as the consequence of immoral behaviour. In so doing, it linked sin to suffering. I have said something about the methods associated with the curative traditions of the temple, but I have yet to mention the identities of its major players. Most likely, such parties included the diviners at court, of whom little is known. Based on fragmentary evidence from the Warring States, however, scholars have inferred that most of them were members of hereditary groups, possibly composed of lineages of diviners. Such diviners were not aristocrats themselves but rather the personal attendants or servants attached to noble households (Kalinowski 2009: 394). Diviners were not the only parties involved in temple healing. Anecdotal evidence also hints that aristocratic experts on the numinous realm – a group boasting of the heads of states, ministers, and the occasional noblewoman – tried their hand at identifying the causes of illness. Interpreting dreams and the results of divinations, such figures also attempted to discern which of the gods as well as which of the ruler’s actions were responsible for bouts of sickness. Such parties, finally, played a key role in determining what steps were to be taken to alleviate human suffering; their answers often ranged from offering sacrifices to encouraging rulers to improve the quality of their governance (Brown 2012). Sources from late antiquity suggest that charismatic healers, who sometimes attracted large followings in the countryside, also practiced a form of temple medicine. The leaders of the Celestial Masters movement (Tianshi dao 天師道) present the best-known example of this phenomenon. Beginning in the mid-second century ad, the movement’s leadership and its followers managed to capture southwest China for half a century and establish a theocracy there. Believing human sickness was punishment for sin, the Celestial Masters practiced a form of faith healing. Upon the onset of illness, the afflicted was to be moved into “chambers of quietness,” given “talisman” water to drink, and enjoined to confess sins. If the person improved, then he or she was presumed forgiven; if recovery did not follow, the person was exhorted to redouble his or her efforts to repent (Bokencamp 1997). The temple thus brought together a motley group of healers: diviners, aristocratic statesmen who professed a deep understanding of the numinous realm, and faith healers who conflated sin and suffering. It also drew upon a variety of beliefs regarding the sources of human sickness. For some, illness was the result of the displeasure of the ancestors, whereas for others, any number of spiritual forces – be they angry spirits, indignant gods, and malevolent ghosts – could visit the scourge of sickness upon the living. Finally, the temple encapsulated contrasting views about what measures were believed by the early Chinese to have been effective, including the offering of sacrifices, programs of moral rectification, and acts of penitence.

Healing at home The private residence or home represented another important locus of health care in early China. Received sources such as dynastic histories indicate that elite men and women collected books of medicine for personal use (Harper 1998). This picture has been corroborated by archaeological evidence, as archaeologists have retrieved collections of formulas, therapeutic models, and other works of medicine from the tombs of aristocrats and officials. While we are not privy to the reasons why they chose to be buried with such materials, we know that such men did not claim healing as their main vocation. Taken together, the evidence points to a 461

Miranda Brown

tradition of homecare, one that exhibited both striking continuities and differences from the medicine found in the temple. Without a doubt, the medicine of the early Chinese home overlapped with that of the temple. Indeed, echoes of temple medicine recur in the private manuscript collections of Qin and Han officials. The author of the Zhoujiatai 周家臺 manuscripts, which belonged to a Qin official of the third century bc, held similar beliefs about human illness. His discussion of incantations blamed human illness (at least some of the time) on the pesky interference of spirits. According to the author, a person suffering from tooth decay was to perform a ritual move for an exorcism, not once but three times, while looking upon an old wall. The afflicted was then to address the god of the wall and to pledge a pair of black oxen in return for recovery (Hubei sheng Jingzhou shi Zhouliangyuqiao yizhi bowuguan 2001: 5–6).The author of a remedy book, recovered from a second-century tomb, offered a similar solution to the torment of inguinal swelling. The sick person was to grasp a pestle, perform the ritual move for exorcism twice, and utter the following incantation, “Spouter expel the Hu (a demon) once. Spouter expel the Hu twice. Spouter expel the Hu thrice” (Harper 1998: 259). Commonalities notwithstanding, the theories of illness associated with the ancient home sometimes offered a contrast to that of the temple. The manuscripts kept in the personal collections of Han noblemen and officials also supplied new answers to the venerable question regarding the causes of illness. Such answers are evinced in the first references to mai 脈: the vessels believed to connect the inner body to the outside environment and to transport the blood and qi (Porkert 1974: 77–98; Harper 1998). Unlike older views of illness, this theory focused on dysfunctions in the vessels. In this regard, consider a manual that instructed readers on the diagnosis of illness. Nowhere in this manual did the author mention the role of malevolent spirits or angry ancestors in causing physical discomfort or sickness. The numinous realm, in fact, never once came up.When explaining the roots of sickness, the author stuck to disorders in the vessels. For example, he claimed that a disturbance in one of the twelve vessels, the Foot Lesser Yin Vessel, would engender a host of dissimilar ailments, among them rage, paranoia, impaired vision, a shortage of qi, and a darkened complexion (Brown 2015: 51). Starting in the third century bc, the authors of medical treatises and the makers of therapeutic models also presented novel solutions to the problem of human sickness, solutions that focused on qi. The newly discovered Laoguanshan 老關山 tomb, dating to the early second century bc, presents one of the earliest signs of this development. Among the many possessions of the elite occupant, archaeologists discovered a miniature lacquer figurine in the tomb. Painted on the surface of the figurine were the pathways of the vessels and the location of the acupoints (Liang 2015). Such marks reveal that ancient men and women attempted to restore balance in the body by stimulating specific points on the skin, points believed to represent the interface between the vessels and internal processes, on the one hand, and the external environment, on the other (Porkert 1974: 200). Other archaeological discoveries, roughly contemporaneous, similarly point to the rising emphasis placed on new-fangled techniques used to alleviate sickness. The author of one treatise of the third century bc explained how to “control and treat” ailments at the sites of vessels. Towards this end, he elaborated a host of techniques other than sacrifice or exorcism, particularly the use of moxibustion and acupuncture, to correct imbalances of qi (Zhangjiashan ersiqi hao Han mu zhujian zhengli xiaozu 2006: 115–128). Besides attesting to the rising popularity of acupuncture, manuals also suggest that the private residence represented a site for a flourishing tradition of drug formulary. Manuscripts from the late third century bc offer our first view of the beginnings of this tradition. Kept with directions for incantations, these formulas employed strange and exotic substances. Consider the following excerpt: “For those with a lump in the abdomen, heat a sword with a square end, 462

‘Medicine’ in early China

tempering it by dipping it in fine alcohol. Women should use it twenty-seven times, men seven times. Give this to drink to [the afflicted] and the lump will disappear” (Hubei sheng Jingzhou shi Zhouliangyuqiao yizhi bowuguan 2001: 129). One manuscript from the northwest frontier indicates that the tradition achieved a level of sophistication by the mid-first century ad. There, archaeologists discovered ninety odd fragments of text on wood, mostly covered with drug formulas. Taken together, these fragments reveal that the Han elite were in the habit of using a wide array of substances – among them sixty-three herbs, twelve animal products, and sixteen minerals (Xie 2005: 79). Such substances came together in complex prescriptions, involving as many as fourteen precisely-measured substances, to treat an impressive list of ailments: various kinds of coughs, lumps in the abdomen or heart, internal coldness, cold damage disorders, and swollen scrotums that ooze from their bases a yellow juice. The drugs also came in a wide range of preparations (pills, pastes, powders, and tinctures), as well as differing styles of administration (before meals, after meals, with alcohol, and with porridge) (Xie 2005: 79–80). Consider, for example, the following, used to relieve an ailment affecting the throat: Two liang [27.6 g] each of these eight substances: Shu; Fangfeng; Chinese wild ginger; ginger; Cassia twigs; Szechuan aconite root; Szechuan pepper; balloon flower. Pulverize, combine, and mix all of these ingredients. Before supper, drink a squareinch spoonful [2.76 ml] of the medicine with millet porridge. (Yang 2017: 260–61). Healing at home was not limited to treating sickness but also encompassed self-care and the preservation of vitality (Lo 1998). Take the Pulling Book (Yinshu 引書), a manuscript that once belonged to a Han official of the early second century bc. It supplied directions for therapeutic stretches and breathing exercises. According to the author of the Pulling Book, these techniques were helpful not merely for alleviating symptoms (back discomfort, knee pain, abdominal swelling, exhaustion, and so forth); more remarkably, they also purportedly prevented the onset of illness altogether. On the latter point, this author recommended that readers harmonize their patterns of eating, sleeping, and sexual activity with the seasons. By his account, such regimens worked to shore up the body’s store of vital essence and offered protection against illnesses resulting from the lack of harmony between the body and cosmic cycles (Lo 2014). In the same regard, elites also kept copies of manuals for sexual cultivation, containing techniques not only for enhancing sexual performance but also for ensuring bodily well-being and procreative success. The silk manuscript, “The Ten Questions” (Shiwen 十問), dating probably to the third century bc, remarked upon the health benefits of properly executed sexual activity. As its author put it, “I have heard that the Master practices coitus with Yin to become strong, and sucks in heaven’s essence to achieve lengthy longevity” (Harper 1998: 402). Thus far, we have covered what sorts of texts elite men and women kept among their personal effects, but this leaves the question whether such texts were actually read as manuals. After all, it is conceivable that the early Chinese elite regarded such manuscripts as being something akin to Joyce’s Ulysses, works that they collected for their prestige value but had little intention of ever reading. While these questions must be kept in mind, our sources do indicate that, like their counterparts in the Roman empire, the early Chinese had something of a do-it-yourself mentality. Anecdotal evidence hints that the early elite accumulated collections of formulas for the purpose of self-medicating – and sometimes to disastrous effect (Brown 2015: 137–140). Moreover, if one scans the contents of the formula books, it becomes evident that the early elite found it useful to have remedies on hand. Many of the ailments discussed in those collections did not 463

Miranda Brown

require a telling touch or trained eye to detect. In fact, most of them were the mundane illnesses one might expect anyone to recognize: fevers and chills, moles or birthmarks, atrophy in the limbs, abscesses, tooth decay, scabs, erectile dysfunction, metal and heat wounds, and the many varieties of haemorrhoids (Brown 2015: 48). The substances were moreover items that one could procure on the market, grow in the garden, or prepare at home. Consider a relatively complex treatment, recovered from a first-century site, which offers further evidence of a tradition of homecare: A formula for a decoction to treat a persistent cough, as well as qi reversal and counterflow: Aster, 7 bundles. Lily-turf [or perhaps Chinese asparagus?], 1 sheng [200 ml]. Tussilago, 1 sheng [200 ml]. Coltsfoot, 1 sheng [200 ml]. Gypsum, half sheng [100 ml]. White [1 illegible character], 1 [1 illegible character]. Cassia twigs, 1 chi [23.75 cm]. Honey, ½ sheng [100 ml]. Jujube, 30 pieces (mei). Panhsia, 10 pieces (mei). Mince all ten of these substances. Do not mince the Panhsia. Pour one dou [2,000 ml] and six sheng [1,200 ml] of water into a pot, bringing it to a boil six times. Drain and get rid of the sediment. Drink a small cup warm three times a day. If the medicine is used after the night, it should be boiled again. The illness will be cured in three to four days. (Yang 2017: 288). The formula is complex, as there were many ingredients and many steps in the preparation. It did not however require the services of someone specifically trained in the curative arts. The preparation – mincing the materia medica, decocting the mixture, and then reboiling it – was straightforward. The techniques involved, in fact, would have been familiar to anyone experienced with cooking.The medicine was thus something that any elite man or woman – or better still, their servants – could have made. Of course, not all of the techniques detailed in ancient manuals could be learned just from casual reading. Certain techniques – particularly reading the hue or applying needles and heated stones – did require at least some training (Sivin 1995; Lloyd 1996: 29). While one might expect that training implied the existence of a profession, such an assumption is unwarranted. Our sources suggest that members of the early elite studied healing as a serious pursuit, but not necessarily as a vocation or career. Consequently, some of them sought out famous masters, who transmitted their secret collections of formulas, the contents of which were first memorized and then copied down for personal use (Sivin 1995: 182–183). Despite the fact that they were not career healers, some of these men and women became quite well known for their curative prowess. They treated not merely members of their own households, but also neighbours, colleagues, and friends (Fan 2010: 334; Fan 2013: 69). Indeed, it became something of a trope for a young man to study medicine to take care of his parents (Brown, Forthcoming). Such men also authored treatises on subjects such as formulary and acupuncture, distinguishing themselves 464

‘Medicine’ in early China

through their compilations and reconstructions of ancient medical works. Still other men, also of independent means and aristocratic birth, trained disciples in their private residences (Sivin 1995: 180–182). We have seen that the varieties of medicine found in the personal residence were many. Some of them resembled the sorts of healing connected to the temple, namely, medicine focused on identifying and pacifying the divine roots of human illness. Others, in contrast, presented different, though not necessarily incompatible, theories of illness, including those that chalked up human sickness to blockages or obstructions of qi. Home remedies also stressed the use of techniques aimed at restoring the health through the administration of drug formulary, acupuncture, moxibustion, sexual cultivation, and stretching, as well as sacrifice, exorcism, or prayer.

Markets We know much less about the healers who populated the market, that is, the men and women who made their living peddling their goods and services. We can blame this on our sources, which did not record the beliefs and experiences of itinerant healers in ancient times. Such a silence is unsurprising considering that some, if not many, of these itinerant healers were probably illiterate.The elite bias of our sources, which favours the words and deeds of the ruling elite, is also at fault, for they leave us only with isolated anecdotes, snobbish comments, and evidence of social restriction. Although they say little about the beliefs held by itinerant healers, such sources nevertheless provide us with some hints as to the social position of those in the market. Limitations of sources notwithstanding, we can infer that the market was where the early Chinese procured the raw materials with which to make their remedies at home – the pellets, plasters, and decoctions that we read about earlier. Indeed, Han works abound with stories, some undoubtedly apocryphal, telling of immortals or hermits who concealed themselves from the world, disguised as lowly medicine peddlers. For example, the classical scholar and official Zhang Kai 張楷 (fl. ad 140) reportedly fled from his many clients and pupils by disguising himself as a medicine peddler (DeCrespigny 2007: 1060). While the story’s veracity is hard to assess, other sources, of a more prosaic quality, including a register of drugs listing the price and the name of the person who sold the materia medica, confirms the existence of a medicine market (Yang 2017: 301). The market was also the place where itinerant healers – required by Han code to be registered in the marketplace and subject to state regulation – offered their services (Jin 2005: 25). In this regard, collectors have recovered seal stamps with references to such individuals as “Han, the Fever-Sore Remover” (Han Quchen 韓去疢), “Abscess Meng” (Meng Yong 孟癰), and “Sickly Su” (Su Shou 蘇瘦). Some scholars suggest that such seals once belonged to men and women of base origin who performed the dirty work of health maintenance: lancing abscesses and other ungainly growths on the bodies of their social betters; scraping off scabs; puncturing the flesh with needles; palpating the ankle bones; and plastering animal manure on wounds (Luo 1979 7/20; Chen 1988: 288–292; Li 2000: 65). While silent about the individual experiences of medicine peddlers, our sources do reveal that some, if not most, of these individuals were of debased station. For example, the daybook recovered from an early third-century tomb suggests that no parent dreamt of his or her child becoming a healer. “If the child is sired on a ren or yin day,” the author warned, “it will be inauspicious, for the daughter will become a healer” (Shuihudi Qinmu zhujian zhengli xiaozu 1990: 253). Ritual classics corroborate this picture. One work, from the first century bc, admonished gentlemen to “avoid becoming familiar” with healers, invocators, and craftsmen (Brown 2015: 13). 465

Miranda Brown

Though not all itinerant healers were of low station, some of them longed to move up in the world. Consider the case of Lou Hu 樓護 (fl. ad ca. 25), born to a family of healers (shiyi 世 醫). While still in his teens, Lou Hu mastered the medical canon. In the words of one chronicler, “Lou Hu was able to recite for memory the classics of medical formulas, pharmacopeia, and long essays regarding the occult arts.” Lou Hu then followed his father to the capital, and together they set up a practice. His talents soon impressed rich and powerful patrons, who urged the youth to seek greener pastures in officialdom. “With your talents,” the youth was told, “Why not acquire the learning befitting an official?” Sensing an opportunity, Lou Hu abandoned the craft of his father, devoting himself instead to the study of the classics, and achieved success as an official (Loewe 2000: 412). The case of Lou Hu proves that being from a family of hereditary healers did not bar a man, at least in principle, from entering the official ranks. Still, the fact that patrons urged Lou Hu to abandon the family trade implies that the life of an itinerant healer, even a successful or well-connected one, left something to be desired. The evidence regarding itinerant healers – the men and women who made their living by offering their services in the market – discloses that they lived without the security of diviners or the prestige enjoyed by the noble experts associated with the temple. Low status notwithstanding, it would be a mistake to assume that such parties were unimportant. As Nathan Sivin points out, such healers represented a significant portion of the health care picture in premodern times (Sivin1995: 195). These individuals – who perform foot and head massages, or apply acupuncture and moxibustion in train stations today – remain a force to be reckoned with in the present (Hsu 1996: 429).

Bureaucracy A great deal of healing in ancient China happened in the bureaucracy, at the hands of the functionaries responsible for running bureaus and counties, supervising corvée and conscript labourers, and maintaining the military colonies in the northwest. Our knowledge about the bureaucracy and its system of health care derives from a number of sources: received sources such as dynastic histories, the administrative records preserved in trash deposits throughout the Qin and Han empires, and bureaucratic manuals recovered from the personal tombs of early imperial officials. From these sources, it becomes clear that early officials were involved in health care. They maintained supplies of medicaments and sometimes served as healers who would administer acupuncture or moxibustion treatments (Xie 2005: 98). They also kept records of drug formulary, something apparent from the sheer volume of information about remedies found at administrative sites. Before delving into the distinctive features of the bureaucracy, we must first explain its relationship to the medicine of other contexts: the private collections of drugs discovered in tombs were similar to the fragments of formulas unearthed in the trash deposits associated with former military outposts. Consider the following formula, taken from a Han administrative site located in the southeast, a formula like those recovered from the personal tombs of officials and aristocrats: “For those stricken with a violent eruption of heart pain that burns, mince two seeds of Pennycress. Pulverize one dried ginger and one Cassia twig. Combine all three of these substances.Take a three-finger pinch, knuckle deep, warming it with concentrated alcohol” (Brown 2015: 80). In large part, the overlap between the medicine of the bureaucracy and that of the household is unsurprising. Many of the tombs that preserved the personal collections of formulas once belonged to Qin and Han officials, including those who had worked at the local level or on the frontier. We know furthermore that many Han officials held medical positions within 466

‘Medicine’ in early China

the bureaucracy for several years of their career, making it likely that many of them became familiar with the basics of drug formulary and acupuncture while still in the service (Brown, Forthcoming). Commonalities aside, the medicine of the bureaucracy also stood apart from other varieties of healing. Surviving bureaucratic documents did not blame illness on ghosts or demons (Yang 2016). Nor did they allude to the therapeutic use of sacrifice, prayer, exorcism, or moral reformation to relieve pain and sickness. The medicine of the bureaucracy also offered a contrast to that of the household in other respects. It focused on acute illnesses, serious conditions such as abscesses and wounds from metal objects, as opposed to therapeutic stretches designed to enhance the general well-being and vitality of officers. Such documents also depart from the interest in sexual health or procreative success, an interest pronounced in personal collections but of little obvious interest to the bureaucracy. The discussions of medicine practiced in an official capacity were bureaucratic in flavour. Such discussions, in fact, testify to the officials’ mania for documentation. From at least Qin times, the government produced certain kinds of records of mortality. Qin codes, for example, called for the sudden deaths of slaves “who died without illness” to be duly noted in the record (Shuihudi Qinmu zhujian zhengli xiaozu 1990: 24). In addition, officials documented the outbreak of epidemics. The Shuihudi manuscripts provide a case in point, for one model record indicates that authorities commanded healers to examine suspected victims of leprosy (Shuihudi Qinmu zhujian zhengli xiaozu 1990: 156–158; McLeod and Yates 1981: 153). The authorities also kept track of illness within its ranks, recording the names of sick conscript labourers and military officials. The following record, from the southeast, presents a case in point: “Three men are ill: Gu, Mao, and Cheng” (Chen 2012: 225).The Qin practice of recording instances of sickness endured for centuries. One fragment from the northwest reads: Li Qiang, a garrison soldier from Shi’an hamlet in Xiangyuan county of Ping’gan Kingdom, aged thirty-seven: On the forty-fourth day of the second month of the fifth year of the era of Original Beginnings (69 bc), Li became ill from distension in his heart and abdomen. He has died. The Front Assistant to the Agricultural Office of the Right reports. (Brown 2015: 82) Such records, increasingly elaborate after the first century bc, not only listed the names of sick men along with extended descriptions of the illnesses and the periods of incapacitation – they also commented upon the nature and results of treatment and sometimes stated the name of the person who administered acupuncture or moxibustion. In this connection, one excerpt noted that a solider had “a headache on the eighth day of the fourth month. He suffers from hot and cold spells, has drunken five dosages of medicine but has yet to improve” (Brown 2015: 82). The emphasis within the bureaucracy on treating and tracking acute illnesses makes sense. The bureaucracy’s interest in health care reflected utilitarian concerns, namely, its need to mobilize manpower, something clear from the fact that officials felt compelled to resolve illnesses that struck members of their own ranks. The account of an architect of the Imperial Palace, who lived in the first century ad, further hints that officials in Han took responsibility for the health of those under their command. According to the account, at one point, an epidemic struck the labourers working on the project overseen by the official.The architect “made rounds of inspection and went to the sick corvée labourers, personally seeing to the preparation of the medicines and porridge for the sick.” Such efforts apparently bore fruit, as “the number of men who died from illness was reduced” (Brown 2015: 84). The architect clearly had pressing reasons for 467

Miranda Brown

wanting to manage the situation himself. Qin and Han officials were judged on their ability to finish projects on time and within budgets, so the architect probably determined that his future career prospects depended on the conscript labourers making a speedy recovery. The Chinese bureaucracy thus represented an important locus for health care. The bureaucracy maintained a rudimentary health care system for soldiers, one connected to the tradition of health care associated with the private residence. Although there were actors, techniques, and ideas common to both traditions, the rationale for practicing medicine in the bureaucracy offered a contrast to that of the household in several respects. To begin with, the goals of home medicine were broader, encompassing the treatment of ills as well as the enhancement of vitality. By contrast, the medicine of the bureaucracy was narrower in its scope, revolving around the healing of acute illnesses that struck the population at large or the ranks of officials, conscripts, and soldiers. All this, of course, is to be expected. The bureaucracy’s interest in health was more than academic or humanitarian (despite what the imperial state would have its subjects believe). Much like famine relief or flood control, the health of subjects – specifically, of corvée labourers, soldiers, and conscripts – had ramifications for imperial revenues and for the success of public works projects and military campaigns.

Court We are ready to tackle the last of the contexts in which healing occurred, the court. By court, I mean not only the seats of the Shang, Zhou, Qin, and Han dynasties but also the various regional centres of culture and power. As we will see, health care at court was intimately connected with a broader system of noble patronage. The importance of this patronage system cannot be emphasized enough, for it drove the priorities as well as the shape of court medicine. The first glimpses of medicine at court date to the Warring States period (453–221 bc), found in works of persuasion and historical annals relating the past to make a specific political point. There, early healers appear mostly as fleeting figures: sometimes as itinerant physicians who issued stern warnings to benighted nobles and at others as the personal attendants dispatched to relieve foreign allies of illness. We also have a highly idealized picture of court medicine from one canonical text, the Rites of Zhou (Zhouli 周禮), which suggests that royal healthcare was organized, at least by the third century bc. According to it, the court was staffed by attendants who catered to the diverse needs of the ruler’s body and his administration: from the ruler’s wounds to his food and drink to the animals in his stables to the epidemics that wreaked havoc upon his realm (Li 2000: 66). While scholars sometimes emphasize – or even exaggerate – the extent to which court medicine differed from the varieties seen in other contexts, particularly in the temple, we must also bear in mind the commonalities. The Classic of the Vessel or Maijing 脈經 (third century ad), compiled by the Jin court physician Wang Xi 王熙 (ad 180–270?), discloses one such commonality: the long-standing belief that human illness reflected the workings of ghosts and spirits. In this seminal work, a classic of vessel theory, Wang instructed readers on how to interpret the various patterns of the pulse. Interestingly, Wang did not blame the irregularities in the pulse just on dysfunctions in the vessels – for example, blocked or stagnated qi. He also alluded to the malevolent interference of the spirits, which could be detected by taking the pulse. Remarking on one such pattern, Wang wrote, “it means that someone in the sick person’s household must have died from the harm afflicted by ghosts, demons, and winds.” (Wang 1929–1936.2/6; Li 2007; Li 2009). Though related to the varieties seen in other contexts, court medicine nevertheless had its own features. The healers who practiced their arts at court – ranging from the women charged 468

‘Medicine’ in early China

with overseeing royal nursing to the masters of medicine and the imperial physicians – were recruited as personal retainers of powerful figures. In other words, they were part of a broader system of noble patronage, and thus no different from the myriad of retainers who flanked the court: military advisors, political persuaders, astrologers, bards, jesters, and dwarfs. Such healers, often of genteel or even aristocratic origin, were awarded positions after being summoned to court for an audience, usually on the basis of their reputation or their facility with the various arts. Take Li Shaojun 李少君 (fl. 130 bc), the notorious favourite of Emperor Wu 武 (r. 141–87 bc) of Han. This future Master of Medicine received his position at the emperor’s side after it became known that he, Li, knew how to forestall the effects of aging (Sima 1982: 12.454– 12.455). As one might expect, such positions came with their share of rewards, material riches as well as bureaucratic positions. Two aristocrats, Zhou Ren 周仁 (fl. 142 bc) and Diwu Lun 第五 倫 (fl. ad 50), illustrate this point, as both of them parleyed their initial appointments as regional court physicians into something bigger. Zhou Ren achieved the rank of Grand Counsellor of the Palace and the Superintendent of the Palace, whereas Diwu Lun subsequently became the Excellency over Works, or one of the senior officials of the realm (Ban 1996: 70.2003, DeCrespigny 2007: 145). Given the stakes involved, it is unsurprising that the healer behaved like any other courtier. Like all potential favourites, court healers faced fierce competition from rivals, looking to upstage competitors and to win themselves a cosy position at the ruler’s side. If the account of Chunyu Yi 淳于意 (fl. ca. 216–150 bc) can be trusted, renowned healers found their prognoses routinely challenged by other men, who had everything to lose from a colleague’s rising star (Hsu 2010: 85). Such men would flatly contradict the famed healer’s prognoses and mock him in the presence of other influential men. For their part, rulers did nothing to encourage collegiality or a sense of brotherhood among court healers. On the contrary, such rulers encouraged cutthroat competition by summoning figures known for their curative powers – in spite of already having personal medical attendants in their retinues. Making matters worse still, they also called upon their guests to diagnose patients or cases of illness previously seen by one of their own attendants (Hsu 2010: 76–77). Like any courtier, the healer found himself in the position of being something of an entertainer. The story told of Guo Yu 郭玉 (first century ad), the emperor’s assistant medic, points to the fact that a good courtier was not only an able healer but also the supplier of diversion. According to the chronicler, Guo Yu’s skills at healing piqued the interest of the emperor, who decided that he should be tested (like any other courtier). The emperor selected a eunuch with delicate wrists, concealing his face and body behind a curtain.Then a girl was also placed next to the eunuch with her arm outstretched. Guo Yu was then ordered to examine the pulse of both individuals. Much to the delight of the emperor, Guo Yu sniffed out the ruse, declaring that one of the pulses was distinctly male whereas the other was female (DeWoskin 1983: 75). The incorporation of healers into the system of patronage also shaped the aims of medicine at court, aims that included the pursuit of immortality as well as the curing of sickness. To see this, we must bear in mind that Qin and Han emperors, as well as members of their entourage and regional princes, were notorious for their quest for super-longevity, if not immortality. Such an interest led members of the ruling elite to spend personal fortunes buying cinnabar and other ingredients for elixirs. It drove them to adorn the walls and tiles of their tombs with pictorial representations of immortals carrying life-enhancing herbs and to paint gorgeous reliefs of the paradises of the immortals (Campany 2009: 52). The noble patron’s fixation, so marked among the ruling elite, naturally left its imprint on the image of the ideal healer, who ancient authors sometimes conflated with would-be immortals. This is clearest in the case of Hua Tuo 華佗 (d. ca. 208 bc), also called the God of Medicine. In various sources, Hua Tuo’s traits – specifically, his 469

Miranda Brown

penchant for evading patrons and transforming his appearance – all matched the qualities commonly attributed to immortals. Even Hua Tuo’s curative exploits – his successful deployment of mundane substances such as garlic purees to exorcise snakes and his ability to perform internal surgery – are reminiscent of the stories told of immortals, also credited with incredible powers of healing (Brown, Forthcoming). Indeed, Hua Tuo was believed by his contemporaries to have maintained a youthful appearance when already a century old, a trick he purportedly passed onto his disciples. One of them, a highly regarded healer in his own right, reportedly had at ninety “bright eyes and keen ears,” and, more improbably, a full set of working teeth (DeWoskin 1983: 140–153). Accordingly, Hua Tuo became synonymous with the quest for immortality. His name became a selling point for formulas advertising their ability to forestall aging, if not to confer deathlessness on consumers (Fan 2010: 15–16). In certain regards, medicine at court was a far cry from the sort found in other contexts, particularly the bureaucracy. Unlike the prosaic descriptions of healing in the bureaucracy, the stories told of court healers focused instead on the miraculous and entertaining aspects of their practice. The tangled relationship between patron and client moreover left its imprint on the content of court medicine. Unlike the varieties found in the bureaucracy or even the temple, medicine at court could not be disentangled from the quest for immortality, a quest that expressed itself in the popularity of formulas believed to confer superhuman powers and that insinuated itself in the image of the King of Medicine.

Conclusion In sketching the various contexts in which health care occurred in early China, I have attempted to give readers a sense of why the phrase “Chinese medicine” might be misleading. Lest I be misunderstood: one should not see each of these contexts (temple, home, bureaucracy, and court) as unrelated microcosms. As I have demonstrated, there were actors who moved between contexts, taking with them ideas and technologies from other spheres of life. For example, the official came into contact with health care in different circumstances: in a private capacity, as the head of a household, as an administrator, charged with tracking and treating illnesses that threatened the well-being of the body politic, and as a courtier, seeking to make his name. Yet despite the connections that existed between the various spheres in which healing was practiced, medicine in early China was not a monolith. The diversity of medical practice, I would argue, largely reflected the fact that the aims of health care were many. In some contexts, particularly that of the temple and home, the spirits and sacrifices figured prominently in efforts to ensure the physical well-being of patients. In other contexts, such as the bureaucracy, healers endeavoured to treat the ills of the body, or to rid the patient of his or her suffering. In the private residence or at court, health care was more expansive in its scope. In such contexts, the art of medicine not only encompassed the technologies of healing but also the general pursuit of well-being, longevity, and even immortality. Given such diversity, it is not surprising that the men and women who healed were a motley bunch. As Nathan Sivin and others have pointed out, no single profile for the therapist existed in early China, as such figures hailed from dissimilar backgrounds and were divided in their motives for practicing their arts (Sivin 2012: 224; Fan 2010: 334). In closing, it would be worth ruminating on the relationship between the situation in the ancient period and those of later eras of Chinese history, asking whether such diversity also existed. Admittedly, the situations in later periods were dissimilar to those of ancient times. One finds, for example, beginning in the early middle period, influence from foreign traditions of medicine, notably, those derived from India, the Middle East, and even the Hellenistic world 470

‘Medicine’ in early China

(Fan 2005; Salguero 2014; Nappi 2009: 114, 141). In addition, the social practice of medicine was transformed due to vigorous state patronage under the Tang, Song, and Yuan, as well as the emergence of more accepting social attitudes that facilitated the rise of occupational healers as a social group. Finally, the face of medicine was altered by the arrival of modern biomedicine at the end of the nineteenth century and the creation of a dual health care system (Shinno 2007; Fan 2013; Hymes 1987; Lei 2014). Interestingly, such developments did little to alter the fact that health care in China has been a heterogeneous phenomenon for most of its history and remains so to this day. Chinese health care consumers continue to consult a wide array of actors: the priests who pray to deities and ghosts, the charismatic soothsayers and millenarian sect leaders who promise their acolytes long life and freedom from sickness, the acupuncturists who treat back pain by piercing or heating the skin, the classically-trained doctors who mix herbal concoctions to restore balance in the qi, and the grannies who nurse coughs and tend to wounds. The contexts in which people seek relief from bodily suffering or to enhance their vitality are similarly diverse (Hsu 1999). People in China not only receive health care in the Western-styled hospital, but also in the temple, the marketplace, the local acupuncturist’s office, and, of course, the home.

Works cited Ban, Gu 班固 1996. Hanshu 漢書, Beijing, Zhonghua shuju. Bokencamp, Stephen 1997. Early Daoist Scriptures, Berkeley, University of California Press. Brown, Miranda 2012. Who Was He? Reflections on China’s First Medical ‘Naturalist’. Medical History 56. Brown, Miranda 2015. The Art of Medicine in Early China: The Ancient and Medieval Origins of a Modern Archive, New York, Cambridge University Press. Brown, Miranda Forthcoming. Who Was Hua Tuo? Reflections on an Art Without a Role. Unpublished manuscript. Campany, Robert Ford 2009. Making Transcendants: Ascetics and Social Memory in Early Medieval China, Honolulu, University of Hawai’i Press. Chen, Wei 陳偉, Chen Youzu 何有祖, Lu Jialiang 魯家亮, Fan Guodong 凡國棟 2012. Liye Qin jiandu jiaoshi 里耶秦簡讀校釋, Wuchang, Wuhan daxue. Chen, Zhi 陳直 1988. Wenshi kaogu luncong 文史考古論叢, Tianjin, Tianjin guji. Decrespigny, Rafe 2007. Biographical Dictionary of Later Han to the Three Kingdoms (23–220 ad), Leiden, Brill. Dewoskin, Kenneth 1983. Doctors, Diviners, and Magicians of Ancient China, New York, Columbia University Press. Fan, Ka Wai 2005. Couching for Cataract and Sino-Indian Medical Exchange from the Sixth to the Twelfth Century ad. Clinical and Experimental Ophthalmology, 33, 188–190. Fan, Ka Wai (Fan Jiawei 范家偉) 2010. Zhonggu shiqi de yizhe yu bingzhe 中古时期的醫者與病者, Shanghai, Fudan daxue. Fan, Ka Wai (Fan Jiawei 范家偉) 2013. The Period of Division and the Tang Period. In: Hinrichs, T. J. H. A. L. L. (ed.) Chinese Medicine and Healing: An Illustrated History, Cambridge, MA: Belknap Press of Harvard University Press, 65–97. Harper, Donald 1998. Early Chinese Medical Literature:The Mawangdui Medical Manuscripts, London and New York, Kegan Paul International. Hsu, Elisabeth 1996. Innovations in Acumoxa: Acupuncture Analgesia, Scalp and Ear Acupuncture in the People’s Republic of China. Social Science Medicine, 42, 421–430. Hsu, Elisabeth 1999. The Transmission of Chinese Medicine, Cambridge, Cambridge University Press. Hsu, Elisabeth 2010. Pulse Diagnosis in Early China: The Telling Touch, Cambridge, Cambridge University Press. Hubei sheng Jingzhou Shi Zhouliangyuqiao Yizhi Bowuguan 湖北省荊州市周梁玉橋遺址博物館 2001. Guanju Qin Han mu jiandu 關沮秦漢墓簡牘, Beijing, Zhonghua shuju. Hymes, Robert 1987. Not Quite Gentlemen: Doctors in the Sung and Yuan. Chinese Science, 8, 9–76. Jin, Shiqi 金仕起 2005. Gudai yizhe de juese – jian lun qi shenfen yu diwei” 古代醫者的角色-兼論 其身份與地位. In: Li, J. (ed.) Shengming yu yiliao 生命與醫療, Beijing: Zhongguo dabaike quanshu chubanshe.

471

Miranda Brown Kalinowski, Marc 2009. Diviners and Astrologers Under the Eastern Zhou. In: Kalinowski, J. L. A. M. (ed.) Early Chinese Religion. Part One: Shang Through Han (1250 bc–220 ad). Leiden: Brill. Keightley, David N. 2001. The ‘Science’ of the Ancestors: Divination, Curing, and Bronze-Casting in Late Shang China. Asia Major, 14, 143–187. Lai, Guolong 2005. Death and the Otherworldly Journey in Early China as Seen Through Tomb Texts, Travel Paraphernalia, and Road Rituals. Asia Major, 3rd Series 18, 1–44. Lei, Sean 2014. Neither Donkey Nor Horse: Medicine in the Struggle over China’s Modernity, Chicago, University of Chicago Press. Li, Jianmin 李建民 2000. Sisheng zhi yu : Zhou Qin Han maixue zhi yuanliu 死生之域 : 周秦漢脈學之源 流, Taipei, Zhongyang yanjiuyuan lishi yuyan yanjiusuo. Li, Jianmin 李建民 2007. Xianqin Lianghan bingyinguan ji qi bianqian – yi xin chutu wenxian wei zhongxin” 先秦兩漢病因觀及其變遷-以新出土文物為中心. Guwenzi yu gudai shi 古文字與古代 史, 1, 453–480. Li, Jianmin 李建民 2009. They Shall Expel Demons: Etiology, the Medical Canon and the Transformation of Medical Techniques Before the Tang. In: Kalinowski, J. L. A. M. (ed.) Early Chinese Religion. Part One: Shang Through Han (1250 bc–220 ad). Leiden: Brill. Liang, Fanrong 梁繁榮, Ceng Fang 曾芳, Zhou Xinglan 周興兰, Xie Tao 谢濤, Lu Yinke 卢引科, Wang Yi 王毅, Jiang Zhanghua 江章華 2015. Chengdu Laoguanshan chutu jingxue xiuqi renxiang chutan 成都老官山出土經穴髹漆人像初探. 中國針灸, 35, 91–93. Lloyd, Geoffrey 1996. Authorities and Adversaries: Investigations into Ancient Greek and Chinese Science, Cambridge, Cambridge University Press. Lo,Vivienne. 1998. The Influence of Yangsheng Culture on Early Chinese Medical Theory. Ph.D School of Oriental and African Studies. Lo, Vivienne 2014. How to Do the Gibbon Walk: A Translation of the Pulling Book (ca. 186 bce), Cambridge, Needham Research Institute Working Papers. Loewe, Michael 2000. A Biographical Dictionary of the Qin, Former Han, and Xin Periods (221 bc–ad 24), Leiden, Brill. Luo, Fuyi 羅福頤 1979. Han yinwen zizheng 漢印文字徴, Beijing, Wenwu chubanshe. McLeod, Katrina C.D. and Robin D.S. Yates 1981. Forms of Ch’in Law: An Annotated Translation of the Feng-chen shih. Harvard Journal of Asiatic Studies, 41, 111–163. Nappi, Carla 2009. The Monkey and the Inkpot: Natural History and its Transformations in Early Modern China, Cambridge, MA, Harvard University Press. Porkert, Mansvelt 1974. The Theoretical Foundations of Chinese Medicine: Systems of Correspondence, Boston, Massachusetts Institute of Technology Press. Salguero, Pierce 2014. Translating Buddhist Medicine in Medieval China, Pennsylvania, University of Pennsylvania Press. Shinno, Reiko 2007. Medical Schools and the Temples of the Three Progenitors in Yuan China: A Case of Cross-Cultural Interaction. Harvard Journal of Asiatic Studies, 67, 89–133. Shuihudi Qinmu Zhujian Zhengli Xiaozu, 睡虎地秦墓竹簡整理小組 1990 Shuihudi Qinmu zhujian 睡 虎地秦墓竹簡, Beijing, Wenwu chuban. Sima, Qian 司馬遷 1982. Shiji 史記, Beijing, Zhonghua shuju. Sivin, Nathan 1995. Text and Experience in Classical Chinese Medicine. Knowledge and the Scholarly Medical Traditions, Cambridge, Cambridge University Press. Sivin, Nathan 2012. Therapy and Antiquity in Late Imperial China. In: Louis, P. N. M. A. F. (ed.) Antiquarianism and Intellectual Life in Europe and China, 1500–1800, Ann Arbor: University of Michigan Press. Wang, Shuhe 王叔和 1929–1936. Wangshi Maijing 王氏脈經, Shanghai, Shangwu yinshuguan. Xie, Guihua 2005. Han Bamboo and Wooden Medical Records Discovered in Military Sites from the North-Western Frontier Regions. In: Lo, C. C. A.V. (ed.) Medieval Chinese Medicine:The Dunhuang Medical Manuscripts, London: RoutledgeCurzon. Yang, Yong 楊湧. 2016. Xianqin Qin Han shiqi de yiliao shi yanjiu 先秦秦漢時期的醫療史研究. Ph.D Wuhan University. Yang, Yong and Miranda Brown. 2017. The Wuwei Medical Manuscripts: A Translation and Study. Early China, 40, 241–301. Zhangjiashan Ersiqi Hao Han Mu Zhujian Zhengli Xiaozu, 張家山二四七號漢墓竹簡整理小組 2006. Zhangjiashan Han mu zhujian (ersiqi hao mu) 張家山漢墓竹簡(二四七號墓) : (釋文修訂本), Beijing, Wenwu.

472

22 MATHEMATICS

KARINE CHEMLAMATHEMATICS

Karine Chemla1

Two types of sources for the history of mathematics in early China In the twentieth century, hundreds of Chinese manuscripts were discovered in the sands of the northwestern part of China as well as in tombs and pits that date from the last centuries bce, unless they had unfortunately to be bought on the antiquities market. Among them, several book-length documents are devoted to mathematics. These new sources have brought about a historiographical revolution in the history of mathematics in early China. They shed light on periods for which until recently we had no related evidence.They also provide a historical background against which we understand in new ways later mathematical books, which, by contrast to these manuscripts, were handed down through the written tradition and had until recently been the only source for historians of mathematics. To this day, only two of these mathematical manuscripts have been published. Writings on Mathematical Procedures (Suanshu1 shu. Hereafter, I abbreviate the title as Writings) was the first manuscript of that kind discovered (Peng 2001; English translation in: Cullen 2004; Dauben 2008). Its title is written on the reverse of one of the 190 bamboo slips composing the document. Writings was found during the winter 1983–1984 at Zhangjiashan (Hubei province), in tomb no. 247. The archaeologists have suggested that the tomb occupant was an official active at a local (county) level in the bureaucracy (Group of editors 2001). One of the other manuscripts found in the same tomb contains a set of administrative regulations, whose title, Statutes and Edicts of Year 2 (Er nian lü ling), is usually interpreted as referring to the second year of the Empress Lü’s reign (r. 187 bce–180 bce), that is, 186 bce (this is discussed in Barbieri-Low and Yates 2015: 72, and fn 9). This year is also the last one occurring in the calendar excavated from the tomb. Accordingly, it is generally assumed that the tomb was sealed around that year. Writings thus testifies to a material connection, in ancient China, between the practice of mathematics and the imperial bureaucracy, in particular the practice of administrative regulations. Its text confirms this connection, since Writings contains many problems and data reminiscent of Statutes and Edicts of Year 2, and also of administrative regulations for the Qin dynasty discovered at Shuihudi (also in Hubei province), Qin Statutes in Eighteen Domains (Hulsewé 1985; Peng Hao 2001, Forthcoming; Zou Dahai 2005, 2007). Notably, Writings quotes a regulation concerning grains, whose text can also be found in the latter set. We return to this quotation later. 473

Karine Chemla

The second book-length mathematical manuscript recently published, Mathematics (Shu1), was part of a set bought in two lots, in 2007 and 2008, on the antiquities market. The editors argue that it was composed not later than 212 bce, thus dating it from the Qin time period. Its title is also on the reverse of one of the 219 slips on which it is written (Xiao 2010; Zhu and Chen [gen. eds.] 2011). Similarly, in the same set, which arguably came from a single tomb, there was a manuscript of administrative regulations. If the common origin of the two manuscripts can be taken for granted, then, for the same reasons, Mathematics also attests to a physical relation between mathematics and the imperial bureaucracy, that is, more specifically again, the practice of regulations. We examine this connection in greater detail later. These two manuscripts, from the Qin and the Han time periods respectively, attest to another important fact. They prove that “mathematics” was an actors’ category in ancient China, and they embody one conception of the topic at the time (among possibly others). Their titles designate their content from the viewpoint that it brings mathematical procedures (shu1) into play, in general, in relation to mathematical problems, and that the related practice of mathematics makes use of counting rods (suan). These two features also characterize the earliest extant Chinese mathematical books that were handed down through the written tradition. The title of The Nine Chapters on Mathematical Procedures (Jiu zhang suanshu2, abbreviated hereafter as The Nine Chapters) also refers to mathematics in this way. Like the manuscripts, the book was mainly composed of mathematical procedures, also in general in relation to problems, and refers to the use of counting rods. Qian Baocong (1963, vol. 1: 83–84) held the view that The Nine Chapters had been completed in the first century ce. Although his view has recently been contested, I still concur with Qian for at least one more reason than those he put forward (Chemla, in Chemla and Guo 2004: 201–205, 475–478). Until the beginning of the Common Era, the highest measurement of capacity used in administrative practice was the dan. During Wang Mang’s interregnum (9–23), the name of this unit was changed into hu, and this name remained after the reestablishment of Han rule. The key fact here is that The Nine Chapters only uses hu. In what follows, we return to these two measurement units, and we will see that this is not a minor point. In fact, it relates to the issue of the official management of grains, as does the following important clue:The first known historical mention of the title of The Nine Chapters occurs on a standard vessel issued by the Grand Minister of Agriculture in 179 (Guo, in Chemla and Guo 2004: 57). This high official was in charge of managing the grain levied as tax and used to pay salaries to officials, and he was also responsible for enacting correct standards for capacity measurements units (Bielenstein 1980: 43–47).We thus once again encounter the same connection between mathematics and the bureaucracy as was mentioned earlier. We will see the crucial meaning of this connection when we address the topic of grains. The exact date of completion of the other book from this period that was handed down through the written tradition, The Gnomon of the Zhou (Zhou bi), is also disputed. Qian Baocong (1963, vol. 1: 3–4) considered that it was completed in 100 bce, and Cullen (1996: 153–156) at the beginning of the first century ce. I will present here my own view, which lies in between. Procedures also play a key role in it, and counting rods are used in computations. However, this last book differs markedly from those presented so far, since it deals with a specific topic, presenting only mathematical knowledge related to cosmography and calendar making (Cullen 1996 gives an English translation). From another perspective, The Nine Chapters and The Gnomon of the Zhou differ from the manuscripts in that all their ancient editions also include ancient commentaries (Martzloff 1997: 123–136). Such commentaries yield important historical evidence, since they allow us to observe how as early as in the third century ce, readers perceived these ancient texts and 474

Mathematics

interpreted them. Ancient commentaries were selected in the seventh century, in the context of the preparation of an imperially commissioned anthology of Ten Mathematical Canons (Suanjing shi shu), placed under Li Chunfeng’s (602?–670) responsibility. The Nine Chapters and The Gnomon of the Zhou, which, already in their ancient commentators’ eyes, were considered “canons” (Chemla 2008), were the oldest two in the anthology. Liu Hui’s commentary, completed in 263, was chosen for the former, whereas Zhao Shuang’s commentary, from the third century, and Zhen Luan’s (ca. 570) subcommentary were selected for the latter. Li Chunfeng and his colleagues edited these texts and also composed a subcommentary on them. (Chemla and Guo 2004 gives an edition and French translation of The Nine Chapters and all these commentaries; on these writings, see Li 1990; Guo 1992). Ten Mathematical Canons was presented to the throne in 656 (for a critical edition, see Qian Baocong 1963.) Together with two auxiliary writings, these canons were immediately afterwards used as textbooks in an official mathematical teaching given in the “School of mathematics,” an institution based in the “State Academy Directorate” (Volkov 2014). The crucial part that the project carried out under Li Chunfeng’s supervision, along with this institutionalization of mathematical education at the time, played in the transmission of mathematical texts of the past can be perceived through two remarks. Firstly, no ancient book survived that had not been used as a textbook at the time (or, at least, that ancient actors believed to have been). Secondly, no ancient editions of the two oldest canons survived that did not have the commentaries selected, and the subcommentaries composed, in the seventh century. What is more, although we have evidence that other commentaries had been composed before this time, none of them survived. The seventh century thus represents a key moment – and Li Chunfeng, a key figure – for the shaping of an ancient mathematical heritage in China. In the subsequent centuries, these books were regularly re-edited, and even printed, in the context of state enterprises, and they played a key part in specialized mathematics education. The bodies of knowledge and practices to which Ten Mathematical Canons attests more widely left their hallmark on experts’ mathematical activities in Chinese until at least the fourteenth century. This remark features the impact of mathematical books composed in early China on the centuries to come. However, recently Zhu Yiwen (2016 [2015]) identified a separate mathematical tradition in the seventh century subcommentaries on the Confucian canons. Zhu showed how in these annotations, when needed, subcommentators used mathematical knowledge to justify the correctness of ancient canons and commentaries. However, their mathematical practices differ strikingly from what the mathematical canons testify. Moreover, these commentators do not seem to have carried out specific research on mathematics, and in particular, none of them is credited with having composed a separate writing devoted to mathematics. In conclusion, we understand from our survey of sources, that at the present day all the documents with which we can explore mathematical knowledge and activities in early China shed light on mathematics as practised in, or selected by, specific state institutions. The same bias affects the evidence we have about the actors who were versed in mathematics (Cullen 2009). Cullen (2009) follows an approach to the mathematics of early China that focuses on the latter type of evidence. In this chapter, I will rather focus on what actual writings tell us about mathematical knowledge and practices in early China, and also about some of the social contexts in which mathematical activity was carried out.

Mathematical manuscripts Manuscripts like Writings differ from sources handed down through the written tradition in that they are objects that were once used by those in whose tombs they were found. Their material 475

Karine Chemla

features thus sometimes provide precious evidence. To explore this facet of documents of this kind, we will focus mainly on Writings, since it is the single manuscript available today that was found in regular excavations.

How material features of manuscripts document the context of their production The 190 bamboo slips composing Writings are divided into sixty-nine sections (and some damaged slips that we must leave aside) which are marked by the fact that their text begins with a heading and usually ends with an empty space left at the end of the last slip. Incidentally, Mathematics is also divided into sections, marked by a final empty space, but sections have no heading. A paleographic analysis carried out by Daniel Morgan has revealed that the manuscript of Writings had been written by two hands. Hand A wrote the main text of forty-eight sections, whereas Hand B wrote sixteen sections. Much more importantly, both hands appear in the main text of five other sections, and even alternate according to a fixed pattern in one of them. Finally, if we set aside one exception, only Hand B wrote the headings. Hands A and B were thus in the same space; the distribution of hands, and the types of text written, suggesting that Hand B was more advanced than Hand A and even apparently guiding Hand A. A study of such material features has led us to the assumption that the document was produced in a learning environment and reflected education practices (Mo (Morgan) and Lin (Chemla) 2016, revised into Morgan and Chemla 2018 [2017]). The manuscript appears to be notes written in relation to learning more than, strictly speaking, a book, and these notes were considered an appropriate object for the tomb occupant to take to the afterworld. From another angle, material features of Writings suggest that its content can be decomposed into two types of text. Most of the sections contain continuous writing, which only occasionally uses punctuation marks. They usually present mathematical problems and procedures. Other sections contain instead sequences of clauses that all share a similar format and are separated from each other by the systematic use of punctuation marks. These sections can be defined as numerical tables, similar to our “multiplication tables” (Chemla 2016). They record tables for multiplying between powers of ten, or between measurement values, in the context of the computation of areas or volumes. They also include tables for computing with fractions as well as for converting, for instance, between units of weight or between capacity amounts of different types and different states of grain. Indeed, a significant number of the tables that Writings contains are connected with grains, some stating equivalences, others listing related sequences of procedures, like rules of three. Similarly, the text of Mathematics can be divided into continuous writing and tables. The types of tables that this manuscript contains are quite similar to those in Writings, the greatest number of tables being, in this case as well, devoted to grains. However, the textual format that this manuscript as a rule uses for tables differs from that in Writings. The clauses are written in cells that are created in a set of contiguous slips (placed vertically) by means of empty spaces running across the slips and by the cords binding the slips together. These two devices thereby shape horizontal registers (sometimes three (upper, middle and lower), and sometimes more). A photograph of some contiguous slips of another (not yet published) mathematical manuscript, Mathematical Procedures (Suan shu2), written before 157 bce and newly discovered in tomb no. M77 at Shuihudi (Hubei province), also shows the use of the same textual format for writing tables of multiplication or procedures for conversion (Hubei sheng wenwu kaogu yanjiusuo et al. 2008; Chemla and Ma 2011). By contrast, the mathematical books handed down from the Han period contain only very few tables (e.g., a single table, devoted to grains, in The Nine Chapters, and a single table of solar shadows 476

Mathematics

at specific moments of the year in The Gnomon of the Zhou), and they never show the use of punctuation to write them down. In addition to these rare numerical tables, mathematical books handed down also contain continuous text, which, as for manuscripts, is commonly composed of mathematical problems and procedures. To illustrate these elements of text, let me quote one paradigmatic set from Mathematics (Zhu and Chen [gen. eds.] 2011: 25), which I chose among problems and procedures dealing with grains: A granary has a width of five zhang, a length of seven zhang and a height of the heap of two zhang. [The name of the dimensions points out that its shape is that of a cuboid.] Suppose there is unhusked grain (millet) in it, in such a way that it fills the granary up to the highest level. One dan [some scholars read shi] of unhusked grain (millet) occupying two chi seven cun, one asks the amount, respectively, of the chi of the volume of the granary and the unhusked grain (millet) it contains (rong). It is said: the chi of the volume are seventy thousand chi, and it contains twenty-five thousand nine hundred and twenty-five dan twenty-five twenty-sevenths dan. Procedure: width and length being multiplied by each other, one further multiplies this (result) by the height, hence the chi (of the volume). One takes two chi . . . [the slip on which the final part of the procedure was written is missing]. Here, zhang, chi and cun are measurement units of decreasing size, in a decimal system for length (1 zhang = 10 chi and 1 chi = 10 cun). In early China, these units were also used in a specific way to denote volume. This is the case when the outline of the problem states a given quantity of millet “occup[ies] two chi seven cun”. In fact, The Nine Chapters contains exactly the same convention of using measurement units for length, and the commentators explain the meaning when this use first occurs. We can thus establish that “two chi seven cun” meant a volume equal to that of a cuboid with a square basis of 1 chi, and a height (or a depth) of “two chi seven cun”. The remark illustrates why the commentators’ explanations are useful for the modern reader. It also highlights an element of continuity between manuscripts and books handed down. The other measurement unit used in the outline of the mathematical problem (the dan, or shi) is less straightforward to interpret. Here, the dan that occurs in the text could have been the highest measurement unit of weight at the time. It could also have designated the highest measurement unit of capacity, which, in administrative practice at the time, bore the same name as the highest unit for weight. However, here dan is none of these. It designates yet another measurement unit that was used in the context of managing grains and had not been so far identified. Sketching the argument will illustrate how mathematical manuscripts complement regulation texts and administrative texts and so enable us to understand in a new way the official system of grains at the time, and in particular the measurement units used in this context (Chemla and Ma 2015 [2014]). The description will also show how mathematical knowledge is at the core of the system.

How manuscripts clarify the relationship between one form of mathematical activity and regulations Mathematical manuscripts contain three types of evidence that make an essential contribution to help us better understand the administrative construction of the system of grains in Qin and early Han dynasties. 477

Karine Chemla

First, as was mentioned earlier, Writings quotes a regulation about grains that has a parallel in the Qin Statutes. This regulation involves six types of grain: millet (the main crop in northern China), rice (the main crop in the south), barley, and hemp, soy, and mung beans. In addition, for each of these grains the regulation mentions different states. Here, a second type of evidence comes into play: the formulation of procedures in both Writings and Mathematics yield numerical values and variations in the names of the states of grain, which proves essential to identify these different states. The key point to note here is that in these mathematical manuscripts, the regulation occurs with other documents with which it was used and which hence help interpret its text. We thus understand that for both millet and rice, the regulation introduced a sequence of different states, obtained from one another by applying different types of technical operations to the grain. Moreover, the regulation defined these states by stating quantitative relationships between them. For example, the sequence deriving from millet started from the plant just harvested and included four different states of the grain: unhusked grain (su), coarsely husked grain (mi), finely husked grain and highly finely husked grain. The ratios between the quantities of these various states of the grain that derive each from the previous one using a technical operation, as given by the regulation, are the same as the ratios between the numbers in the sequence 50:30:27:24. In this way, the regulation defined these states of the grain relatively to one another, using simple numbers. Following exactly the same pattern, the sequence deriving from rice that the regulation gave started from paddy stalks and included this time three states: unhusked grain (here, paddy rice, still su), coarsely husked grain (rice, still mi), and highly refined rice. The ratios between the quantities deriving each from the previous one, using a technical operation applied to the grain, are in this case represented by the sequence of numbers: 6:3:2. Mathematical knowledge is thus put into play in the very definition of the states of grains. But there is more. In the rules of three, presented in Mathematics, Writings and also in The Nine Chapters, the numbers 50, 30, 27, 24 (resp. 6, 3, 2) are used to compute equivalent quantities of different states of the same grain. In the manuscripts, these simple numbers derive from the Qin regulation by reworking the quantities as stated there (using measurement units). By contrast, the single numerical table that The Nine Chapters contains gives the relationship between the different states of grain using directly abstract numerical values adapted to these rules of three. In this sense, this table corresponds to the Qin regulation but presents the relationships between grains using a different type of numbers. The numbers governing the exchange between the same two types and states of grain have nevertheless the same ratio in The Nine Chapters and in the manuscripts, thereby displaying a first element of continuity between them in the management of grain. However, the type of numerical values used in the regulation reveals a key point. For both millet and rice, the regulation starts with a weight of 1 dan for the plant just harvested (this is the highest measurement unit of weight) and, for each state, tells the quantities of grain deriving from this amount using capacity measurement units. It thus appears that in each case, the quantity of the state of grain we translated as “coarsely husked grain (mi)” deriving from this weight (1 dan) of the plant is 1 dan (however, this time as a capacity unit, that is, the highest unit in the system of capacity). This fact has several interesting consequences. First, it suggests an explanation for why the highest measurement units for weight and capacity used in administrative practice have the same name. The link is established through grains. This is but one way of perceiving a fact essential in early China regarding metrology: grains played a key part in the definition of the official measurement units. Second, even though in the sequences deriving respectively from millet and rice all the other states have different quantities, for the state mi the quantities are constantly 1 dan. This state in fact allowed practitioners to establish a bridge between the various types of grain: for all 478

Mathematics

the grains, 1 dan (capacity) of mi had the same value (see the discussion that follows). Mi thus appears to have been the pillar of the whole system of grains: it related dan as weight and dan as capacity, and it also brought systems of states of grain deriving from different plants into relation. In The Nine Chapters, the structure of the table corresponding to the regulation is not manifest. Moreover, The Nine Chapters reflects a situation, established during Wang Mang’s reign, where hu has been substituted for dan as the name of the highest capacity unit. However, the insight given by the manuscripts allows us to see that the system of grains described in The Nine Chapters has the same structure. In the same way, its pillar is still the state mi, common to all types of grain. The key part played by the state mi for all grains is confirmed by another key feature of the whole system, which the mathematical writings highlight.This brings us back to the interpretation of the measurement unit also called dan in the outline of the mathematical problem quoted earlier. To understand what this “dan” refers to, we need the third type of evidence that mathematical manuscripts provide about grains, that is, the tables stating relationships between different types and different states of grain, and also different types of measurement units. For instance, Mathematics gives a table of the various capacity amounts of different types, and different states, of grain corresponding to a weight of 1 dan. This table allows us to disprove the idea that the unit dan in the granary problem corresponds to a weight.The volume stated in the problem for 1 dan (One dan of unhusked grain (millet) occupying two chi seven cun) further implies that this dan is not a capacity value (1 dan as a capacity unit has a different volume). Another numerical table in Mathematics lists amounts (expressed in volume units) of grain corresponding to this third meaning of the unit also called dan. The Nine Chapters likewise provides evidence showing that the same phenomenon affects the unit now called hu. Although in some cases hu clearly refers to a capacity unit, in other cases, exactly in the same way as dan in the manuscripts, the unit hu cannot be interpreted as capacity.This simply derives from the fact that in these other uses of hu, the volumes denoted differ from the volume of the capacity unit hu. In fact, these other values for the volumes are the same for dan in the manuscripts and for hu in The Nine Chapters. What, then, do these “dan,” and “hu,” refer to? The solution to the puzzle is given by the seventh-century commentator on The Nine Chapters, Li Chunfeng. We thus see again the important historical evidence provided by commentators. Li Chunfeng states that in this other use hu is in fact a measurement unit of value.The same interpretation can be shown to hold for dan. Accordingly, it appears that, in fact, the regulation quoted in Writings states, for any given type and state of grain that could be used in the bureaucracy, the capacity amount of this type and state of grain that was needed to make a unit of value of 1 dan. Barley as well as hemp, soy and mung beans were also inserted in the system so as to fit this structure. With respect to the unit of value, the state mi for all grains again played a special part: it was the only state for which 1 dan capacity corresponded to 1 dan value. The Nine Chapters and its commentators further evidence the fact that for each type and state of a given grain, the related hu unit of value was embodied by a special vessel also called hu. Excavated vessels show that vessels measuring value must have existed for the Qin and Western Han periods (Chemla, in Chemla and Guo 2004: 201–205; MA and Chemla 2018). In fact, standard vessels that embodied the official highest unit of capacity (whether this unit was dan [Qin and Western Han time periods] or hu [Eastern Han time period]) were also vessels measuring the unit of value for the state of grain mi. This state of grain thus also appears to have been the basis for defining the official vessels measuring capacity. Discussions of the dimensions and volumes of these vessels regularly occur in mathematical documents, where they represent an important topic. In another respect, the rules of three contained in all mathematical manuscripts and in The Nine Chapters could be used to measure units of value for any type and state of grain, using any vessel available to the user. 479

Karine Chemla

This long development allows us to draw important conclusions. First, taken in combination with other types of documents, mathematical writings disclose a fact that had remained so far hidden: the sectors of the Qin and Han bureaucracies that were managing the resources of the state (that is, primarily, grains, which were the main medium for levying taxes and also played a key role in the payment of salaries to officials) used units of value (called, respectively, dan and hu) based on capacity. Second, in the same way as the meaning of the dan capacity unit derived from the meaning of the dan weight unit through grains (more precisely, through the state mi of any type of grain), the meaning of the dan (or hu) value unit derived from the meaning of the dan (or hu) capacity via grains (again, specifically, using the state mi.) For value, too, the meaning of the unit was rooted in grains, and it was related with other measurement units via grains. This had led us to suggest calling this state of grain “standard mi”. Finally, whether we look at the recently excavated manuscripts or at The Nine Chapters, mathematical writings display the same system for grains, with an identical structure based on the universal state mi and a similar use of units of value. The system has remained unnoticed so far by modern historians. Bringing it to light reveals a great deal of continuity in the administrative regulations of grains throughout the first centuries of the Chinese empire. Some changes in the measurement units used can nevertheless be perceived (notably, but not only, the change of dan into hu). This invites further research to explore the exact nature of this continuity. For all these conclusions, the newly discovered mathematical manuscripts have provided crucial evidence and have shed new light on the received texts, allowing us to interpret them in a new way. These conclusions illustrate more generally the importance of mathematical knowledge and practice for the state bureaucracy. For this, too, a great deal of continuity can be observed. On the one hand, as we have seen, the structure of the system of grains that the state bureaucracy used was permeated with mathematics. On the other hand, mathematical writings and administrative documents have proved to be intimately connected to one another. In the case of the Qin regulation, the elucidation of administrative practice with grains required taking mathematical writings into account. In the case of The Nine Chapters, the converse held true. The statement of the commentator Li Chunfeng that allowed us to elucidate the nature of the value unit hu comes from a technical monograph that Li Chunfeng wrote on measurement units for an official history. Since the value units dan, or hu, were central for the economic activities of the bureaucracy, the clarification in both contexts is essential for economic history. The tombs where mathematical manuscripts were found, and the other objects they contained, illustrate a connection between mathematics and the bureaucracy, which the content of the manuscripts thus confirms. In the previous section, we saw how material features specific to manuscripts provided historical evidence about the contexts of mathematical activity. In this section, we have seen how evidence yielded by manuscripts, for time periods for which we previously had no sources handed down through the written tradition, also proved important in providing insight about a social role played by mathematics at the time. Finally, manuscripts provide a historical depth that has allowed us to read a text handed down through the written tradition in a new way. From this perspective, we have been able to establish a strong continuity in what appeared to have been a social role devoted to mathematics. We could also perceive a fair amount of continuity between the two types of mathematical documents that we had distinguished earlier. Mathematical manuscripts and The Nine Chapters contain the same type of administrative data regarding grains, reflect the same management and contain similar mathematical tools, essential to handle the calculation of equivalent quantities of different grains. In what follows, we will keep exploring continuities and also discontinuities 480

Mathematics

between manuscripts and books handed down, as they can be perceived from the viewpoint of mathematical knowledge and practices.

Manuscripts and canons: similarities Topics The continuity between mathematical manuscripts, on the one hand, and between manuscripts and The Nine Chapters, on the other, is manifest when we examine further the topics that these writings dealt with. First, with respect to numbers and quantities, except for the replacement of dan as a capacity unit by hu at the beginning of the first century ce, all Chinese mathematical documents use the same measurement units. Moreover, both types of documents introduce fractions similar to those we use today, that is, as a pair consisting of a numerator and a denominator. Note that this concept is not found elsewhere in the world at the time. Another concept of fraction was used around the Mediterranean Sea from the beginning of the second millennium until at least the seventh century, which, as such, has disappeared from present-day mathematical practice – there only remains an interest for mathematical questions raised by computation with this other type of fractions (Benoit, Chemla and Ritter 1992). Fractions as a pair of numerator and denominator occur in the earliest known Chinese mathematical documents as results of division (Chemla 2013). More precisely, these results are given as integral amounts of units, increased by a fraction, whose numerator is the remainder of the operation and whose denominator is the divisor. In this context, fractions were thus systematically smaller than the last unit used to express the integral part. Division also played a key role in the procedures that allowed practitioners to carry out arithmetical operations on fractions. Mathematics,Writings and The Nine Chapters all contain such procedures. If we now turn to procedures more generally, the mathematical manuscripts, and The Gnomon of the Zhou to a certain extent, as well as The Nine Chapters to a lesser extent, devote a relatively significant amount of space to procedures carrying out division between all types of quantities (think, for example, of the division of an area by a capacity amount). The operation of division appears to have been perceived as difficult, and especially so in the last centuries bce. We return to this issue later. Let us note for the moment that the same conclusion emerges from considering all mathematical traditions worldwide at the time. Mathematics, Writings and The Nine Chapters further share common procedures for unequal sharing (which solve problems like “how to divide something between officials who have different ranks in the bureaucratic hierarchy, and will accordingly receive different shares proportional to their ranks”).These writings also all devote some space to rules that survived in mathematical practice worldwide until the seventeenth century under the name of “rules of false double position” (Chemla 1997). Finally, we find in these three writings computations of volumes related to civil works to be carried out, or to granaries – as in the problem translated earlier.The similarity in topics and approaches confirms the strong historical connection between all these texts that we have noted here. In contrast, The Gnomon of the Zhou appears at first sight as different. We return to this issue later. In these respects, manuscripts are alike. However, manuscripts also show differences that await explanation. For instance, Writings does not deal with the computations of areas, except that of the rectangle, whereas both Mathematics and The Nine Chapters devote several problems and procedures to the determination of areas for different geometrical shapes in the plane and the sphere. Other as-yet-unpublished manuscripts likewise seem to devote significant space to areas 481

Karine Chemla

(Han Wei 2013). The measurement of areas was essential to determine the amount of tax to be paid in grain for a cropland that the state had allotted to a household. All manuscripts, including Writings, have problems addressing this issue. Manuscripts differ on another topic. The right-angled triangle occurs in a problem of Mathematics, whereas it remains invisible in Writings. On the other hand, it is the topic of a whole chapter in The Nine Chapters, and it also plays an important role in The Gnomon of the Zhou. We also return to this issue. In conclusion, if the mathematical topics dealt with, and the procedures used, in the manuscripts present significant overlap with The Nine Chapters, only some of them appear in The Gnomon of the Zhou. One could argue that the procedures that are common to all these texts are the most fundamental ones: arithmetic operations, computations with fractions, rules of three or unequal sharing. Further, the (comparatively more significant) bodies of knowledge and practice shared by the manuscripts and The Nine Chapters might represent the mathematics available in some sectors of the bureaucracy, especially those related to the management of the state.

Practices and epistemological values However, continuity does not occur only at the level of topics dealt with. It is also noticeable at the level of mathematical practice. We have already noted that in the manuscripts like in The Nine Chapters mathematical knowledge takes the form of problems, procedures and also tables. Problems used as paradigms and related procedures represent one of the main supports of mathematical activity. However, neither The Nine Chapters nor the manuscripts refer to any visual auxiliary. In these respects, too, The Gnomon of the Zhou is different. Although it presents procedures, it is not organized around mathematical problems. Moreover, its opening part refers to a graphical process (Chemla 2005, 2010a;Volkov 2007). The Nine Chapters and Writings share another element of practice, which is more unusual (Chemla 2006). These two documents regularly give texts of mathematical procedures outside the framework of any problem. These texts are, however, related to mathematical problems, for which specific procedures are also given. In other words, in the two writings, the operations carrying out the task that the statement of the mathematical problem requires to solve correspond to two texts of procedure of two types– a higher-level text and a lower-level one. This is, for example, the case when Writings presents a procedure aimed at computing the unit price for a specific unit and a given commodity, when the price paid for a quantity is known. Why, one wonders, should one give two texts for a procedure? We get insight into the reasons for this, if we observe that in The Nine Chapters and Writings, the higher-level texts of procedures share key features. By contrast with the lower-level texts, they cover different cases of the same problem and cases that require the use of different lists of operations. Accordingly, the higher-level texts usually contain conditionals. Specific textual features and a specific handling of the text allowed the practitioners to derive from these texts different lists of operations, depending on the case dealt with (Chemla 2015 [2016]). By contrast, the text of lower-level procedures can be followed in the order in which the operations are listed. Both texts of procedures embody forms of writing in a general way; however, the generality of the upper-level procedures extends more widely. Higher-level texts have another feature in common: unlike the lower-level texts, which refer directly to operations using arithmetical terms, the higher-level texts contain terms referring indirectly to operations by stating a reason why the operation should be carried out. For instance, in The Nine Chapters, one of these texts designates operations by the use of a heavily loaded philosophical term: “make communicate tong.” The term refers to specific operations by pointing 482

Mathematics

out the aim to be fulfilled, that is, to allow mathematical entities to enter into communication (to enter into the same operations) by transforming them. The term occurs in different mathematical contexts. The operations it designates change in relation to the context. However, they share the same formal goal of enabling the entities to which they are applied to enter into the same operations. The term occurs in the ancient “Great Commentary” of the Book of Changes. Its occurrences convey a reflection on the ability of entities to enter in relation to one another through transformations to which they lend themselves. The use of the term in the context of mathematics points out that mathematical situations are read as falling under this general pattern. As a result, the texts of higher-level procedures give reasons for the correctness of the procedures they state. Evidence suggests that these texts are a locus where we find the practice of abstraction as an actors’ category in early China. We saw earlier how mathematical activity was intimately related to administrative practice. We now see that this description fails to capture another, more theoretical, side of mathematical activity which the mathematical writings of early China amply evidence. Both Writings and The Nine Chapters testify to this theoretical practice, albeit with specificities in each case. What was just described displays a continuity of mathematical practice between a manuscript and a canon also at this level. However, apparently Mathematics does not attest to this practice of abstraction. The differences that manuscripts display in this respect await further study. In conclusion, we can perceive certain continuity in the epistemological values to which The Nine Chapters and Writings testify, while Mathematics reflects other choices. Finally, mathematical texts all refer to the use of the same artefact to compute, that is, counting rods. The earliest description of the physical appearance of these rods is due to the scholar Liu Xin (46 bce–23 ce) and was inserted in the “Monograph on the Calendar and Pitch-pipes” of the official history The History of the Han, composed by Ban Gu (32–92) and his sister Ban Zhao (ca. 45–ca. 116) (Volkov 2001). Mathematical texts also yield clues about how these rods were used. For instance, they all use the verb “to place” to prescribe representing a numerical value on a calculating surface using counting rods. They also regularly indicate positions in which these values should be placed like “one puts to the left . . .”. However, none of them includes illustrations of the computation. These common features do not mean that the way of using the rods was uniform throughout the period considered. On the contrary, the fact that manuscripts abound in tables, including tables for multiplying between powers of 10 (found in Writings and the manuscript from Shuihudi Mathematical Procedures), and that these tables disappear from the later canons suggests a transformation in modes of computation. This remark leads us to consider now more systematically differences between our various witnesses.

Manuscripts and canons: differences A key difference in bodies of knowledge and related practices Arguably, the evidence offered by the newly found manuscripts indicates that a major transformation took place between the second century bce and the first century ce with respect to arithmetical operations, and in particular division. Moreover, in parallel, arithmetical operations like division appear to have been a basis on which an important theoretical work was carried out. Hence, once again, manuscripts and canons like The Nine Chapters not only reflect use of mathematics in administrative practice, they also show directions taken by theoretical work. Against the backdrop of continuity examined earlier, the manuscripts now allow us to perceive a historical process. 483

Karine Chemla

One of the key clues that reveal this transformation is given by a set of changes related to the use of the term chu. In the manuscripts, this verb only designates subtraction, including sometimes repeated subtraction. On the other hand, the manuscripts do not testify to any prescription of a division that would use a single verb. Only complex expressions refer to this operation, chu occurring within some of them, arguably also with the meaning of subtraction or repeated subtraction. In correlation with these features, division seems to have been perceived at the time as a tricky topic, not least on account of the space that is devoted to it in Writings. In the divisions that occur in the manuscripts, operands are quantities expressed using a sequence of measurement units, and sometimes a fraction of the last unit (or a fraction alone). Their results are of the same nature, and the quotients are progressively obtained as measurement values, along a sequence of measurement units, and sometimes end with a fraction. The contrast between this situation and what we find in The Nine Chapters offers a measure of the work carried out. In the canon, division is now sometimes prescribed by a simple verb, and this is precisely the verb chu. The meaning of chu thus appears to have undergone a shift. The key point is that the change of terminology appears to have gone along with a radical change in the execution of division.The execution is now based on a place-value decimal system, similar to ours.The similarity holds for two essential features. First, 10 is the basis of the construction of both number systems. Second, positions are used in the same way. When we write 123, 1 means a hundred, and 2 twenty, in relation to their positions in the array of digits. Similarly, although at the time of The Nine Chapters digits were written with counting rods on the calculating surface, the position of digits within arrays of digits was used in the same way to write numbers. Note that by contrast with the manuscripts, The Nine Chapters thus appears to offer the first known pieces of evidence for the use of a place-value decimal system of this kind. Accordingly, the execution of division now relies on numbers separated from their measurement units and yields quotients digit by digit, and sometimes a final fraction. Only at the end of the execution are quotients related to measurement units. Arguably, division was also executed on the calculating surface on the basis of a vertical array of three positions, in the context of a new practice of the instrument; we return to this point later (Li and Du 1987: 15–19). Related to the appearance of this new terminology and the new execution for division, it is noteworthy that in The Nine Chapters, a whole array of phenomena related to the use of the term chu, and also to the use of positions, can be found. Manuscripts contained procedures for carrying out the operation of square root extraction. In these early writings, the operation was not compared to division. In contrast, in The Nine Chapters the terminology of square root extraction changes and now highlights how square root extraction is a specific type of division chu. In correlation with the change of terminology for square root extraction, the execution on the calculating surface also changes, likewise showing how square root extraction is a form of division chu – we return to this point later. This is the first element of the shaping of a structure in a set of operations. In addition, three new operations of root extraction (“circular”, cube, “spherical”) are introduced. “Circular” and “spherical” root extractions rely on square and cube root extractions whose modes of prescription systematically point out that they both derive from chu. Similarly, quadratic equations are introduced as arithmetical operations, and whether we look at them from the viewpoint of the terminology used to prescribe them, from that of the names of their operands or from that of their execution, we observe they are shaped as a type of operation dependent on the division chu. The chapter of root extractions and equations indicates how knowledge previously available is reshaped and expanded in a way that relates to the new division chu. Another chapter also appears in The Nine Chapters, which can be interpreted as a new theoretical development based on the division chu, namely, the solutions of systems of linear equations (fangcheng, which I translated as “measures in square”) (Li and Du 1987: 46–50; Chemla 2000). 484

Mathematics

All these topics, which The Nine Chapters relates to the division chu, are dealt with in the canon, but they do not occur in the manuscripts (or, if they do, as is the case for square root extraction, they are not dealt with in the same way). This body of knowledge thus appears to have been the result of an intense theoretical work based on the arithmetical operations. At the same time, the shift that I have sketched could perfectly account for why the numerous tables that might have been related to older modes of executing multiplication and division disappear from The Nine Chapters, and more generally from the canons (Chemla 2018). Interestingly, the transformation outlined is perceptible not only from the viewpoint of mathematical knowledge but also from the viewpoint of mathematical practice. Indeed, all the new developments just mentioned are related to a use of “position” on the counting surface.We have already noted that positions, in horizontal arrays, are used to write numbers. Positions, in vertical arrays, are also used to carry out division and root extraction and to find the roots of quadratic equations. Positions, in horizontal and vertical arrays, are likewise used to write down systems of linear equations (fangcheng) and solve them, the procedure relying heavily on the arrangement of data into positions. In fact, all the operands of these operations, as well as the specific terms occurring in the process of execution, correspond to positions. Positions have precise mathematical meanings, and they are also the basis for the execution of the algorithms. Finally, positions play a key part in a theoretical practice that takes shape on the basis of the calculating surface and that explores relationships between the different operations on the basis of the relationship between their executions on the surface. Let me illustrate this assertion with an example.This is the point where we return to the execution of the division chu and related operations. Through this practice, the relation of opposition between the operations of multiplication and division appears as a relation of opposition between the processes executing them. Multiplication and division are both carried out on an array of three lines (see figure 22.1). Multiplier and multiplicand are in the lower and upper positions, respectively.Through the execution, the multiplicand is erased digit by digit, whereas the product is formed progressively in the middle row. In opposition to this, dividend and divisor are in the middle and lower positions, respectively. Throughout the process of division, the dividend is progressively diminished, whereas the quotient is formed digit by digit in the upper line. The observation of the events taking place in the three positions during the executions shows how the processes of execution of the two operations are opposed to each other. In exactly the same way, relationships of similarities are built between division and root extractions. Square root extraction and cube root extraction appear to be executed using positions and events that are comparable to positions and events in a process of division. In this sense, as in many others, the pair of opposed operations, division and multiplication, plays a role of reference for the executions of other operations. This remark suggests that this practice in mathematics is related to the exploration of more general cosmological ideas and their reflection within mathematics. By “cosmological ideas”, I mean here universal ideas about how processes in the natural, the cultural and the social realms unfold. In this case, mathematical practice seems to reflect a more general meditation on the part played by opposed and yet complementary elementary operations to account for how processes unfold.The practice just described illustrates how this reflection might have developed through the observation of the execution of operations on the calculating surface. Other types of evidence indicate that cosmological reflections also permeated the perception of rods or other calculating tools (Volkov 2001). Commentaries on Han canons likewise testify to cosmological concerns in mathematics (Chemla 2010b). Before the canons, we have no hint that this practice existed. However, there is evidence of the same calculating practice until at least the Song-Yuan period, and later authors also refer to the cosmological reading of operations on the calculating surface (Chemla 2018). As we have already mentioned, other mathematical traditions existed in China, and their culture of 485

Karine Chemla

Multiplication Upper/Full Middle/Empty Lower

1

5

7

2

3

5

7

2

3

1

5 5

2

3

1 2

5 3

1

5 2

3

Delete

Division

3

1 2

1 3

1

3 2

1 3

1

3 2

5 1 3

1

5 6

2

3

1

5 6 2

1 3

5 6 2

7 1 3

Fill

1

5

7

Full

2

3

7

1

7 1

7 1

1

3

1 2

7 1 3

1

3

1 2

1 3

Delete

Empty Full

Empty Full

1

/ Fill 1

1

Figure 22.1 The execution of a multiplication and a division, according to the Mathematical Canon by Master Sun. Digits written with rods are replaced by Hindu-Arabic numerals.

computation differed drastically, not least because they did not use any material tool for computation (Zhu Yiwen 2016 [2015]). In this sense, arguably The Nine Chapters testifies to the emergence of a new culture of computation in early China. Although, as in manuscripts, this culture still makes use of counting rods, it is characterized by its number system, its use of positions, and the related practice of working on processes of execution. In correlation with the emergence of this culture, we perceive theoretical work carried out in particular on the operation and execution of division. A fact that strongly suggests that the two changes are intimately related to one another is that the terms used in texts of procedures to refer to positions are all related to the terms designating the operands of division. In conclusion, let me emphasize a point. I have evoked several types of theoretical developments to which either both manuscripts and The Nine Chapters, or only the canon, testify. We could grasp some of these developments in the use of philosophical terms like “tong make communicate”. I have also suggested interpreting as a theoretical development the structure of a set of operations based on chu because of the terms used and the ways in which texts of procedures were formulated. I have further suggested that a form of theoretical practice which had cosmological dimensions was shaped on the basis of the calculating surface and that we had 486

Mathematics

to interpret the processes of computation as stating relationships. Early texts never explain these theoretical facets in a fully discursive fashion. We can only rely on clues in the documents to grasp them. And after all, it is exactly with the same working method that I established earlier a relationship between mathematical writings in early China and some sectors of the bureaucracy. In this context, I had emphasized how ancient commentators offered crucial evidence. Here too, noteworthy is the fact that ancient commentaries on The Nine Chapters also appear to read terms like tong or chu that are used in the canons as conveying meanings. Their commentaries testify to how they make sense of these terms. The evidence they thereby offer is not decisive, but it is essential to keep it in mind when we attempt to interpret clues in texts from early China.

A key difference in emphasis: the right-angled triangle and the mathematics of the calendar We have noted another important difference between the manuscripts and the two canons. Although knowledge about the right-angled triangle is not entirely absent from the manuscripts, as far as we can tell on the basis of the available evidence, it does not constitute a principal theme. In contrast, The Nine Chapters devotes a whole section to the topic, and it is also essential for The Gnomon of the Zhou, whose opening section consists of a dialogue, in which what we call the “Pythagorean theorem” is the main topic of discussion. Here, we need to be more precise. We usually associate the “Pythagorean theorem” with the figure of the right-angled triangle. The association also holds true in The Nine Chapters, where a right-angled triangle is designated using its three sides: gou and gu, respectively, the shorter and the longer sides that form the right angle of the right-angled triangle, and xian, its hypotenuse. However, in The Gnomon of the Zhou, the statement concerns gou and gu, and in the whole book the hypotenuse xian is never mentioned (Li Jimin 1993). In this canon, the shape that the “Pythagorean theorem” is about is, in fact, embedded in a rectangle, and what for us, and for The Nine Chapters, is the third side of the triangle, is, in the context of The Gnomon of the Zhou, the diagonal of the rectangle. Notwithstanding the clear relationship between the two canons on the topic, this remark underlines a fundamental difference. This remark brings us back to The Gnomon of the Zhou and its date. The observations developed in the previous section about the terminology of division yield essential evidence. In fact, in some parts of this canon chu is used to prescribe division, in exactly the same way as it is used in manuscripts, that is, in the context of complex expressions. In other parts, chu is used as a single verb to refer to division. Most interestingly, in three of these cases, either the third century commentator Zhao Shuang emphasizes that the part in question did not belong to the “original text” or there is an editorial problem (Chemla 2013). These remarks seem to suggest that a part of The Gnomon of the Zhou is definitely closer to the time when our manuscripts were produced, and probably, as Qian Baocong (1963) suggested, dates from around 100 bce, whereas another part is more recent, and probably dates from the early first century ce, as Cullen (1996) suggested (even though the more modern sections are not the same as those he suspected), or even slightly later. More generally, Cullen was perfectly justified to assume that the canon was probably the result of the compilation of various pieces of text. The canon illustrates mathematical knowledge that could be taught for cosmographical issues and calendar making. In the seventh century ce, it constituted the part of the curriculum taught in official mathematical education that was clearly related to these issues.We will examine its contents and practices from this viewpoint (for the astral sciences, see the chapter devoted to “astronomy” in this volume (Chapter 23)). 487

Karine Chemla

As in all the other mathematical writings from early China examined so far, the statement of the “Pythagorean theorem” that is inserted in the opening part of The Gnomon of the Zhou takes the form of a procedure. Moreover, the same passage refers to a visual process through which a diagram is cut and rearranged.The third-century commentator Zhao Shuang interprets this part of the text as an argument aiming at establishing the correctness of the procedure corresponding to the “Pythagorean theorem” (Chemla 2005). We will also note that this is the earliest piece of evidence that we have regarding the use of diagrams in the context of mathematical practice in early China. Later evidence is also found in the third-century commentaries by Liu Hui on The Nine Chapters and by Zhao Shuang on The Gnomon of the Zhou. In these other contexts, too, diagrams look like material objects that are cut into pieces and whose pieces are rearranged (Chemla 2010a). Moreover, diagrams there also serve the function of being a basis on which to establish the correctness of procedures. Arguably, in early China visual auxiliaries might have been introduced into mathematical practice in relation to the astral sciences. The Gnomon of the Zhou also refers to other graphical devices. In particular, the canon refers to a device embodying the cosmography “Heaven (like a) canopy gaitian” that it describes (Cullen 1996: 221–223). According to this geometrical model for the cosmos, heaven is considered to form a plane above that of the earth, rotating around a vertical axis. Seven concentric circles represent the paths of the sun at specific moments of the year. Moreover, what people see about heaven from earth depends on a range of visibility beyond which phenomena remain invisible. From early on, The Gnomon of the Zhou has been considered the main theoretical text attached to this cosmography. In the canon, the “Pythagorean theorem”, along with the rule of three, is associated with the introduction of this cosmography. It also provides a key tool for the use of mathematics in the context of the astral sciences. An instance thereof is particularly telling. According to the cosmography described in The Gnomon of the Zhou, when the sun is due east on the day of the winter solstice, it is out of the range of human sight. The cosmography, however, assigns it to a perfectly clear geometrical position. The canon asserts: On the day of the Winter solstice, when it is exactly East or West one does not see the sun. One thus looks for it using computation: the distance between the point below the sun and the Zhou (gnomon) is two-hundred fourteen thousand five hundred fiftyseven li and a half. (Qian 1963: 39–40, my emphasis) Clearly, here mathematical computation is asserted to be a substitute for observation, when the latter becomes impossible. This essential statement captures the perception of the role that The Gnomon of the Zhou assigns to mathematics. How is computation carried out in this context? According to the interpretation offered by the commentator Zhao Shuang, the cosmographical model is a basis on which one can develop a geometrical analysis of the situation. He thus identifies a right-angled triangle and how to apply the procedure corresponding to the “Pythagorean theorem” to it, so that computation yields the numerical value that could not be derived from observation. In his view, the cosmography, geometric shapes, and mathematical computations must thus be combined to allow mathematics to determine the positions of heavenly bodies. In this sense, geometrical diagrams are essential in the astral sciences. This is, however, not the only part that The Gnomon of the Zhou assigns to mathematics in the astral sciences. Constructing the calendar presupposes that one has selected numerical values for the following magnitudes: after how many days does the sun return to the same stellar background (or else, how many days in the year)? After how many days are the moon and the sun in 488

Mathematics

conjunction again (or else, how many days in the month)? How many of these months are there in a year? How many degrees does the moon travel daily eastwards? However, these values are not independent of each other. Some can be chosen as basic and determined using observations (and also mathematical procedures), and the others can be deduced from them. The Gnomon of the Zhou describes precisely how mathematical computations and geometric reasoning about the relationship of these cycles with one another allow practitioners to build a logically coherent set of values. Mathematics is thus shown to play a key role in the organization of systems of astronomical data. As far as we can tell, these bodies of mathematical knowledge, which are typical of the astral sciences and constitute the core of The Gnomon of the Zhou, are absent from the manuscripts, whereas one finds them in The Nine Chapters. One is thus tempted to assume that the canon The Nine Chapters might have represented a synthesis of bodies of mathematical knowledge that took shape and were cultivated in relation to various sectors of official activity (Chemla in Chemla and Guo 2004: 473–474).

A key difference in practice: the meaning of chapters in The Nine Chapters Finally, an interesting difference puts The Nine Chapters in opposition to all the other known documents. Mathematical knowledge in this canon is divided into chapters. One might be tempted to consider this as a minor and merely material detail. However, there are reasons to believe that this is a fundamental feature of the book (Chemla and Zou 2018). In Liu Hui’s preface to his commentary, the commentator relates this feature to an assertion found in a book highly valued in part of the Confucian tradition since the Han dynasty, The Rites of the Zhou (Boltz 1993). In the latter canon, indeed, mathematics is asserted to be one of the six arts to be taught to the scions of the elite. The book adds that “mathematics” there refers to “nine shu1”. For Liu Hui, the “nine chapters” are a “development of the nine shu1”. It is difficult to ascertain the meaning of shu1 in these two contexts, shu1 usually meaning “number” or “reckoning”, and in some contexts “procedures” (see my glossary of technical terms used in mathematics, Chemla and Guo 2004: 984–986). The first scholar whose commentary on The Rites of the Zhou is still extant, Zheng Xuan (127–200), mentions an interpretation of this expression using a list of items which strongly evokes the list of titles of chapters in The Nine Chapters. Most of the items of Zheng Xuan’s list coincide with these titles, although not all. Moreover, his order is not totally identical to that of The Nine Chapters as we read it today. In fact, closer examination shows that the titles of the chapters in The Nine Chapters are also more or less (I skip the detail of the argument) the names of the first procedure in each chapter. Further, commentaries offer evidence showing that in the commentators’ view, the placing of a procedure at the beginning of a chapter amounts to the claim that this procedure is the “origin yuan” of the procedures in this chapter. In other words, the other procedures derive from it. These facts suggest the interpretation that the chapters of The Nine Chapters each state a fundamental procedure which gives the related chapter its name and whose development takes form from the set of other procedures recorded in this chapter. Accordingly, one of the goals pursued by the commentators might have been what we find in the actual commentaries: showing how chapters gather procedures that derive in some sense from their opening procedure (Chemla 2008). If we accept this view, then the list given by Zheng Xuan can be interpreted as a list of fundamental operations in mathematics that for him referred to the essence of the teaching of the subject as meant by The Rites of the Zhou. This thus suggests an interpretation of the expression 489

Karine Chemla

“nine shu1” that occurs in this canon as “nine fundamental procedures”. Furthermore, the fact that during the Han dynasty there existed alternative lists implies that the identification of the “nine fundamental procedures” meant by The Rites of the Zhou was one of the goals of mathematical activity in the first centuries ce and received different answers, probably in contention with each other. In any event, The Nine Chapters as a book allows us to appreciate the mathematical work that led to the establishment of such a structure of mathematical knowledge. This is clearly work of a theoretical type. It attests to a valuing of generality in a specific way. It also evokes the type of abstraction that I have discussed earlier, with respect to the “higher-level” texts of procedures. Interestingly, The Gnomon of the Zhou records a dialogue between a master (Chen zi) and a would-be disciple (Rong Fang), in the context of which the master outlines what in his eyes are the characteristic features of superior knowledge in mathematics. In particular, the master emphasizes specific types of procedures in the following words: Among the procedures of the Way, those whose expression is simplified and whose uses are wide are those that are the most enlightening for knowing categories. To raise a problem related to a category and thereby to understand ten thousand situations, this is what is called ‘to know the way’. (Qian Baocong 1963: 24; see another translation in Cullen 1996: 177) This characterization might well fit with the texts of procedures that are placed at the beginning of the chapters in The Nine Chapters. Arguably the type of generality and abstraction they represent accounts more generally for the structure of the Canon, a structure in “chapters” that does not occur in the manuscripts so far known. These chapters in fact embody the various domains of mathematics on which later expert work in mathematics will concentrate. This is one of the ways in which the history of mathematics in early China shaped mathematical knowledge and practices for centuries to come.

Note 1 It is my pleasure to express my most sincere gratitude to Jeremy Gray for his close reading of this text and his comments on it. The research leading to these results has received funding from the European Research Council under the European Union's Seventh Framework Programme (FP7/2007–2013) / ERC Grant agreement n. 269804.

Works cited Barbieri-Low, A. J. & Yates, R. D. S. (2015) Law, State, and Society in Early ImperialChina: A Study with Critical Edition and Translation of the Legal Texts from Zhangjiashan Tomb no. 247, Leiden and Boston: Brill. Benoit, P., Chemla, K. & Ritter, J. (eds.) (1992) Histoire de fractions, fractions d’histoire, Basel and Boston: Birkhäuser Verlag. Bielenstein, H. (1980) The bureaucracy of Han times, Cambridge, GB: Cambridge University Press. Boltz, W. G. (1993) ‘Chou li 周禮’, in Loewe, M. (ed.) Early Chinese Texts: A Bibliographical Guide, Berkeley: The Society for the Study of Early China and the Institute of East Asian Studies, University of California. Chemla, K. (1997) ‘Reflections on the world-wide history of the rule of false double position, or: How a loop was closed’, Centaurus. International Journal of the History of Mathematics Science and Technology, 39: 97–120. Chemla, K. (2000) ‘Les problèmes comme champ d’interprétation des algorithmes dans les Neuf chapitres sur les procédures mathématiques et leurs commentaires’, Oriens Occidens: 189–234.

490

Mathematics Chemla, K. (2005) ‘Geometrical figures and generality in ancient China and beyond: Liu Hui and Zhao Shuang, Plato and Thabit ibn Qurra’, Science in Context, 18: 123–166. Chemla, K. (2006) ‘Documenting a process of abstraction in the mathematics of ancient China’, in Anderl, C. & Eifring, H. (eds.) Studies in Chinese Language and Culture – Festschrift in Honor of Christoph Harbsmeier on the Occasion of His 60th Birthday, Oslo: Hermes Academic Publishing and Bookshop A/S. Online at: http://halshs.archives-ouvertes.fr/halshs-00133034, www.instphi.org/Festschrift.html. Chemla, K. (2008) Classic and Commentary: An Outlook Based on Mathematical Sources, Berlin: Max-PlanckInstitut für Wissenschaftsgeschichte. Chemla, K. (2010a) ‘Changes and continuities in the use of diagrams tu in Chinese mathematical writings (3rd century–14th century) [I]’, East Asian Science,Technology, and Society: An International Journal, 4: 303–326. Chemla, K. (2010b) ‘Mathematics, Nature and Cosmological Inquiry in Traditional China’, in Dux, G. & Vogel, H-U. (eds.) Concepts of Nature in Traditional China: Comparative Approaches, Leiden: Brill. Chemla, K. (2013) ‘Shedding some light on a possible origin of the concept of fraction in China. Division as a link between the newly discovered manuscripts and The Gnomon of the Zhou [dynasty]’, Sudhoffs Archiv, 97: 174–198. Chemla, K. (2015 [2016]) ‘Proof, Generality and the Prescription of Mathematical Action: A Nanohistorical Approach to Communication’, Centaurus, 57: 278–300. Online at: http://onlinelibrary.wiley.com/ doi/10.1111/1600-0498.12111/full. Chemla, K. (2016) ‘Numerical Tables in Chinese Writings Devoted to Mathematics: From Early Imperial Manuscripts to Printed Song-Yuan Books’, EASTM (East Asian Science, Technology and Medicine), 44: 69–123. Chemla, K. (2018) ‘How Has One Approached, and How Could Have One Approached the Diversity of Mathematical Cultures?’ In: Skutella, M. (ed.) Proceedings of the European Congress of Mathematics 2016, Berlin: European Mathematical Society Publishing House. Chemla, K. & Guo, S. (2004) Les neuf chapitres: Le Classique mathématique de la Chine ancienne et ses commentaires, Paris: Dunod. Chemla, K. & Ma, Biao 馬彪. (2011) ‘Interpreting a Newly Discovered Mathematical Document Written at the Beginning of Han Dynasty in China (before 157 b.c.e.) and Excavated from Tomb M77 at Shuihudi 睡虎地’, Sciamvs, 12: 159–191. Chemla, K. & Ma, Biao 馬彪 (2015 [2014]) ‘How Do the Earliest Known Mathematical Writings Highlight the State’s Management of Grains in Early Imperial China?’ Archive for History of Exact Sciences, 69: 1–53. Chemla 林力娜, K. & Zou, Dahai 鄒大海. (2018) ‘Parts in Chinese Mathematical Texts: Interpreting the Chapter form of The Nine Chapters on Mathematical Procedures’, in Bretelle-Establet, F. & Schmitt, S. (eds.) Pieces and Parts in Scientific Writings. Dordecht, Holland: Springer. Cullen, C. (1996) Astronomy and Mathematics in Ancient China: The Zhou bi suan jing, Cambridge [England] and New York: Cambridge University Press. Cullen, C. (2004) The Suan shu shu 筭數書 ‘Writings on Reckoning’: A Translation of a Chinese Mathematical Collection of the Second Century bc, with Explanatory Commentary, Cambridge: Needham Research Institute. Cullen, C. (2009) ‘People and Numbers in Early Imperial China’, in Robson, E. & Stedall, J. (eds.) The Oxford Handbook on the History of Mathematics, New York: Oxford University Press. Dauben, J.W. (2008) ‘算數書. Suan Shu Shu (A Book on Numbers and Computations). English Translation with Commentary’, Archive for history of exact sciences, 62: 91–178. Guo, Shuchun 郭書春 (1992) Gudai shijie shuxue taidou Liu Hui 古代世界數學泰斗劉徽 (Liu Hui, a leading figure of ancient world mathematics), Jinan: Shandong kexue jishu chubanshe. Han, 韓 Wei 巍 (2013) ‘Beida Qin jian “Suanshu” tudi mianji lei suanti chushi 北大秦簡“算書”土地面積 類算題初識’, Jianbo 簡帛, 8: 29–42. Hubei sheng wenwu kaogu yanjiusuo 湖北省文物考古研究所 & Yunmeng xian bowuguan 雲夢縣博物 館 (2008) ‘Hubei Yunmeng Shuihudi M77 fajue jianbao 湖北雲夢睡虎地M77發掘簡報 (A Concise Report on the Excavation of the Tomb M77 at Shuihudi, in Yunmeng, Hubei)’, Jiang Han kaogu 江漢 考古 (Jiang Han archeology), 109: 31–37 + Plates 11–16. Hulsewé, A. F. P. (1985) Remnants of Ch’in Law, Leiden: Brill. Li, Jimin 李繼閔 (1990) Dongfang shuxue dianji Jiuzhang suanshu ji qi Liu Hui zhu yanjiu 東方數學典 籍 – 《九章算術》及其劉徽注研究 (Research on the Oriental mathematical Classic The Nine Chapters on Mathematical Procedures and on its Commentary by Liu Hui), 西安 Xi’an: 陝西人民教育出版社 Shaanxi renmin jiaoyu chubanshe.

491

Karine Chemla Li, Jimin 李繼閔 (1993) ‘ “ Shang Gao dingli ” bianzheng “商高定理”辯證 (Textual research on the “Theorem of Shang Gao”)’, Ziran Kexueshi Yanjiu 自然科學史研究 (Research in the History of Natural Sciences) 12: 29–41. Li,Yan & Du Shiran (1987) Chinese Mathematics: A Concise History, Oxford [England]: Clarendon Press. Ma, Biao 馬彪 & Lin Lina 林力娜 (Karine Chemla) (2018) ‘秦、西汉容量“石”诸问题研究 (Research on various problems raised by the capacity measuring unit ‘dan’ in the Qin and Western Han time periods)’, 中国史研究 Zhongguo shi yanjiu (Research in Chinese History). Martzloff, J. C. (1997) A History of Chinese Mathematics, Berlin and New York: Springer. Mo, Zihan 墨子涵 (Daniel Morgan) & Lin Lina 林力娜 (Karine Chemla) (2016) ‘ “也有輪著寫的： 張家山漢簡《筭數書》寫手與篇序初探 (There is Also Writing in Turns: Initial Investigation of the Hands and Compilational Order of the Han Bamboo Manuscript Suan shu shu (Writings on mathematical procedures) from Zhangjiashan)’, Jianbo 簡帛, 12: 235–252. Translated into English as Morgan 墨子涵, D. P. & Chemla 林力娜, K. (2018 [2017]) ‘Writing in Turns: An Analysis of Scribal Hands in the Bamboo Manuscript Suan shu shu 筭數書 (Writings on Mathematical Procedures) from Zhangjiashan tomb no. 247’, Silk and Bamboo, 1: 152–190. Peng, Hao 彭浩 (2001) Zhangjiashan Han jian “Suanshu shu” zhushi 張家山漢簡《算數書》注釋 (Commentary on Writings on mathematical procedures, a document on bamboo strips dating from the Han and discovered at Zhangjiashan), Beijing 北京: Kexue chubanshe 科學出版社 (Science Press). Peng, Hao 彭浩 (Forthcoming) ‘Official Salaries and State Taxes as Seen in Qin-Han Manuscripts, with a Focus on Mathematical Texts’, in Michel, C. & Karine Chemla (eds.) Mathematics and Administration in the Ancient World. Dordecht, Holland: Springer Qian, Baocong 錢寶琮 (1963) Suanjing shishu 算經十書 (Qian Baocong jiaodian 錢寶琮校點) (Critical punctuated edition of The Ten Classics of Mathematics), Beijing 北京: Zhonghua shuju 中華書局. Volkov, A. (2001) ‘Capitolo XII: La matematica. 1. Le bacchette’, in Karine Chemla, with the collaboration of Francesca Bray, Fu Daiwie, Huang Yilong & Métailié, G. (eds.) La scienza in Cina, Roma: Enciclopedia Italiana. Online at: www.treccani.it/enciclopedia/la-scienza-in-cina-dai-qin-han-ai-tangla-matematica_%28Storia-della-Scienza%29/, accessed September 30, 2015. Volkov, A. (2007) ‘Geometrical Diagrams in Traditional Chinese Mathematics’, in Bray, F., Dorofeeva-Lichtmann,V. & Métailié, G. (eds.) Graphics and Text in the Production of Technical Knowledge in China.The Warp and the Weft, Leiden: Brill. Volkov, A. (2014) ‘Mathematics Education in East- and Southeast Asia’, in Karp, A. & Schubring, G. (eds.) Handbook on the History of Mathematics Education, New York: Springer. Xiao Can 蕭燦 (2010) ‘Yuelu shuyuan cang Qin jian “Shu” yanjiu 嶽麓書院藏秦簡《數》研究 (Research on the Qin strips Mathematics kept at the Academy Yuelu)’, Ph.D. Thesis in History, Hunan University 湖南大學. Zhangjiashan er si qi hao Han mu zhujian zhengli xiaozu 張家山二四七號漢墓竹簡整理小組 Group of editors of the bamboo strips from the Han tomb 247 at Zhangjiashan (2001) Zhangjiashan Han mu zhujian (erbai sishi qi hao mu) 張家山漢墓竹簡（二四七號墓）Bamboo slips from a Han tomb at Zhangjiashan (Tomb number 247), Beijing 北京: Wenwu chubanshe 文物出版社. Zhu, Hanmin 朱漢民 & Chen Songchang 陳松長 zhubian 主編 (gen. eds.) (2011) Yuelu shuyuan cang Qin jian (er) 嶽麓書院藏秦簡（貳）(Qin Bamboo slips kept at the Academy Yuelu [2]), Shanghai 上海: Shanghai cishu chubanshe 上海辭書出版社. Zhu,Yiwen 朱一文 (2016 (Available online 23 April 2015) ‘Different Cultures of Computation in Seventh Century China from the Viewpoint of Square Root Extraction’, Historia Mathematica, 43: 3–25. Zou, Dahai 鄒大海 (2005) ‘Shuihudi Qin jian yu xian Qin shuxue 睡虎地秦簡與先秦數學 (The Qin slips from Shuihudi and pre-Qin mathematics)’, Kaogu 考古 (Archeology): 537–545. Zou, Dahai 鄒大海 (2007) ‘Shuihudi Bamboo Strips of the Qin Dynasty and Mathematics in PreQin Period’, Frontiers of History in China, 2: 632–654. Online at: http://dx.doi.org/10.1007/ s11462-007-0030-8.

492

23 ASTRONOMY

DAVID PANKENIERASTRONOMY

David Pankenier

Ever since the people have existed, when have successive rulers not systematically calendared the movements of the sun, moon, stars and asterisms? Sima Qian, “Treatise on the Celestial Offices”

In common parlance today astronomy deals with the dispassionate scientific study of extraterrestrial phenomena, but in ancient times the distinction between astronomy and the pseudoscience later termed “astrology” was not clear-cut. Indeed, in the West the first meaning of the quadrivium subject of astronomia was what we mean by “astrology.” Close observation of the movements of celestial bodies for the purpose of timekeeping, and ultimately calendar making, may well have arisen about the same time as did speculation about supra-visible influences associated with them. After all, the diurnal variation of darkness and light and seasonal changes and their profound effects would have been blindingly obvious then as now.The choice of rock shelters and the deliberate placement of doorways of dwellings to face south or southeast tell us as much. Both facets of tianwen 天文 “celestial patterns” as well as cosmology are well represented in early China and were not functionally distinguished until the imperial period.

Neolithic beginnings Images of sun (or stars) and moon frequently appear on Neolithic pottery, along with domestic animals, fish, deer, dwellings, plants, and trees.1 Some of the most striking depictions are found on Dawenkou 大汶口, Liangzhu 良渚, and Yangshao 仰韶 culture artifacts, leading to speculation about the possible prevalence of solar cults among these late Neolithic culture groups.The depiction of such objects in various aspects and configurations, whether as symbols of ownership or something more significant, implies the existence of vocabulary capable of distinguishing them. When evidence of regular, deliberate observation of celestial phenomena first makes its appearance in the archaeological record, it does so in striking fashion. This was the 1987 discovery of a Yangshao 仰韶 culture burial at Puyang 濮阳, Xishuipo 西水坡, in eastern Henan, dating from the late fourth millennium bce.2 The elite individual buried in the center of the exceptionally large, cardinally oriented grave M45 is aligned with his feet to the north and head to the south. At his feet and on both sides are the skeletons of three young sacrificial victims, 493

David Pankenier

presumably acolytes. Around the chief occupant of the tomb are elaborate mussel shell mosaics – a dragon to his east, a tiger to his west, and a third ladle-like figure at his feet to the north, made up of a triangle of shells and two human tibias. Similar, if less elaborately arranged, mosaics were found in separate burials some 50 meters to the south, including one containing the skeleton of an individual whose lower legs are missing. These are presumed to be the tibias found in the main burial. Another burial adds mosaics of a deer and a spider accompanying a dragon and tiger. Still another depicts a human figure sitting astride a dragon. These Yangshao culture burials have excited great interest because of the cosmological associations evoked by the animal “familiars” accompanying the tomb occupant. Astonishingly, the cardinal placement of the dragon and tiger figures are precisely the same as in the correlative cosmology of some 3,000 years later, with the Azure Dragon to the east (a constellation comprising our Virgo–Scorpius) and the White Tiger to the west (corresponding to Andromeda– Orion). The Dragon, a denizen of the riverine east, was associated with spring, and the Tiger, a denizen of the forests and hills of the west, with autumn. The third smaller mosaic in the north has been taken to be a representation of the Northern Dipper, which was nearer the Pole then than at any time since. Rather improbably, some see in tomb M45 a map of the sky, still others a pictorial representation of a cosmological world view, others simply a “shamanic” burial. There is no denying that the head-to-head and tail-to-tail placements of the iconic figures of Dragon and Tiger and their creaturely identities, as well as the cardinal arrangement of all three images, are consistent with the conventional cosmological correlations of the late first millennium bce. The opposition at the ends of the sky of the Dragon and Tiger (Scorpius and Orion) is singled out in early etiological myth as highly significant. It is precisely these stellar configurations, plus the Dipper, that are stressed in classical texts as the three Great Seasonal Indicators (da chen 大辰); that is, crucially important seasonal signposts recognized already by the late Neolithic and very likely much earlier.3 Both constellations or their constituent asterisms appear in the thirteenth century bce Shang dynasty 商 (1554–1046 bce) oracle-bone divinations as recipients of sacrifice.4 Many have pointed out that Orion marked the vernal equinox around 7,000 years ago and Antares in Scorpius marked the autumnal equinox around 5,000 bp. The two dates bracket the epoch of the Puyang tomb.What we appear to have in this tomb is a representation of the prime importance of signal constellations and the Sun in their timekeeping roles, and allusion to their management by uniquely empowered individuals no doubt closely connected with rulership, based on the elaborateness of the burial.The Dipper, later supremely emblematic of the mysterious motive power at the center of the sky, may have been the power to which the cosmo-priest paid homage in life and into whose presence he hoped to be conducted after death. But the arrangement of the burial is equally as revealing of astronomy as it is of archaic cosmology. Taking the burial as a self-contained unit, while the central figure faces north, the obliquely laid out skeleton at his feet points to azimuth 119°, the southeast direction of winter solstice sunrise (and conversely, summer solstice in the northwest). Such an intentional alignment would have been empirically established by paying close attention to the sun’s changing rising points along the horizon throughout the year.5 The Puyang burial thus provides the strongest indication of a systematic focus on solar observation designed to serve both practical and ritualistic ends. This effort will culminate by the Shang dynasty with the creation of a lunisolar calendar. Elsewhere, in the northeast (Hongshan 紅山 culture) and southeast (Liangzhu 良 渚 culture), rigorous cardinal alignment of square sacrificial platforms and earthworks points to a ritualistic focus on a sacred locus defined by four quarters, signaling the identification of four cardinal directions, north, south, east, west.6 Here we have the roots of the perennial Chinese

494

Astronomy

cosmography, whose chief structural features are a circular dome of the sky (circular altars are already present in Hongshan) above a flat, square earth.7 The next major advance documented in the archaeological record is the systematic (albeit ritualized) observation of the sun throughout the year carried out at an elaborately designed and constructed platform dedicated to the purpose. At Taosi 陶寺, in the heartland of the first semilegendary Chinese dynasty, the Xia 夏 (ca 1953 – 1555 bce), at a late stage in the development of a large urban center adjacent to the Fen River, a multi-layered elevated terrace was constructed adjoining the city wall and accessible only from the elite residential area. On the uppermost level facing east was a curved structure that matches in length the range in degrees along a mountain ridge in the distance traveled by the sun from south to north between the solstices. Remnants of the structure include a pounded earth wall perforated by slot-like apertures through which the sun could be observed to rise at specific intervals during its progress along the horizon. Although subsequently razed and its associated elite burials desecrated when Taosi was destroyed by conquest, enough remains of the site to identify it as a likely precursor to the storied ling tai 靈台 or Numinous Terrace attested in the earliest textual sources. Taosi’s location very near Pingyang 平陽, reputedly the birthplace of Yao 帝堯, the legendary founding “emperor” of the semi-legendary Xia 夏 dynasty, has naturally fueled speculation that Taosi is, in fact,Yao’s capital. To Yao is attributed an important chapter in the earliest surviving text on governance, purportedly dating from the early first millennium bce, the Shangshu 尚 書 or Venerated Writings. A seminal chapter at the beginning, the “Canon of Yao,” enjoys pride of place in this earliest canonical compilation.This classic, to which we will return in a moment, highlights the crucial importance of the calendar as a symbol of the legitimacy of the ruler, which explains its appearance in a venerated text on governance.

The first written records The Shang dynasty set the stage for all further development in astronomy, astral omenology, and cosmology. Several decades of research on the divinations on shell and bone from the late Shang dynasty (ca. late thirteenth to eleventh centuries bce) have greatly advanced our understanding of astronomy in the second millennium bce. Because it is the first historical and recognizably Chinese dynastic state, the vast numbers of inscriptions on shell and bone that record divinations about diverse aspects of the religious life and rituals of the late Shang are richly informative about this crucial formative period. From the inscriptions we learn that the Shang were intensely concerned with the timing of sacrifices, royal hunts, military sorties, weather, and the like, and with apotropaic rites. These rites were designed to discern how the supra-visible powers and ancestral spirits were disposed toward high officials and affairs of state; that is, which of them might either favorably or unfavorably intervene and how they might be mollified. Given the Shang preoccupation with timeliness and precise dating, it has been possible to glean much information about the luni-solar calendar conventionally in use. For example, we know that months were numbered, and from sporadic thirteenth-month dates we also know that inter-year intercalation was already being carried out to synchronize the lunar cycle and the tropical year. How regularly this was done is still unclear, since the entire calendar has not been reconstructed. The days were also enumerated serially from one to sixty using the sexagenary binomial terms generated by sequentially pairing the ten members of a set of “heavenly stems” with each of the twelve “earthly branches.”This system has been in uninterrupted use ever since. Initially, it is thought that Shang calendar months were all 30 days long and that every third month began with day “one” in the sequence, jiazi 甲子, so that the sexagenary designations of

495

David Pankenier

the days remained constant. But already by the late Shang this practice had been abandoned in favor of alternating short and long months of 29 and 30 days in order to approximate the 29.54 day length of the synodic lunar month. Each month was divided into three 10-day weeks xun 旬, but the one-to-one correspondence between days of the month and sexagenary designations was lost.8 The vast majority of divinations are only indirectly records of historical events, so that references to celestial phenomena appear sporadically in the formulaic inscriptions only insofar as they were subjects of divinatory concern. Not all have been deciphered. Those that have been include mention of eclipses, sunspots, and comets; that is, anomalous occurrences that excited concern as ominous signs from the supra-visible powers.9 Some phenomena routinely appear as a matter of institutionalized ritual, such as sacrifices to the rising and setting sun and to a small number of seasonally important stars or asterisms, such as Scorpius (as the Fire Star, huo 火), Orion’s Belt (Triaster, shen 參), Orion’s Sword (Attack, fa 伐) and the Dipper (dou 斗).10 Other “probables” are also mentioned, e.g., Dragon, Tiger, Woman, but these have yet to be confirmed.11 The five planets visible to the naked eye which move about independently of the panoply of stars were evidently thought to be minions of the Supernal Lord, Shangdi 上帝, and were referred to collectively as the Five Minister Regulators, wu chen zheng 五臣正. As we shall shortly see, intimate gatherings of all five were thought to have tremendous significance. Only Jupiter has been tentatively identified in the divinations (Jupiter was called sui 歲 “year” because its annual period of visibility approximates one year). It is impossible to believe that such a striking object as Venus did not attract cultic attention as elsewhere in the ancient world, leading one to suspect that it has not yet been identified in the inscriptions. Given the existence of the continuous 60-day cycle of day-dates, the fact that divinations are so regularly dated as to month and day raised the tantalizing possibility of precisely dating eclipses mentioned in the inscriptions. Unfortunately, in the case of solar eclipses, in all but one case the divinations are prospective, that is the diviner is concerned with whether an eclipse of the sun might occur on a given date some days hence. These are clearly not records of an actual occurrence. In only one case is it obvious that the inscription contains a confirmation after the fact of a prior prognostication: Divining on day yisi: wine libation ritual day, we should perhaps X sacrifice to [royal ancestor] Xiao Yi. Use this. The sun was eclipsed; report this tonight to [first high ancestor] Shang Jia together with nine cattle. Number one [test]. (甲編 0755 (He ji 合集 033696) The preface is one of a series of “charges” to the oracle on the day yisi, regarding a regular libation rite for the royal namesake of stem-day yi 乙; that is, King Xiao Yi 小乙. He was the immediate predecessor of King Wu Ding 武丁 (late thirteenth century bce), longest reigning ruler of the late Shang. It was first during Wu Ding’s reign that the divinations began to be punctiliously inscribed on the cracked bones, preserving for posterity the earliest recognizable form of Chinese writing. The proposition put to the oracle is whether a certain sacrifice should be offered to the deceased King Xiao Yi in addition to the regularly scheduled wine libation on that day. The reading of the response affirmed the proposition, hence the notation “use this” was made on the shell. What then follows is a verification, after the fact, that an eclipse occurred that day, which it was deemed important to report that very night to the dynastic high ancestor, Shang Jia, together with a sacrifice of nine cattle. Chronological and calligraphic evidence in tandem with a search of all yisi-day eclipses potentially visible in the vicinity of the Shang capital suggests 496

Astronomy

to some scholars that this late period inscription plausibly refers to a large partial eclipse (mag. 0.91 at Anyang) on the morning of 31 Oct 1161 (0.95 at Zhengzhou, location of a major urban center 100 miles to the southwest), but this is unconfirmed.12 The situation with lunar eclipse records was more promising. Applying the same methodology to a series of five lunar eclipse inscriptions during King Wu Ding’s reign enabled scholars to narrow the possible candidates down to five total eclipses, which occurred between 1201 – 1181 bce. They are: i ii iii iv v

night of day guiwei, 07/12/1201 bce; night of day shenwu, 11/04/1198 bce; night of day jiwei-gengshen, 12/28/1192 bce; night of day renshen, 10/25/1189 bce; night of day yiyou, 11/25/1181 bce.13

The end result is that with the aid of astronomy a crucial benchmark in the chronology of the Shang dynasty was credibly established, dating King Wu Ding’s reign to 1250–1192 bce. By the late Shang dynasty, knowledge of the solar and lunar cycles and observation of solar shadows using a gnomon had long since achieved a level of sophistication that permitted the royal officials to maintain an accurate luni-solar calendar. This is to be expected given the attention given to solar observation at Taosi some eight centuries earlier. At some point during that long interval written records must have begun to be kept on perishable materials, probably bamboo or wooden slips tied in bundles, as suggested by the Shang period form of the graph for “documents” ce 冊, a pictograph of precisely that. Drawing signs and glyphs on pottery using brush and ink had been in wide use since the mid-Neolithic. The evidence suggests that the luni-solar calendar declined in relative importance as a result of the strict sequence of five regular sacrifices dedicated to dead kings and consorts. By the end of the dynasty those rituals had expanded to require an entire calendar year. To all appearances, the meticulous timing and performance of sacrifices had become the main preoccupation of the officials in charge of ritual activity, which consumed an inordinate amount of resources. References to celestial phenomena disappeared from the divinations. In terms of cosmology, it is clear that the late Shang elites were even more intensely focused on cardinal orientation than their predecessors, a pattern that would persist until the end of the imperial period. Every important structure was axially oriented north-south, from palaces and royal tombs to commoners’ graves and even burial pits for sacrificial victims. The mysterious void at the Pole, lacking a true pole star throughout the formative period of Chinese civilization, was seen as the abode of the Supernal Lord, whose invisible emanations obliged the entire cosmos to revolve about his locus at the apex. The Shang cultic center called zhong Shang 中商 in the inscriptions, later called the Ruins of Yin (yinxu 殷墟) at Anyang in Henan, served as their axis mundi. Presumably, the turtle’s flat plastron with its scutes and sutures evoked the regions and rivers of the terrestrial realm, while its dome-like carapace resembled the overarching sky. The idea that the turtle was a simulacrum of the cosmos may have persisted since the Neolithic. This numinous significance may explain the adoption of turtle plastrons as the intermediary endowed with the spiritual capacity to facilitate oracular communication with the unseen powers. The Four Quarters si fang 四方 figure importantly in the divinations, and the iconic role of the mantic Dragon and Tiger persists in their symbolic association with the riverine east and the western wilds, respectively. Their images are virtually ubiquitous on bronze, jade, and pottery. One exceptional, stylistically eccentric bronze comes from the tomb of a Shang elite individual, presumably a cosmo-priest. The dragon-shaped ewer displays on its sides realistic depictions 497

David Pankenier

of alligators juxtaposed with dragons, pointing to their close association in the minds of the manipulators of powerful symbols. At various points on the lid of the ewer are roundels whose strategic placement and irregular sizes appear to correspond to prominent stars in the Dragon constellation, with the raised mushroom-cap knob in the middle serving as the bright Fire Star Antares at the Dragon’s heart. If this interpretation holds, this stylistically eclectic Northern Zone bronze is unique in reproducing the earliest recognizable depiction of a constellation, and also the earliest attempt to show the relative brightness of stars.14

The calendar, astral omens, and heavenly legitimation As mentioned earlier, the “Canon of Yao” is the earliest text to describe the length of the year, seasonal stars and asterisms, positional observation, the practice of intercalation to synchronize the solar and lunar calendars, and the perennial royal duty to “bestow the seasons” shou shi 授時 on the people. This function was ideologically emblematic of the “celestial” legitimacy of the sovereign and his management of the imperative to conform to Heaven.To this day, “bestow the seasons,” serves as the name of China’s National Time Service Center, whose astronomical and navigational functions correspond to those of the U.S. Naval Observatory. [Yao] then commanded Xi and He, in reverence to August Heaven, to calendar the [astral] signs, Sun, Moon, stars, and seasonal asterisms, so as to respectfully bestow the seasons on the people; to host the rising sun, and regularly arrange the initiation of affairs in the east; to take leave of the setting sun, and regularly arrange the completion of affairs in the west; to regularly attend to [the Sun’s] change [of direction] in the north; to regularly attend to [the Sun’s] southward displacement and reverently [mark] the solstice . . .The day being of medium length and the asterism being Bird [α Hydra], he thereby determined mid-spring; the day being longest and the asterism being Fire [α Scorpius], he thereby determined mid-summer; the night being of medium length and the asterism being Ruins [β Aquarius], he thereby determined mid-autumn; the day being shortest and the asterism being Topknot [7 Taurus], he thereby determined mid-winter.15 Many efforts have been made in the past to date this text based on the astronomical details it contains, but its very nature as a late compilation of early fragments, along with the vagueness of the passage with regard to the time of observation, means a definitive solution will remain elusive. What can be said with assurance, based on analogous concepts, practices, and terminology appearing in the Shang oracle-bone inscriptions, is that some techniques described certainly pre-date the text’s compilation by centuries, and possibly up to a millennium. One of those techniques, illustrated by the seasonal stars identified in the text, shows that a precursor to the scheme of twenty-eight lunar lodges so central to later positional astronomy was already in existence by the late Shang dynasty. Only four asterisms are mentioned, but these same asterisms anchor the four celestial quadrants of the later scheme as well. They are associated in “Yao dian” 堯典 with a given season by their achronical rising (evening at dusk), from which the knowledgeable official would also know where the sun would be three months, six months, and nine months later. This schematic technique of indirectly deriving positional information from an established pattern is typical of the non-geometrical approach of later astronomy. Ancient Chinese sky-watchers almost entirely ignored heliacal risings of stars, another characteristic that clearly distinguishes Chinese astronomy from those of Mesopotamia and ancient Egypt. 498

Astronomy

The four cardinal stars or asterisms correlated with the seasons and four directions tell us that at least a rudimentary correlation between sky above and earth below was already in process of formation by the late Shang. At the same time, the single most important topographical feature of north China was the Yellow River, so it is hardly surprising that it would come to play the central organizing role in the scheme of astral-terrestrial correspondences which gradually took shape. In the course of time, the spaces between the cardinal asterisms in the path of the moon were filled in with additional stellar configurations, so that during the first half of the first millennium bce the now familiar scheme of twenty-eight lunar “lodges” came into being, one for each night of the lunar month, during which time the moon completed a circuit of the heavens. More illustrative of the practical application of early astronomy is the use of astral alignments to orient high-value structures in the landscape. By the late second millennium bce, astral officials had discovered that the east and west sides of the asterism known to us as the Square of Pegasus were accurately aligned on true north, if observed when perpendicular to the horizon in late autumn (Figure 23.1). Called Ding 定, “determiner, establisher” (or “Pure Temple”), this particular location marked by α Pegasus played a seminal role in astral lore as the place where an unforgettable sign from the Supernal Lord had occurred. In late February 1953 bce, the Five Minister Regulators, the visible planets, gathered there in the most compact grouping of the five in human history. Given the pole’s cosmo-political significance as the abode of the Supernal Lord, and the absence of a bright star at that spot (the closest, Kochab or β UMi, was more than 6° from the pole and inadequate for the purpose), it was imperative to devise a method to accurately bring down to the ground the north-south alignment bestowed by the Supernal Lord. Although in theory the use of a gnomon and solar shadows should have rendered such a technique redundant, later texts make it clear that even a millennium later this astral method was still ritually prioritized. It is worth noting that in this uniquely Chinese approach the stellar alignment was observed and the baseline staked out on the ground while facing Pegasus/Ding on the meridian due south, with one’s back to the Pole.The angular distance separating the Pole from Pegasus-Ding makes it impossible to see both simultaneously. The early Chinese sky-watchers distinguished themselves in observing three impressive planetary groupings in the second millennium bce, no doubt because of the role played by those “five pacers” as deputies who conveyed the Supernal Lord’s intentions. This shows that, although documented only sporadically, or merely alluded indirectly to in the earliest texts, astral omenology based on the behavior of the five planets coupled with a nascent scheme of astralterrestrial correspondences would have been on a theoretical footing by the late Shang dynasty. The ideological interpretation given to the second most spectacular conjunction of the planets, in late May 1059 bce, tells us as much.16 That conclave of the five celestial “Minister Regulators” occurred between or just ahead of the “beak” of the Chinese constellation called the Vermilion Bird. By chance this was the very astral location associated with the fortunes of the Zhou polity, then a comparatively minor power in the Wei River valley of Shaanxi in the west. History records that the Zhou rulers, erstwhile defenders of the western flank of the Shang dominions, promptly set about plotting to overthrow the Shang and usurp the royal power, which feat was accomplished thirteen years later in 1046 bce, with the aid of numerous disgruntled Shang vassals.With that, the precedent was established that this rarest of astral omens would henceforth be identified as the preeminent sign of the Supernal Lord’s displeasure with the current rulership and that a dynastic transition was in the offing. This cosmo-political watershed was understood to signify the withdrawal of Heaven’s Mandate (tian ming 天命) from the ruling house and its bestowal on another lineage.Whether explicitly cited or only vaguely alluded to, the precedent thereby established concentrated the minds 499

Figure 23.1 Star chart showing the Ding constellation perpendicular to the horizon in late fall midseventh century bce, at the time when this accurate alignment technique is documented in the Book of Odes. (DWP – Starry Night Pro 7.6)

Astronomy

of rulers and usurpers alike ever after. The locations of the planets at several historic turning points subsequently and the significance attributed to such planetary clusters in astrological theory reinforced the idea. In a pinch, if no spectacular conjunction of planets were forthcoming, a comparatively routine alignment ranging over many more degrees in longitude could be pressed into service, as in the case of the founding of the Han dynasty. Remarkably, the line-up of planets in 205 bce, ranging over more than 30°, was centered on the very same location as the epoch-making Zhou conjunction of 1059 bce (Figure 23.2). As Grand Scribe-Astrologer Sima Qian 司馬遷 noted a century after the Han founding: When the Five Planets gather this is a change of elemental-phase. The possessor of virtue is celebrated, a new Great Man is set up to possess the Four Quarters, and his descendants flourish and multiply. But the one lacking in virtue suffers calamities or even extinction.17 Jupiter’s location and especially planetary massings, having become a feature of the ideology of rulership, continued to figure importantly in political calculations at crucial moments throughout history.

Figure 23.2 The conjunction in late May 1059 bce at the “beak” of the Vermilion Bird. The five planets clustered for days within a roughly 7° circle in the northwest. Jupiter’s return to the heart of the Vermilion Bird in 1047–46 bce affirmed that the time was right for the Zhou to launch the campaign to overthrow the Shang dynasty. DWP – Starry Night Pro 7.6

501

David Pankenier

Spring and Autumn period through Warring States In step with the growing complexity of the political landscape, a hermetic science of prognostication, in which Jupiter (the “year star”) figured prominently, gradually took shape. An elaborate system of astral-terrestrial correspondences grew up as an outgrowth of the Shang Four Quarters conception. In keeping with the more complex political landscape of the early to mid-first millennium, the new scheme, called field allocation fenye 分野, matched each of the ancient Nine Provinces of north and central China (later twelve states) with different astral fields according to the provinces’ locations in relation to the Yellow River. The Yellow River was the terrestrial analogue of the Milky Way, the River of Heaven, whose southwest to northeast course determined the directionality ascribed to the terrestrial counterparts of the astral fields (so that, e.g., Shandong peninsula on China’s east coast became “astral” north). Not surprisingly, given China’s advanced civilization, the entire sky was given over to their territory, and peripheral peoples had no place in the scheme. Among the planets, Jupiter played the preeminent role, and prognostications concerning that planet are well represented in the Warring States narrative histories Discourses of the States, Guoyu 國語 and the Tradition of Zuo, Zuozhuan 左傳. Integral cycles of Jupiter provided the basis for predictions of good or ill befalling the various regions. Having Jupiter “in one’s state” was invariably auspicious because it was thought to render one immune to attack.To ignore the astral signs is portrayed as foolhardy and invariably invited calamity. Judging from Han texts, by the Warring States period the other visible planets also figured importantly in prognostications, but their role is poorly documented before the early imperial period. The complete scheme of lunar lodges had little utility in calendrical science, so it may well be that it was the astrological imperative for more fine-grained correlations that spurred its elaboration. Besides the cardinal asterisms mentioned in the Shang oracle-bone inscriptions and the four seasonal lodges named in the “Canon of Yao,” a half-dozen other asterisms (not all of them in the path of the moon) are named in the poem “Great East” in the oldest collection of poetry, the Odes, shi 詩 (Zhou dynasty, early first millennium bce). One ode, “Seventh Month,” dwells on the setting of the Fire Star Antares (heart of the Dragon) as a harbinger of the preparations associated with the onset of winter. Asterisms mentioned in other poems include Determiner (Pegasus), Weaving Maid (Vega), Draught Ox (Altair), Basket (Sagittarius), Dipper (Ursa Major), Hunting Net (Taurus),Topknot (Pleiades), and also Venus as morning and evening star. An archaeological discovery from 433 bce shows clearly that the entire scheme of lodges was complete and in conventional use by that time. A lacquer hamper used to store ceremonial robes unearthed in the tomb of the ruler of a minor state is unique in reproducing the entire roster of lodge names in writing, rather than pictorially or using the ball-and-link style of two to three centuries later. A hint of the process of elaboration from the original four cardinal asterisms to twenty-eight lodges is provided by the division of the asterism Ding (Square of Pegasus) into two separate “walls,” denoted East Aligner (dong ying 東營) and West Aligner (xi ying 西營), both of which point to the role of that constellation in the alignment of high-value structures.18 The table of the determinative stars of the twenty-eight lodges shows their highly irregular angular extension. Many of the lodges are either considerably larger or much smaller than the moon’s mean daily motion of about 13°. The scheme is obviously an idealized schematic representation, which would have been primarily useful in mantic applications or as a heuristic device. It is significant that the only lodges named for the body parts of their associated cardinal emblem are those comprising the Azure Dragon of the east, a reflection of the antiquity of that constellation and its usefulness as an indicator of the passing seasons since the Neolithic.19 To this day, Chinese farmers all know that “on the second day of the second month, the Dragon raises 502

Astronomy

its head.” The Dragon’s horns were marked by the star Spica (α Virginis), whose rising played a similar role in ancient Greece. It seems clear that the lodge system evolved from an initial focus on the Dragon constellation and the Fire Star at its heart as a seasonal indicator throughout the year,20 to the four cardinal asterisms in the “Canon of Yao,” and finally to the twenty-eight lodges and four cardinal “zodiacal” emblems – Dragon, Turtle and Snake, Tiger,Vermilion Bird – of the mature scheme (Table 23.1). Previously, when the early history of Chinese astronomy was still poorly understood, there was a tendency to see the lunar lodges as a transmission of a roughly analogous scheme from India, the nakṣatras. That view is now generally dismissed, as is the claim that Chinese astronomy as a whole diffused from ancient Babylonia.21 The next stage was the gradual identification of additional stars and asterisms by the skygazers of the Spring and Autumn (722–479 bce) and Warring States (403–221 bce) periods, especially pre-imperial astronomers Gan De 甘德 of Lu 魯 and Shi Shen 石申 of Wey 衞, who are

Table 23.1 The four cardinal emblems and twenty-eight lunar lodges Four cardinal emblems (四象)

Lunar lodge (宿)

Azure Dragon of the East (東方青龍) Spring

Black Turtle of the North (北方玄武) Winter

White Tiger of the West (西方白虎) Fall

Vermilion Bird of the South (南方朱鳥) Summer

Name

Translation

Determinative star

 1  2  3  4  5  6  7

角 (Jiao) 亢 (Kang) 氐 (Di) 房 (Fang) 心 (Xin) 尾 (Wei) 箕 (Ji)

α Vir κ Vir α Lib π Sco σ Sco μ1 Sco γ Sgr

 8  9 10 11 12 13 14 15 16 17 18 19 20 21 22 23

南斗 (Nan dou) 牽牛 (Qian niu) 須女 (Xu nü) 虛 (Xu) 危 (Wei) 營室 (Ying shi) 東壁 (Dong bi) 奎 (Kui) 婁 (Lou) 胃 (Wei) 昴 (Mao) 畢 (Bi) 觜 (Zui) 參 (Shen) 東井 (Dongjing) 輿鬼 (Yu gui)

24 25 26 27 28

柳 (Liu) 七星 (Qi xing) 張 (Zhang) 翼 (Yi) 軫 (Zhen)

Horns Neck Base Chamber Heart Tail Winnowing Basket Southern Dipper Ox-leader Serving Girl Ruins Rooftop Align-the-Hall Eastern Wall Stride Pasture Stomach Topknot Hunting Net Beak Triaster Eastern Well Ghost Conveyance Willow Seven Stars Spread Wings Chariot Platform

503

Equatorial extension ca 12° 9° 15° 5° 5° 18° 11°

φ Sgr β Cap ε Aqr β Aqr α Aqr α Peg γ Peg η And β Ari 35 Ari 17 Tau (Pleiades) ε Tau λ Ori ζ Ori μ Gem θ Cnc

26° 8° 12° 10° 17° 16° 9° 16° 12° 14° 11° 16° 2° 9° 33° 4°

δ Hya α Hya υ¹ Hya α Crt γ Crv

15° 7° 18° 18° 17°

David Pankenier

each credited with compiling Canons of Stars (xing jing 星經), among other works, in about the fourth century bce. Their tabulations would have been lists of more or less easily recognizable lodge stars and asterisms, possibly including descriptions of their positions, colors, and brightness. A third ancient astronomer, “Shaman Xian,” Wu Xian 巫咸, who supposedly lived under the Shang, is a much more shadowy figure. Gan De is also noteworthy for a Treatise on Jupiter, Suixing jing 歲星經, and on astral prognostication, Tianwen xing zhan 天文星占, illustrating the dual nature of the profession.22 An early, “old-degree” (gu du 古度) system mentioned in the Han scholar Liu Xiang’s (77–6 bce) Commentary on the Great Plan, Hongfan zhuan 洪範傳 survived in the form of scattered positional data preserved in the Prognostication Canon of the Kaiyuan Reign Period, Kaiyuan zhan jing 開元占經 (ca 729 ce). This is all that remains of the Warring States system, which was still in use in the early Han dynasty, as shown by so-called mantic-astrolabe shi. That early “old-degree” series preferred bright stars as lodge markers, in contrast to a newer scheme appearing, for example, in the astronomical chapter of the late second-century bce text Huainanzi 淮南子.23 By the early imperial period, positions of stars (and planets) were conventionally given in terms of their angular distance in Chinese du along the equator east of the determinative star of the lodge, together with their polar distance. In the Chinese equatorial system, the sun advanced one du per day or 365.25 du during the course of a year, making one du slightly smaller (0.9856°) than one degree. Recently it was demonstrated mathematically that the positional data for the lodges of this second series (attributed to Shi Shen’s Canon) date from the early first century bce, when an equatorial armillary would have been used to measure their extensions.24 Chinese astronomy was resolutely polar-equatorial, and the equator was (and still is) called the “Red Road.” The ecliptic, although recognized and named the “Yellow Road,” was generally ignored until 84 ce when Fu An 傅安 and later Jia Kui 賈逵 (30–101 ce) added an ecliptic ring to the armillary which by then had come into use.

Warring States through Han By the late Warring States period Jupiter was already becoming a species of “time spirit” called tai sui 太歲 or sui yin 歲陰 in esoteric hemerology. This was a natural extension of the planet’s documented early role by the early Spring and Autumn period (722–479 bce) as a counter of years in a twelve-year cycle. Jupiter is annually visible for about 365 days out of its mean synodic period of 399 days and completes a circuit of the sky in 11.86 years. This means that as Jupiter annually travels eastward roughly 30° among the stars, during superior conjunction with the sun each year, while it is unobservable, the planet imperceptibly gains a little ground. As a result, after some seven cycles, the planet would “unexpectedly” appear one Jupiter station (“bivouac” ci 次) farther along than a simple twelve-year count would lead one to expect, a phenomenon called “overtaking a sign” chao chen 超辰 in an early astronomical system, Liu Xin’s 劉歆 (ca. 50 bce–23 ce) “Three Regimes Calendar” San tong li 三統曆.25 According to Sima Qian, it was Gan De and Shi Shen in the Warring States period who first recognized that Mars retrograded. In his “Treatise on the Celestial Offices,” the Grand Scribe Astrologer was the first to assert that, “in my own perusal of the scribal accounts I have examined into [the planets’] events and movements, and in the past one hundred years there has not been one instance when the Five Planets have appeared and not reversed course and retrograded.”26 Prior to this, and despite progress in establishing the regularity of the planetary period, many still held that they could be moved about, speeded up, or slowed down at will by the Supernal Lord. 504

Astronomy

Liu Xin’s “Three Regimes” was not simply a calendar but a comprehensive astronomical and numerological system that attempted to synchronize all celestial periodicities and calendrical cycles. It was the first thoroughgoing reform since the ambitious “Grand Inception” Taichu 太初 calendar reform of 104 bce ordered by Han Emperor Wu 武 (141–87 bce). Emperor Wu’s Grand Inception reform, the Han dynasty’s first after a century in power, was rendered necessary because it had become patently obvious that the system carried over from the Qin dynasty (221–206 bce) had become hopelessly out of date. A telltale problem was that by Emperor Wu’s time the beginning of the year – the location of the sun at winter solstice – was obviously not where the old Warring States system said it should be. Such a fundamental discrepancy would render accurate eclipse prediction impossible. Eclipses being the supremely inauspicious celestial omen affecting the state, this naturally reflected poorly on the emperor. The cause of the error, of course, is the precession of the Earth’s axis about the Pole, which causes the solstices and equinoxes to imperceptibly drift westward over time. Although the consequences were recognized and periodically corrected for, the phenomenon was not properly explained until some five centuries later and continued to prompt revisions of the calendar over the centuries. Talented scholars were recruited from throughout the empire to work on the Grand Inception reform under the supervision of court officials, notably Grand Scribe-Astrologer Sima Tan 司馬談 (d. 110 bce) and later his son Sima Qian 司馬遷 (d. 86 bce), who gives a terse account of the project in the “Treatise on the Calendar,” Li shu 曆書, in the Grand Scribe’s Records, Shji 史記. The Li shu relates that a “mantic specialist” (fang shi 方士), Tang Du 唐都, an expert on the stars, was summoned to reapportion the divisions of the sky, no doubt by standardizing the determinative stars and extensions of the lunar lodges. Another participant in the reform project whose contribution was crucial was Luoxia Hong 落下閎 from the Yi-minority region in the province of Ba 巴 in the far southwest. Recalling that Sima Qian described Luoxia Hong’s role in the Grand Inception calendar reform as having yun suan zhuan li 運算轉歷 “iteratively calculated the revolutional sequence” of the stars identified by Tang Du, so that the “dimensions (du) of the solar divisions conformed to the Xia calendar,” it seems certain that he was using an early form of armillary.The equatorial armillary was fully developed by 52 bce, when the astronomer Geng Shouchang 耿壽昌 added a fixed equatorial ring, although it is hard to imagine how Luoxia Hong could have measured equatorial extensions without one. The exhaustive reform project was completed sometime before 110 bce, but was not promulgated until December 105 bce, when the beginnings of the major cycles – new moon at midnight, winter solstice at dawn, first day of the sexagenary cycle, first year of the Jupiter cycle, etc., all supposedly coincided. The Taichu calendar would have given the synodic periods of all five planets and delineated the motions during one period for each planet. Presumably, these are the same detailed indications found in the planetary section of Sima Qian’s “Celestial Offices.” The Grand Inception system incorporated the now familiar twenty-four solar terms that define the tropical year. These are the jieqi 節氣 “nodes of the materia vitalis” or qi. By definition, every third qi-node, including the solstices and equinoxes, was deemed a “middle qi-node,” zhong qi 中氣, and the rule was that when a month came along that did not contain one it would be designated an intercalary or “repeated” month. Liu Xin’s “Triple Concordance” system a century later, which was fortunately copied whole into the Treatise on harmonics and the calendar, Lü li zhi 律曆志, in the Hanshu 漢書, History of the Former Han Dynasty (first century ce), adopted 29.53 days as the length of a month, and the length of a tropical year was given as 365.25 days (denoted the “quarter-day remainder” si fen 四分 system). It was prescribed that there should be seven intercalary months in nineteen years, that is, every 235 lunar months. 505

David Pankenier

From the Shang dynasty through the Spring and Autumn period, intercalary thirteenth (and occasionally even fourteenth!) months were added at the end of the year. An example of lackadaisical calendar management from the very end of the Spring and Autumn period is the following. Ji Kangzi asked Confucius, “This is the Zhou twelfth month, the Xia tenth month, and still there are katydids, why is that?” Confucius answered, “I have heard ‘after the Fire Star has set, the hibernators have all gone.’ But now the Fire Star is still declining in the west. The officials in charge of the calendar are wrong.” Ji Kangzi said, “By how many months are they off?” Confucius said, “By the tenth month of the Xia calendar the Fire Star has already set. But now the Fire Star is still visible; [they have] missed intercalating twice.”27 Besides documenting the extraordinary need to intercalate two months into the calendar, which shows that nothing had been done for at least five years, this passage also provides a good illustration of how observation of both the phenomena of nature and the stars served as a check on officialdom. It goes without saying that the farmers in Confucius’ home state could hardly have paid much attention to the official calendar, even if it had been promulgated in a timely manner. Until the late Warring States period the cause of eclipses was poorly understood, and even after the astronomers grasped the significance of the coincidence of new moon and solar eclipses many others continued to insist that the moon shone by its own intrinsic yin or “cool” light, so deeply entrenched was the yin-yang duality. Moreover, in the gai tian 蓋天 “canopy sky” cosmology that prevailed until the Han dynasty, in which the sky was imagined to be like a round, tilted carriage canopy rotating above either a flat or dome-like earth, naturally it was impossible to imagine that the earth could cast a shadow on the face of the moon. In any event, the moon seems to have played a surprisingly minor role in astral prognostication. Between the Shang dynasty and the Han there is only one verifiable record of a total lunar eclipse, on March 12–13, 1065 bce. That record evidently survived only because it was seen by the Zhou dynasty founder King Wen (文王 d. 1050) as a portent of his impending demise some years before the successful overthrow of the Shang dynasty. Solar eclipses were an entirely different matter, given their extremely inauspicious implications. No verifiable eclipse record from the Western Zhou (1046–771 bce) has survived, but during the two-and-a-half centuries encompassed by the canonical Spring and Autumn Annals, Chunqiu 春秋 (722–481 bce), thirty-seven solar eclipses are recorded, of which nearly all are accurate. The technical term ji 既, meaning complete or “total,” appears there for the first time, which points to the possible existence of other specialized terminology as well.Three records of comets and one of meteorites are found in the chronicle, obviously many fewer than one might expect, which fact reinforced the traditional interpretation that the inclusion of specific omens, supposedly by Confucius, was intended as a subtle commentary on the deficiencies of the rulership. From the Former Han dynasty (206 bce–9 ce) on, the majority of large solar eclipses affecting China were faithfully recorded in the dynastic histories (128 eclipses from the Han dynasty alone), either in the “Basic Annals” of emperors or in the regular monographs devoted to astronomy and the calendar, including eclipses which were unobserved at the capital but were reported from the kingdoms and distant commanderies. Here are two examples, both from the “Monograph on Astronomy” (tianwen zhi 天文志) in the History of the Former Han: bce

709 Jul 17

Duke Huan of Lu, third year, seventh month, day renchen [29], the first day of the month. The sun was eclipsed. It was total. [Spring and Autumn Annals] 506

Astronomy

bce

89 Sep 29

Emperor Wu, fourth year of the Zhenghe reign period, eighth month, day xinyou [58], the last day of the month. The Sun was eclipsed; it was not total but was like a hook. It was two du in Kang [Libra]. At the hour of fu [LT = 15–17h] the eclipse began from the northwest. Towards the hour of sunset it recovered. Prior to the 1973 discovery of a trove of silk manuscripts in an elite tomb at Mawangdui 馬 王堆 near Changsha in Hunan, the most authentic source for the astronomy of the early imperial period was Sima Qian’s “Treatise on the Celestial Offices.” Among the documents recovered from the Mawangdui tomb, two manuscripts provide insight into the kinds of sources Sima Qian likely had at his disposal in the imperial library. One of these is a remarkable pictorial album of numerous cloud-like vaporous emanations of materia vitalis, qi 氣, some in zoomorphic shapes, together with more than two dozen different types of comets, their names, and brief generic prognostications. The twenty-eighth of the comets depicted in the Mawangdui atlas is called “Chi You’s Banner” after a mythic cosmic miscreant. Its appearance was commonly regarded as a portent of war. The comet, described elsewhere as “long, multi-branched, and splendiferous, yellow above and white below,” is captioned on the manuscript “troops abroad return.” Here is a sample record from 240 bce of Comet 1/P Halley’s earliest identifiable appearance, which appears in the “Basic Annals of Qin” in the Grand Scribe’s Records: In the seventh year [of the First Emperor of Qin]; a broom star initially emerged in the east, was observed in the north, then in the fifth month it appeared in the west . . . [in the ninth year] a broom star appeared in the north, from the Dipper southward for eighty days. More important still is the Mawangdui manuscript entitled “Prognostications of the Five Planets” Wu xing zhan 五星占, whose very existence is testimony to the signal importance attached to the behavior of the planets at the time.28 Buried with its owner in 168 bce, the “Prognostications” begins with an astrological preface, followed by a tabulation of the motions of Jupiter, Saturn, and Venus from the first year of the First Emperor of Qin’s reign as King of Qin in 246 bce, through the fall of Qin, down to the third year of Emperor Wen of Han in 177 bce. Jupiter’s synodic period is said to be 395.44 days, its sidereal period twelve years; Saturn’s periods are said to be 377 days and thirty years; and Venus’s 548.40 days and eight years. These documents were found in the tomb of an official of a kingdom far from the imperial capital, together with a number of other manuscripts dealing with divination, mantic physiological practices, and medicine. Obviously, information of this kind was not a monopoly of the imperial court. Sima Qian alludes to his personal study of a century of detailed astronomical observations of the planets and of those handed down by many earlier scribe-astrologers: one can imagine what riches the Han imperial library must have contained by way of astronomical sources.

Cosmography and cosmology Mention was made of the “Canopy Heaven” gai tian 蓋天 conception of the structure of the cosmos as resembling the round, umbrella-like carriage canopy commonly depicted in stone engravings of the Qin and Han periods. The concave canopy, usually decorated, had a number of ribs that converged on the central apex. A popular myth, already current in the Warring States period, recounted the battles with another cosmic miscreant, Gong, who knocked the pivot of 507

David Pankenier

the heavens askew so that it was no longer directly overhead.The stationary surface of the earth, flat in some versions, concave in others, was parallel to the sky canopy, only a portion of which was visible at a time. During the daylight the Sun rotated into view and during the night the moon and stars.The invisible axis of rotation of the starry dome of the sky connected the North Pole and the center of the earthly plane.29 Naturally, by the early imperial period this earthly center was the capital, since the emperor at the apex of the universal empire was the Supernal Lord’s terrestrial counterpart and, notionally, Heaven’s “son.” In a physical realization of the connection, the Qin and Han imperial capitals were deliberately laid out to mimic the circumpolar constellations and the Milky Way, in this symbolic way laying claim to, or at least invoking, the cosmic power (Figure 23.3). The gai tian theory was probably already being expounded long before the founding of the Qin Dynasty in 221 bce. It is described in detail in the early Han text Mathematical Canon of the Zhou Gnomon, Zhou bi suan jing 周髀算經. The sun’s changing distance from the Pole throughout the year is computed, as well as the radius of the sun’s orbit. The implication is that this oscillation produces the seasons. Employing the gou-gu theorem (勾股定理, the Chinese equivalent of the Pythagorean theorem), the sky was found to be 80,000 li above the surface of the earth. Apart from the discussion of cosmography, the Zhou bi suan jing is mainly concerned with computational procedures and methods involving the length of the shadow of a gnomon of standard height.30 It was about this time too that the new term yuzhou 宇宙 came into prominence, with a meaning more or less equivalent to “cosmos.” It draws on an architectural metaphor of “cosmic eaves and rafters,” resembling the ribs of a parasol, to express the idea of time and space and is especially prominent in the very important Huainanzi chapters dealing with astronomy and cosmology.31

Figure 23.3 The Supernal Lord in imperial garb attended by spirit officials and driving the Dipper like a carriage. According to the “Treatise”: “The Dipper is the Supernal Lord Di’s carriage. It revolves about the center, visiting and regulating each of the four regions. It divides yin from yang, establishes the four seasons, equalizes the Five Elemental Phases, deploys the seasonal junctures and angular measures, and marks the several [celestial] pendants: all these are tied to the Dipper.” After Feng and Feng (1893)

508

Astronomy

It is probably no coincidence that the successor theory to the “Canopy Heaven,” the “Spherical Heaven” hun tian 渾天 theory, made its appearance and supplanted the former soon after the equatorial armillary came into use. Early sources even attribute its invention to Luoxia Hong, who was the first historical figure we know of to use an armillary. Indeed, that device soon came to be called the hun tian yi 渾天儀, or “Spherical Heaven Instrument.” In the hun tian scheme, the earth is a flat plane at the center of a continuous sphere that completes one rotation daily. The pivot of the sphere is inclined at an angle of 35° above the earth-plane, which corresponds to the latitude of the imperial capitals. The capital and the observer, indeed the entire world known to the Chinese, are essentially located at the center of a vast sphere. The most complete account of this geocentric theory was written down about 117 ce by the remarkable Eastern Han polymath and court astronomer Zhang Heng 張衡 (78–139 ce). In his treatise on the armillary, Hun yi zhu 渾儀注, Zhang asserted: The heavens are like a hen’s egg and as round as a crossbow bullet; the earth is like the yolk of the egg, and lies alone in the centre. Heaven is large and earth small. Inside the lower part of the heavens there is water. The heavens are supported by qi (vapour), the earth floats on the waters. The circumference of the heavens is divided into 365¼ du; hence half of it, 182 5/8 du, is above the earth, and the other half is below. This is why, of the 28 hsiu (equatorial star groups) only half are visible at one time. The two extremities of the heavens are the north and south poles; the former, in the middle of the sky, is exactly 36 du above the earth, and consequently a circle with a diameter of 72 du encloses all the stars which are perpetually visible. A similar circle around the south pole encloses stars which we never see. The two poles are distant from one another 182 du and a little more than half a degree. The rotation goes on like that around the axle of a chariot.32 Zhang is also famous as the brilliant inventor and fabricator of a clepsydra and gear-driven armillary sphere, a celestial globe, and an odometer-carriage, as well as the world’s first seismoscope, reputedly capable of indicating the direction from which seismic waves come. His equatorial armillary was the first to incorporate horizon and meridian rings. Using it, he catalogued over 2,500 stars and 124 constellations. In his book, the Numinous Efficacy Manifested, Ling xian 靈憲 (120 ce), Zhang also produced the clearest explanation of the celestial mechanics of eclipses. The Sun is like fire and the Moon like water. The fire gives out light and the water reflects it. Thus the moon’s brightness is produced from the radiance of the Sun, and the Moon’s darkness is due to (the light of) the sun being obstructed. The side facing the Sun is fully lit, and the side which is away from it is dark. The planets (as well as the Moon) have the nature of water and reflect light. The light pouring forth from the Sun does not always reach the moon owing to the obstruction of the earth itself – this is called ‘an xu’ [lit. “dark emptiness” 暗虛], a lunar eclipse. When (a similar effect) happens with a planet (we call it) an occultation; when the Moon passes across (the Sun’s path) then there is a solar eclipse.33 The Moon and the planets are Yin; they have shape but no light. This they receive only when the Sun illuminates them. The former masters regarded the Sun as round like a crossbow bullet, and they thought the Moon had the nature of a mirror. Some of them recognized the Moon as a ball too. Those parts of the Moon which the Sun illuminates look bright, those parts which it does not, remain dark.34 509

David Pankenier

Zhang clearly compares the earth to a sphere – the yolk of an egg floating on water – but although he hints at a shadow phenomenon he does not explain precisely how the earth prevents the light of the sun from reaching the moon or why the moon is not completely dark. Jing Fang 京房 more than a century earlier had already alluded to the fact that some of these views were old. The term Zhang coined to denote the phenomenon, “dark emptiness ~ obscurity” is unique and not the usual word for shadow. A third school of thought, of unknown origins, was called “Obscure Night” Xuan Ye 宣夜. It is much less well documented but evidently influenced Zhang Heng’s thinking and that of other Han scholars. It held that the universe is formless and without bounds and that the celestial bodies move about impelled by the materia vitalis. The independent movements of sun, moon, planets and stars and their varying periods were said to be a consequence of their individual natures. It follows that they cannot be attached to a surface. In Zhang Heng’s words, “what is beyond the celestial no one knows. It is called the ‘cosmos’ yu zhou. It has no end and no bounds.”35

The ‘Treatise on the Celestial Offices’ A certain disorderly arrangement and sketchy coverage of some topics in Sima Qian’s “Treatise,” by comparison with the specialized Mawangdui silk manuscripts, “Prognostications of the Five Planets,” and “Diverse Prognostications on Heavenly Patterns and Formations of Materia Vitalis,” suggests that the received version of the text is merely a summary.This is particularly true of the section devoted to astrological history. In his postface to the Grand Scribes’ Records, Sima Qian has the following to say concerning the historical materials he and his father consulted: The writings concerning the astral bodies and materia vitalis [contain] many and varied inauspicious and propitious omens, but are unorthodox. [We have] inferred their patterns, investigated their correspondences, apart from the outliers, collating them all [to permit] discussion [in reference to their corresponding] deeds and events, verifying them in terms of paths and measures, and ordering them sequentially in compiling the Treatise on the Celestial Offices.36 Although Sima Qian asserts a self-conscious theory formation in advancing their astrological theory of history, almost none of the “discussion” he alludes to can be found in the text. “[Having explored events] from beginning to end, from ancient times to the present, [we have] looked deeply into the vicissitudes of the times, examining the minute and the large-scale,” and “[we have] investigated the boundary between the celestial and the human, thereby to comprehend historical change, past and present, in order to formulate our own thesis [yi jia zhi yan 一家之言],” again barely hints at the deeper meaning of the lessons the Simas drew from their research.37 Given their position, such lessons would probably have been communicated to a select few. For this reason it is possible that the received version of the “Treatise” was not intended to be any more than an executive summary for Emperor Wu and certain court officials. The very title “Celestial Offices” evokes the parallelism we have seen between the stellar patterns and the imperial offices and departments of the Celestial Empire. It was the prevailing conception that celestial events had ramifications on Earth and, conversely, that temporal affairs could perturb the cosmos, such “mutually resonant” (ganying 感應) influences being accomplished by means of the long-established regime of astral-terrestrial correspondences and through the medium of materia vitalis. It was thought that the signs of this reciprocal interaction between Heaven and Man could be discerned at an incipient stage and that the study of historical precedents and the maintenance of careful records could aid in the interpretation of such 510

Astronomy

omens. Observation was crucial to the welfare of the state, so that then as later the emperor had to be kept regularly and accurately informed of celestial events, especially the unanticipated or anomalous. That is why the Simas’ official title Tai shi ling 太史令 is best rendered in English as “Prefect Grand Scribe-Astrologer,” the two responsibilities being functionally inseparable. Although it is the astrological motive that principally informs the “Treatise” and lends it its particular flavor, here are the principal astronomical highlights of what the “Treatise” accomplishes. Sima Qian • • • • •

“imperializes” the names of asterisms, identifying by name and relative location eighty-nine asterisms and some 412 individual stars, apparent magnitude, color, and variability. provides regional variations in the names of stars, asterisms, and prognostics. gives a detailed account of the planetary motions and periods, asserting that all five visible planets retrograde. provides an account of numerous anomalous astral phenomena and their associated prognostications, including the earliest description of the zodiacal light. reformulates the prior “field allocation” astral-terrestrial system as a yin-yang binary scheme, taking account of the new universal empire (yang) and its adversarial relations with frontier peoples (yin).38

Here is a sample from the beginning of the “Treatise’s” discussion of the circumpolar palace with which Sima Qian begins: The brightest among the Celestial Culmen stars in the Central Palace is the constant abode of the Supreme One. The three stars beside it are the Three Eminences, otherwise called the Crown Prince and Cadet Princes. The large star at the end of the curve formed by the four stars in back is the Official Consort; the other three stars are [concubines] belonging to the Inner Palace. The twelve stars framing and guarding [the celestial Pole] are Screening Ministers. Collectively, they are all referred to as the Purple [Tenuity] Palace . . . Ahead of the Dipper’s Bowl is a frame of six stars, [the Privy Council], called the Palace for Promotion of Civic Virtue; the first is called Generalin-Chief, the second is called Lieutenant General, the third is called Chancellor, the fourth is called Overseer of Fate, the fifth is called Overseer of the Middle Way, the sixth is called Overseer of Ranks and Emoluments. Within the Bowl of the Dipper is a Prison for the nobility.39 Here is an example of applied astral portentology at the official level Sima Qian’s time: During the Shiyuan reign period of Emperor Xiao Zhao [of Han; r. 87–74 bce], the Han eunuch Liang Chenghui and the King of Yan’s stargazer Wu Moru saw a fuzzy star [tailless comet] emerge in the west at the east gate of Celestial Market [in Ophiuchus; March, 84 bce]; it traveled past River Drum [Aquila] and entered Align the Hall [Aquarius–Pisces]. [Liang Cheng-]hui said, “when a fuzzy star appears for sixty days, before three years have passed there will be rebellious officials executed in the marketplace below.”40 In another instance, the significance of otherwise unremarkable movements of Venus through the Celestial Court (roughly Leo) only subsequently became clear after a plot to assassinate imperial regent Huo Guang 霍光 (d. 68 bce) was discovered in 80 bce, leading to the public 511

David Pankenier

execution of a king, two generals, and the senior princess. Consequently, the following prognostic was pronounced: Taiwei is the Celestial Court. When Venus moves into it, the gates of the palace are closed, great generals don armor and weapons, and evil officials submit to punishment. The “Treatise” remains the most authoritative and comprehensive account of astronomy and astral portentology down to the Han dynasty. Although astral cartography and instrumentation made great strides in subsequent centuries, the astronomical foundations laid in the early imperial period established the polar-equatorial framework and non-geometric mathematical approaches which remained essentially unchanged until late imperial times.

Notes 1  Pottery sherds with solar images (with halo) from Dahecun 大河村 are in Li (1983). Very early such pottery signs from the Shuangdun 雙墩 site at Bengbu 蚌埠 date from ca. 7,000 bp; see Xu (2008), 78. 2 See Feng (2007), 278, 285; also Pankenier (2013), 338. 3 Needham and Wang (1959), 248, 282. 4 Jao (1998), 33, 36, 39. 5 This burial is a millennium older than the solar observation terrace at Taosi 陶寺 (discussed later), where some burials (e.g. TG5M23, TG4M24) exhibit the same orientation toward winter solstice sunrise. 6 For the Yaoshan 瑶山 (Liangzhu) hilltop altar platform, positioned in the landscape to afford a clear view of solstice sunrises and sunsets, see Pankenier (2013), 30. 7 The inscribed jade plaque from Hanshan, Lingjiatan, found sandwiched between the two shells of a carved jade turtle, is in Anhui sheng wenwu kaogu yanjiusuo (1989); for analysis, see Feng (2007), 505. 8 Smith (2011). The fact that the Taosi astronomers were already working to synchronize the solar and lunar cycles shows that they were developing a luni-solar calendar eight centuries before the Shang. 9 Xu, Pankenier and Jiang (2000). 10 Jao (1998), Feng (2007), 137ff. 11 Jao (1998). 12 NASA Eclipse Web Site: http://eclipse.gsfc.nasa.gov/SEsearch/SEsearchmap.php?Ecl=-11601031 13 Xia Shang Zhou duandai gongcheng bianji weiyuanhui (2000), 52. Liu (2002), 4. http://eclipse.gsfc.nasa. gov/LEcat5/LE-1299–1200.html 14 See Pankenier (2013), 71, 74. 15 Trans. Pankenier (2013), 57. cf. Needham and Wang (1959, 245). The best recent studies of the “Canon of Yao” are Hwang (1996), 607–10, and Liu, Q-Y (2004).The oracle-bone inscriptions leave little doubt that the Shang were observing and sacrificing to the sun on the solstices, and probably other dates as well. 16 See Pankenier (2013), 194. 17 Trans. Pankenier (2013), 218. 18 See Pankenier (2013), 58. 19 Pankenier (2013), 38ff. 20 Ecsedy (1989). 21 Pankenier (2014). 22  Both books survive only as scattered quotations in the Kaiyuan zhan jing 開元占經 (Kaiyuan Reign Period Treatise on Astrology), compiled between 718 and 726 ce.To Gan De has been attributed a possible naked-eye observation of a moon of Jupiter; see Xi (1982), Hughes (1982). 23 Painting the twenty-eight lodges, sun and moon on tomb ceilings was common in Han dynasty elite tombs, often including the Milky Way. For an important example, see Luo (1991). 24 Maeyama (2002), Sun and Kistemaker (1997). For an illustration of a Han period mantic-astrolabe shi 式 inscribed with the 28 lunar lodges and the solar chronograms, see Pankenier (2013), 335. 25 The ancient enumerator chen 辰 (fifth in the sequence of twelve so-called earthly branches dizhi 地支) in this context refers to the series of signs used to denote the positions (here Jupiter stations) occupied by the Sun each month.The twenty-eight lunar mansions were apportioned among the twelve by twos

512

Astronomy and threes. For traditional reasons, the enumeration of the lunar lodges always began with Horns (角 Spica α Virgo) whose achronical rising in spring anciently marked the start of the lunar year, while the solar calendar began with the winter solstice in the sign zi 子. 26 Pankenier (2013), 509. Sima (1959), 27.1350. 27 Trans. Pankenier (2013), 257. 28 Cullen (2011a-b). 29 Needham and Wang (1959), 210ff. 30 Cullen (1996). 31 Major (1993). 32 Trans. Needham and Wang (1959), 217 (modified). 33 Trans. Needham and Wang (1959), 414 (modified). 34 Needham and Wang (1959), 227 35 Pankenier (2013), 369. 36 Sima (1959), 130.3306.Trans. Pankenier (2013), 428. A complete annotated translation of the “Treatise” is an appendix to Pankenier (2013), 444–511. 37 Pankenier (2013), 510, 455. 38 See Pankenier (2013), 444. The original is a planisphere inscribed on a stele from the Northern Song dynasty (860–1126 ce). See Zhongguo shehui kexueyuan kaogu yanjiusuo (1980). 39 Trans. Pankenier (2013), 459–60. 40 Ban (1962), 26.1306.

Works cited Bo, Shuren. (1985). “Sima Qian – The Great Astronomer of Ancient China”, Chinese Astronomy and Astrophysics, 9.4: 261–7. British Library (Ch.85.XIII), mid-seventh century star atlas scroll discovered in the early twentieth century in a Mogao 莫窟 cave at Dunhuang 敦煌. http://idp.bl.uk/database/oo_scroll_h.a4d?uid=190006978 11;recnum=8280;index=12 Chen, J-J. 陳久金. (1978). “Cong Mawangdui boshu Wu xing zhan de chutu shitan wo guo gudai de suixing jinian wenti” 從馬王堆帛書《五星占》的出土試探我國古代的歲星紀年問題, in Zhongguo tianwenxueshi wenji bianjizu 中國天文學史文集編輯組 (ed.), Zhongguo tianwenxue shi wenji 中國天 文學史文集, Liaoning jiaoyu, 48–65. Chen, Meidong 陳美東. (1996). Zhongguo gu xing tu 中國古星圖 (“Ancient Chinese star charts”). Shenyang: Liaoning jiaoyu. Cullen, C. (1976). “A Chinese Eratosthenes of the Flat Earth: A Study of a Fragment of Cosmology in Huai Nan Tzu”, Bulletin of the School of Oriental and African Studies, 39.1: 106–27. Cullen, C. (1993). “Motivations for Scientific Change in Ancient China: Emperor Wu and the Grand Inception Astronomical Reforms for 104 B.C”, Journal for the History of Astronomy, 24.3: 185–203. Cullen, C. (2000). “Seeing the Appearances: Ecliptic and Equator in the Eastern Han”, Studies in the History of Natural Sciences, 19.4: 352–382. de Saussure, L. (1930). Les origines de l’astronomie Chinoise, Paris: Maisonneuve. Gang, X. and Shi,Y-L. (2005). “On the Solar Eclipse Records from the Han Dynasty and their Application to the Analysis of the Precision of the Systems of Calendrical Astronomy of the Han Dynasty”, Chinese Journal for the History of Science and Technology, 26: 107–21. Henderson, John B. (1984). The Development and Decline of Chinese Cosmology, NewYork: Columbia University Press. Henderson, John B. (1995). “Chinese Cosmographical Thought: The High Intellectual Tradition”, in J.B. Harley and D. Woodward (eds.), The History of Cartography, Vol. 2, Book 2, Cartography in the Traditional East and Southeast Asian Societies, Chicago: University of Chicago, 203–27. Ho, Peng Yoke (1966). The Astronomical Chapters of the Chin-shu, Paris: Mouton & Co. Ho, Peng Yoke (2003). Chinese Mathematical Astrology: Reaching out for the Stars, London: Taylor & Francis. Hu, L-G. 呼林貴 (1989). “Xi’an Jiaoda Xi Han ershiba xiu xingtu yu Shiji: Tian guan shu” 西安交大西漢 二十八宿與《史記》天管書, Renwen zazhi 人文雜誌, 2: 85–7. Hu, T-Z. 胡鐵珠 (2000). “Xia xiao zheng xingxiang niandai yanjiu” 《夏小正》星象年代研究, Ziran kexueshi yanjiu 自然科學史研究, 19.3: 234–50. Jiang, X-Y. 江曉原 (2011). Tianxue zhen yuan 天學真原 Nanjing:Yilin.

513

David Pankenier Kalinowski, M. (1986). “L’Astronomie des populations Yi du Sud-Ouest de la Chine”, Cahiers d’Extreme Asie, 2: 253–63. Kelley, D.H. and Milone, E.F. (2011) Exploring Ancient Skies: A Survey of Ancient and Cultural Astronomy, New York: Springer. Lee, Y.K. (2002), “Building the Chronology of Early Chinese History”, Asian Perspectives: The Journal of Archaeology for Asia and the Pacific, 41.1: 15–42. Li, D. (2002).“The Place of Gaitian Theory in the History of Astronomy in China”, Nei Menggu shifan daxue xuebao (Ziran kexue ban) Inner Mongolia Normal University [Natural Science edition], 31.3: 281–9. Li, X-Q. (1992–93). “A Neolithic Jade Plaque and Ancient Chinese Cosmology”, National Palace Museum Bulletin, 27. 5–6: 1–8. Li, Y. 李勇 (1992). “Dui Zhongguo gudai hengxing fenye he fenye shipan yanjiu” 對中國古代恆星分野 和分野式盤研究, Ziran kexue shi yanjiu 自然科學史研究, 11.1: 22–31. Morgan, D.P. (2016). “The Planetary Visibility Tables in the Second-Century bc Manuscript Wu xing zhan 五星占”, East Asian Science, Technology, and Medicine, International Society for the History of East Asian Science,Technology, and Medicine (ISHEASTM), 43: 17–60. Needham, J. and Ronan, C. (1993).“Chinese Cosmology”, in N.S. Hetherington (ed.), Cosmology: Historical, Literary, Philosophical, Religious, and Scientific Perspectives, New York: Garland, 25–32. Pankenier, D.W. (2012). “On the Reliability of Han Dynasty Eclipse Records”, Journal of Astronomical History and Heritage, 15.3: 200–12. Pankenier, D.W., Xu, Z-T, and Jiang, Y-T. (2000). Archaeoastronomy in East Asia: Historical Records of Astronomical Observations of China, Japan and Korea, Amsterdam: Gordon & Breach. Selin, H. (2000). Astronomy Across Cultures:The History of Non-Western Astronomy, Dordrecht: Kluwer. Sivin, N. (1969). Cosmos and Computation in Early Chinese Mathematical Astronomy, Leiden: Brill. Sivin, N. (2011). “Mathematical Astronomy and the Chinese Calendar”, in J.M. Steele (ed.), Calendars and Years II: Astronomy and Time in the Ancient and Medieval World, Oxford: Oxbow, 39–52. Stephenson, F.R. (1994). “Chinese and Korean Star Maps and Catalogs”, in J.B. Harley and D. Woodward (eds.), The History of Cartography, Vol. 2, Book 2, Cartography in the Traditional East and Southeast Asian Societies, Chicago: University of Chicago, 511–78. Teboul, M. (1985). “Sur quelques particularités de l’uranographie polaire Chinoise”, T’oung Pao, 71: 1–39. Xi, Z-Z. 席澤宗 (1989). “Mawangdui Han mu boshu de huixing tu” 馬王堆漢墓帛書的彗星圖, in Shehui kexueyuan kaogu yanjiusuo (ed.), Zhongguo gudai tianwen wenwu lunji, Beijing: Wenwu, 29–34. Xi, Z-Z. and Chen, M. (2003). “Twentieth Century Chinese Scholarship in the History of Astronomy”, Guangxi minzu xueyuan xuebao 廣西民族學院學報, 10.1: 6–11. Xu, F-X. 徐鳳先 (1994). “Zhongguo gudai yichang tianxiang guan” 中國古代異常天象觀, Ziran kexueshi yanjiu 自然科學史研究, 2: 201–8. Xu, F. and He, N. (2010). “Taosi Observatory, China”, in C. Ruggles and M. Cotte (eds.), Heritage Sites of Astronomy and Archaeoastronomy in the Context of the UNESCO World Heritage Convention, ICOMOS – IAU Thematic Study on Astronomical Heritage (June), 86–90. Yi, S-T. 伊世同 (1981). Zhong-xi duizhao hengxing tu biao 1950.0 中西對照恆星圖表 (Maps and tables with Chinese and Western fixed stars juxtaposed, epoch 1950.0). 2 vols, Beijing: Kexue chubanshe. Zhang, P. (2002). “Determining Xia – Shang – Zhou Chronology Through Astronomical Records in Historical Texts”, Journal of East Asian Archaeology, 4: 335–57. Zheng, W-G. 鄭文光 (1979). Zhongguo tianwenxue yuanliu 中國天文學源流, Beijing: Xinhua. Zhongguo shehui kexueyuan kaogu yanjiusuo 中國社會科學院考古研究所 (ed.) (1980). Zhongguo gudai tianwen wenwu tuji 中國古代天文文物圖集, Beijing: Wenwu. Zhongguo tianwenxueshi wenji bianjizu 中國天文學史文集編輯組 (ed.) (1978). Zhongguo tianwenxue shi wenji 中國天文學史文集, Beijing: Kexue.

Suggestions for further reading Anhui sheng wenwu kaogu yanjiusuo 安徽省文物考古研究所 (1989). “Anhui Hanshan Lingjiatan xinshiqi shidai mudi faxian jianbao” 安徽含山凌家灘新石器時代墓地發現簡報, Wenwu 文物 4: 1–9. Ban, G. 班固 (1962). Han shu 漢書 (History of the Former Han Dynasty), Beijing: Zhonghua. Cullen, C. (1996). Astronomy and Mathematics in Ancient China: The Zhou Bi Suan Jing, Cambridge: Cambridge University Press. Cullen, C. (2011a).“Understanding the Planets in Ancient China: Prediction and Divination in the Wu xing zhan”, Early Science and Medicine, 16: 218–51.

514

Astronomy Cullen, C. (2011b). “Wu xing zhan 五星占 ‘Prognostics of the Five Planets’ ”, Sources and Commentaries in Exact Sciences 12: 193–249. Ecsedy, I., Barlai, K., Dvorak, R., and Schult, R. (1989) “Antares Year in Ancient China”, in A.F. Aveni (ed.), World Archaeoastronomy, Cambridge: Cambridge University, 183–6. Feng, S. 馮時 (2007). Zhongguo tianwen kaoguxue 中國天文考古學, Beijing: Zhongguo shehui kexue. Feng, Y-P. 馮雲鵬, and Feng, Y.-y. 馮雲鵷, 1893. Jinshi suo 金石索 Shanghai: Shanghai jishan shuju, first published 1821; also available at http://catalog.hathitrust.org/Record/002252003. Hughes, D.W. (1982). “Was Galileo 2,000 Years Too Late?” Nature, 296: 199. Hwang, M-C. (1996). “Ming-Tang: Cosmology, Political Order, and Monuments in Early China,” Ph.D. dissertation, Harvard University. Jao, T-Y. 饒宗頤 (1998). “Yin buci suojian xingxiang yu shen shang, long hu, ershiba xiu zhu wenti” 殷卜 辭所見星象與參商、龍虎、二十八宿諸問題, in Zhang, Y-S. 张永山 and Hu, Z-Y. 胡振宇 (eds.), Hu Houxuan xiansheng jinian wenji 胡厚宣先生纪念文集, Beijing: Kexue, 32.5: 37. Li, C-T. 李昌韜 (1983). “Dahecun xinshiqi shidai caitao shang de tianwen tuxiang 大河村新石器時代彩 陶上的天文圖像”, Wenwu 文物 8: 52–4. Liu, C-Y. (2002). “Astronomy in the Xia-Shang-Zhou Chronology Project”, Journal of Astronomical History and Heritage, 5.1: 1–8. Liu, L-X. 劉樂賢 (2004). Mawangdui tianwen shu kaoshi 馬王堆天文書考釋, Guangdong: Zhongshan daxue. Liu, Q-Y. 劉起釪 (2004). “Yaodian Xi He zhang yanjiu” 《堯典》羲和章研究, Zhongguo shehui kexue yuan lishi yanjiusuo xuekan 中國社會科學院歷史研究所學刊, 2: 43–70. Luo, Q-K. 雒启坤 (1991). “Xi’an Jiaotong daxue Xi han muzang bihua ershiba xiu xingtu kaoshi” 西安 交通大學西漢墓葬壁畫二十八宿星圖考釋, Ziran kexue shi yanjiu 自然科學史研究, 10.3: 236–45. Maeyama, Y. (2002). “The Oldest Star Catalog of China, Shih Shen’s Hsing Ching”, in Y. Maeyama (ed.), Astronomy in Orient and Occident: Selected Papers on Its Cultural and Scientific History, Hildesheim: Georg Olms, 1–34; rpt. from Y. Maeyama and W. Saltzer (eds.) (1977) Prismata: Festschrift für Willy Hartner, Wiesbaden. Major, J.S. (1993). Heaven and Earth in Early Han Thought, Albany: State University of New York. Needham, J., with the research assistance of Wang L. (1959). Science and Civilisation in China,Vol. 3. Mathematics and the Sciences of the Heavens and the Earth, Cambridge: Cambridge University Press. Pankenier, D.W. (2013). Astrology and Cosmology in Early China: Conforming Earth to Heaven, Cambridge: Cambridge University Press. Pankenier, D.W. (2014). “Did Babylonian Astrology Influence Early Chinese Astral Prognostication, xing zhan shu?” Early China, 37.1: 1–13. Sima, Q. 司馬遷 (1959). Shiji 史記, Beijing: Zhonghua. Smith, A.C. (2011). “The Chinese Sexagenary Cycle and the Ritual Foundations of the Calendar”, in J.M. Steele (ed.), Calendars and Years II: Astronomy and Time in the Ancient and Medieval World, Oxford: Oxbow, 1–38. Sun, X-C. and Kistemaker, J. (1997). The Chinese Sky During the Han: Constellating Stars and Society, Leiden: Brill. Xi, Z-Z. (1982). “The Discovery of Jupiter’s Satellite Made by Gan De 2,000 Years Before Galileo”, Chinese Physics, 2.3: 664–7. Xia Shang Zhou duandai gongcheng bianji weiyuanhui 夏商周断代工程編輯委員會 (2000). Xia Shang Zhou duandai gongcheng 1996–2000 chengguo baogao: jianben 夏商周断代工程1996–2000年阶段成果 报告: 简本 [The Xia-Shang-Zhou Chronology Project Report for the years 1996–2000 (abridged)], Beijing: Shijie tushu. Xu, D-L. 徐大立 (2008). “Bengbu Shuangdun yizhi kehua fuhao jianshu” 蚌埠雙墩遺址刻畫符號簡述, Zhongyuan wenwu 中原文物, 3: 75–9. Xu, Z-T, Pankenier, D.W. and Jiang,Y-T. (2000). Archaeoastronomy in East Asia: Historical Records of Astronomical Observations of China, Japan and Korea, Amsterdam: Gordon & Breach.

515

INDEX

Note: Page numbers in italics indicate figures and page numbers in bold indicate tables on the corresponding pages. Agricultural Archaeology 302 agriculture 300 – 3; agronomic developments and 307 – 9; on the Central Plain 39 – 42, 40; climate and ecology 303 – 4; diets in different regions 304 – 5; farm and peasant life and 310 – 13; malefemale differences during the Neolithic and 48 – 9; religion and 313 – 15; shift toward wheat and barley in Yangshao 52; silviculture and 310; stable isotope perspective on Yangshao diets and 42 – 9, 45, 46, 47 agronomy 307 – 9 Allan, Sarah 380 – 1 All-Under-Heaven see politics Analects see Lunyu ancestral cult 238 – 40, 375 animal husbandry 304 – 7; animal bones in the Chinese Neolithic as evidence of 20 – 1; in the Central Plain 39 – 42, 40, 52 – 6, 54, 55 antiquity, end of 191 – 3 Anyang: excavations at, 1928 – 37, 1950–present 64 – 6; excavations at Zhengzhou as local antecedent to 66 – 8; human sacrifice and 75 – 6; oracle inscriptions at, connecting traditional written record with material remains 63 – 4, 74 – 5 archery 133 aristocratic lineages and politics 287 – 90 art: Bronze Age 433 – 50, 454, 456; cursive writing 455 – 7, 456; Eastern Zhou 132 – 9, 134 – 5, 136 – 8; late Neolithic 425 – 3, 426 – 7, 429 – 32; Qin and Han 450 – 7, 451 – 2, 454 – 5 astral omens 498 – 507 astrology 493

astronomy 493; calendar, astral omens, and heavenly legitimation in 498 – 507; cosmography and cosmology in 497, 507 – 10, 508; first written records on 495 – 8; Four Quarters conception 502 – 4, 503; Neolithic beginnings of 493 – 5; Spring and Autumn period through Warring States period 502 – 4, 503; “Treatise on the Celestial Offices” 510 – 12; Warring States period through Han 504 – 7 Baez, Joan 408 Bai, Lord 125 Bai, Shouyi 118 Bai Gong 130 Balanced Discourses see Lunheng Bamboo Annals see Zhushu Jinian Ban Gu 187 – 8, 190, 421 – 2, 483 banliang 351, 351 – 2; unification and 353 – 5 Banpo site 45 – 6 Ban Zhao 381 Baoshan Tomb 2 243 Barbieri-Low, Anthony 386 barley agriculture in Yangshao 52 battlefields 140, 142, 141 Battle of Changping 326 Baxter, William H. 206, 207, 211, 214 bean agriculture, development of 42 Biannian ji (Chronicle) 387 Biao Qiang 114, 115 Bielenstein, Hans 184, 185 bird hunting 133 Bodde, Derk 314 Book of Changes see Yi

517

Index Book of Fan Shengzhi 308 Book of Huang Shigong,The 186 – 7 Book of Lord Shang see Shangjun shu Book of Submitted Doubtful Cases see Zouyan shu Boyer, Pascal 234 Bray, F. 42 Bronze Age 4, 5, 22, 26, 35; art 433 – 50, 454, 456; before the Zhou dynasty 61 – 79 Brown, Miranda 467 Buddhism 1, 277 bureaucracy and medicine 466 – 8 burial sites see graves and tombs Bu Shi 173 – 4

development of social complexity in 25 – 31, 27 – 8; early villages and village life in 21 – 5, 23; gender relations in 48 – 9; legacy of 34 – 5; processes during the pre-Neolithic period as deep background to 16 – 21, 17, 19 – 20; regional variation and cross-cultural interactions in 31, 31 – 3; women in 367 – 70 Chronicle see Biannian ji Chuci 307, 419 – 20 Chunqiu 118, 283, 409, 413 “Chunqiu Dashi Biao” 118 Chunyu Yi  469 Chu state 118, 122 Civil War,The 319 Classic of Law 388 Classic of the Vessel 468 climate and ecology effects on food 303 – 4 coins: banliang 351, 351 – 5; changing the material of 355, 355; Wang Mang’s reforms to 358 – 62, 359 – 62; wuzhu 356 – 7, 356 – 8 collapse of late Neolithic societies 33 commercialization and production in Eastern Zhou 128 – 32,  131 Complete Writings of the Four Treasuries see Siku quanshu Confucius 6, 122, 153, 217, 264, 266, 267, 280, 281, 409, 506; Analects 266, 268, 412 – 13; bamboo documents 223; on existence of spirits 252, 268; “Five Classics” 173, 174 – 5, 177, 296; on the intellectual and the state 291 – 2; Spring and Autumn Annals 409, 413; Wang Mang’s monetary reforms and 358 – 9 Cook, C. 374, 379 cosmic nature of ritual 273 – 6 cosmography 507 – 10,  508 cosmology 497, 507 – 10,  508 court, medicine at 468 – 70 cowries, sea 336 – 9,  337 cribra orbitalia 51 criminal procedures during the Qin and Han era 395 – 401,  396 crossbows 326 cross-cultural interactions, Neolithic 31, 31 – 3 Cui Zhu 6 Cullen, C. 475, 487, 490 currency: ancient spades and ancient knives as 341, 341 – 4, 343 – 4, 346, 346 – 50, 348, 350; banliang 351, 351 – 5; birth of money 336 – 44, 337, 340 – 1, 343 – 4; changing the material of 355, 355; ideal 355 – 7, 355 – 62, 359 – 62; increase in production of 339 – 41, 340; intrinsic value and its limits 349 – 52, 350 – 1; monetarization of society and 344 – 5, 345 – 6, 348 – 51, 354; seacowries as 336 – 9, 337; unification of 353 – 5; Wang Mang’s reforms to 358 – 62, 359 – 62; Warring States period 344 – 9, 345 – 6, 348 – 51, 354; wuzhu 356 – 7, 356 – 8

Caesar, Julius 319 Cai Yong  422 calendar and astronomy 498 – 507 Cambridge History of Ancient China 111 Campbell, R. 372 Cang Jie 217 Cao Cao 1 Cato 302 cavalries 325 Central Asia, Former Han empire expansion into 171 – 2 Central Plain, the: animal story in 52 – 6, 54, 55; currency in 342 – 3; dental caries in 49 – 50, 50, 50 – 1; development of agriculture and animal husbandry on 39 – 42, 40; dietary variation in 44 – 8; geography of 39; grave sites 43 – 4; malefemale differences during the Neolithic in 48 – 9; porotic hyperostosis and cribra orbitalia in 51; shift toward wheat and barley agriculture in 52; stable isotope perspective on diets in 42 – 9, 45, 46, 47; women in 368 ceramic wares 26, 27 Chahai site 22, 23 Chang, K. C. 32 Chao Yuen Ren 204 chariots 323 – 4, 327; artwork 451, 451 – 3 Chemla, Lin 476 Chen, Chusheng 337 Chen, Kang 337 Chen, Pan 118 Chen, Qiyou 274 Cheng, Junying 263 Cheng Qiao 353 Chengtoushan site 29 Chen She 156 – 7, 162 Chen Siting 270 Chen state 121 Chen Wei  244 China, defining 2 – 5 Chinese Neolithic, the 15 – 16, 16; art of 425 – 33, 426 – 7, 429 – 32; beginnings of astronomy in 493 – 5; Central Plain during (see Central Plain, the); collapse of late societies in 33;

518

Index cursive writing 455 – 7,  456 cypraea moneta 338

Eastern Zhou dynasty 89, 103, 105, 108, 110, 111; commercialization and production in 128 – 32, 131; currency 339 – 40; farm and peasant life in 317; life in peace and war 132 – 9, 134 – 5, 136 – 8; sources on 109 – 13; sub-periods 109 – 10; toward unification of 139 – 42, 141; urbanization and population in cities of 125 – 8, 127; women in 376 – 81; see also territorial lords EBA see Early Bronze Age (EBA) Educating Women see Jiao nü education of women 381 Egypt, ancient 77, 77, 77 – 8,  221 Ehrenberg, Margaret 369 Eighteen Qin Statutes see Qinlü shibazhong Engels, F. 48 Eno, Robert 238, 375 – 6 Erligang empire 69 – 70 Ernian lüling 387, 389 – 90, 391, 473 Essential Techniques for the Peasantry see Qimin yaoshu Expedition of Cyrus,The 319

dafu aristocrats 117, 120 Daoism 1, 277, 414; motivation for war and 330 Dawenkou site 46; pottery 428, 429 De Agricultura 302 death see graves and tombs Debate on Salt and Iron see Yantie lun Deng Man 410 Deng Mu 287 Denouncing Cursive 456 – 7 dental caries 49 – 50, 50, 50 – 1 dietary variations in the Yangshao sites 42 – 9, 45, 46, 47, 53, 54 Ding constellation 499, 500 Di Xin, King 84 – 5, 239, 321 Discourses of the States see Guoyu Documents see Shu domesticated animals see animal husbandry Dong, Kun 337 Dong Hu 6 – 7 Dong Zhuo 180, 183, 184 – 5, 189, 191 Dong Zuobin 64, 225 Dou Gu 187 – 8 Dou Wu  190 Dou Xian 187 – 8, 190 Downer, G. B. 211 Dreyer, Edward 329 Drought of the Great King of Jian,The see Jian dawang bohan Duàn Yùcái  205 Du Mu 154 Durrant, Stephen 249, 265, 269, 330, 410

Falkenhausen, Lothar von 263, 373 – 4, 377 Fan, Hui 133 Fang, Qing 118 Fan Ye 181, 187 farming see agriculture Father spirits 238 – 40 fengjian system 117, 122 Five Forms of Conduct,The see Wuxing food 300 – 15; agronomic developments and 307 – 9; climate and ecology effects on 303 – 4; diets in different regions 304 – 7; farm and peasant life and 310 – 13; farming, and religion 313 – 15; silviculture and 310; sources and scholarly trajectories on 301 – 3 Former Han empire 160 – 8, 191; Early 161 – 8, 328; end of 177; Late 174 – 7, 329; Mid-168 – 74; military troops 328; warfare 326; see also Latter Han empire fortified sites 29 – 30; urbanization and 125 – 6 Four Quarters conception 502 – 4, 503 Four Types of Descriptions Including That of Lawsuits see Weiyu deng zhuang sizhong Fu Hao, Lady Hao 65, 71, 321, 371 – 2 funeral practices see graves and tombs Fu Xi 370

Early Bronze Age (EBA): Anyang oracle inscriptions in 63 – 4; Erligang empire and 69 – 70; excavations at Anyang, 1928 – 37, 1950–present 64 – 6; human sacrifice in 75 – 6; predynastic Zhou as statesmen or barbarians in 73 – 4; Sanxingdui pits and 70 – 1; scope, aims, and sources in study of 61 – 3; short history of archaeology in 63 – 74; unplanned archaeology in Yangtze region and 72 – 3; writing and literacy in 74 – 5; Xingan tomb and 71 – 2;Yanshi Erlitou excavations, 1959–present 68 – 9; Zhengzhou excavations as local antecedent to 66 – 8 Early China 1 early Chinese history: defining “Chinese” in 2 – 5; defining “early” in 1 – 2; defining “history” in 5 – 8 early Chinese writing 217; basic types and characteristics of characters in 220 – 1; content of 221 – 4; formation and development of characters for 217 – 20; reflections on script and 224 – 7

gai tian theory 508 Gallic War,The 319 Gan De 504 Gao, Q. 49 Gaozu, Emperor 166; system of submitting doubtful cases under 390 – 1 Gardner, Charles S. 5 Geling Tomb 1 244 gender relations during the Neolithic 48 – 9

519

Index Gnomon of the Zhou,The see Zhoubi Goldin, P. 374 Gong, King 337 gong aristocrats 117 – 18 Gong Ji 375 Gongsun Yang  114 Gong Yu  176 Gong Zhiqi 250 – 1 Gouwan site 44, 45 – 6 government and bureaucracy, Western Zhou 98 – 100 grammar, Old Chinese 208 – 10 Grand Scribe’s Records see Shiji graves and tombs 26, 28, 28; cong in 431 – 2, 432 – 3; cross-cultural interactions evidence in 31, 31 – 2; Qin and Han artwork on 451; Wang Fengshi 246 – 7; of women 374 – 5;Yangshao sites 43 – 4; of Zeng Houyi 111 – 12 Great Drought of Lu see Lubang dahan Great Wall 168, 171 grinding stones 18 – 19, 20 Guandong region 182, 189 Guanjia site 44, 48 Guanzhong region 182 – 3, 189 Guanzi 292, 302 Guishen zhi ming 267 Gu Derong 109, 139 Gu Donggao 118 Gu Ji 243 Gumbrecht, Hans Ulrich 230 – 1 Guo Moruo 225, 263 guoren 3, 327 Guoyu 6, 117, 411 – 12, 502 Guo Yu  469

Herodotus 409 hierarchies, social 28 – 9; politics and 287 – 90; territorial lords and 116 – 18 Historian’s Records see Shiji historical narrative 409 – 12 History of the Former Han Dynasty see Hanshu History of the Latter Han see Hou-Han shu Holocene, early 18 – 21; millet agriculture development in 41; women in 367 home medicine 461 – 5 Hongshan culture 233 – 5; pottery 430, 431 Hou-Han shu 180, 456 Huainan kingdom 166 – 7, 419 Huainanzi 300 Huang, Xiquan 337 Huang Zongxi 287 Huan Tan  176 Hua Tuo 469 – 70 Hui, King 148, 269 Hulsewé, A. F.P. 386 human health indicators,Yangshao 49 – 51, 50, 50 – 1 Human Nature Emerges from the Endowment see Xing zi ming chu human sacrifice 75 – 6 “Hundred Schools of Thought” 296 Hun yi zhu 509 Huo Guang 511 – 12 hyperporosity of cranial bones in Yangshao 51

Hafner, Arnd Helmut 393 Han dynasty 1, 147 – 8, 158 – 9; art 450 – 7, 451 – 2, 454 – 5; astronomy 504 – 7; commercialization and production in 128 – 9; cosmography and cosmology in 497, 507 – 10, 508; criminal procedure during 395 – 401, 396; currency 354; farm and peasant life in 312; latter 180 – 93; legal system (see legal system, Qin-Han); literature of 422; mathematics in 475 – 81; medicine in 466 – 8; Shennong deity of agriculture and 314; women in 380; Zhou Li and 129; see also Former Han empire; Latter Han empire Han Fei 307 Hanfeizi 416 – 17 Hanshu 187, 421, 483, 505, 506 – 7 Han state 119, 139 – 40; agronomic developments in 308 – 9; farming in 306; silviculture in 310 He, Emperor 190 heavenly legitimation 498 – 507 He Jin 190 Hemudu site 22 Henan sites 44

jade 234 – 5, 430, 430, 430 – 1 Jaspers, Karl 280 Ji, Xusheng 270 Jian dawang bohan 269 – 70 Jiang, Gang 128 Jiangou site 30 Jiang Yuan  262 Jiangzhai site 43 Jiao nü 381 Jia Sixie 306 Ji Kangzi 506 Ji Li 90 Ji Liang 250 Jing, Emperor 166, 168 Jing, Lord 112 – 13 Jing Gong, Lord 112 – 13 Jin Jing Gong, Lord 120 Jin state 113 – 14, 117, 119 – 20, 139 Jiuzhang suanshu 474 – 80; key difference in bodies of knowledge and related practices in 483 – 7, 486; meaning of chapters in 489 – 90; practices and epistemological values 482 – 3; right-angled

identity, Chinese 4 – 5 infantries 324 – 8 initial consonants, Old Chinese 206 – 8 interactions, Neolithic cultural 31, 31 – 3

520

Index triangle and the mathematics of the calendar in 487 – 9; topics covered in 481 – 2 Jizi’s Instructions for Women 381 Joyce, James 463 Ju state 122 Kaiyuan zhan jing 504 Kang, King 222 Kang, Lord 125 Karlgren, Bernhard 202 – 6 Keightley, David 98, 238, 239, 310, 368 – 9 Kern, Martin 373 Khayutina, M. 376 Kinney, A. 378 knives as currency 341, 341 – 4, 343 – 4, 346, 346 – 8,  348 Korean peninsula 175, 326 Kuang Heng 176 language, Old Chinese see Old Chinese (OC) language Lao Ai 149 – 50 Laozi (Daode jing) 226 – 7, 285, 330, 417 Latter Han empire 180 – 4; as end of antiquity 191 – 3; external challenges to 184 – 8; internal conflicts 189 – 91; military training in 329; see also Former Han empire Lee, K.Y. 49 Legalists 280 legal system, Qin-Han: adjudication against magistrates in 399 – 400; “Answers to Questions Concerning Qin Statutes” 388, 391; Book of Submitted Doubtful Cases 387, 391 – 2, 397 – 8, 400; conclusion on 401 – 2; criminal procedure in 395 – 401, 396; “Eighteen Qin Statutes” 387 – 8; introduction to 386 – 7; Longgang tomb no. 6 389; major content of submitted doubtful cases 390 – 1; “Miscellaneous Excerpts from Qin Statutes” 388; “Models for Sealing and Investigating” 388; original of trial document excavated from Tuzi shan 400 – 1; penal system and criminal procedure 393 – 5; role of imperial secretary in supervision in litigations 398 – 9; Shuihudi tomb no. 11 387 – 9; “Statutes on Auditing” 388; submitted doubtful cases in the Qin manuscripts in possession of the Yuelu Academy 392 – 3; unearthed legal materials and their nature 387 – 9; Zhangjiashan tomb no. 247 389 – 92 Legge, James 274, 379 Lei Haizong 192 Lewis, Mark E. 284 Li, Wai-yee 265, 269, 410 Liangchengzhen site 29 Liangzhu culture 233 – 5; pottery 431 – 2, 431 – 2 Li Chunfeng 475 Lienü zhuan (Biographies of Eminent Women) 381, 421

521

Li Fang-kuei 203 – 6 Liji see Records of Rites Li, King 93, 95 Li Ji 64 – 6 Li Kui 388 Li Ling 328 lineages in Western Zhou society 100 – 2, 102 Ling xian 509 Li Si 150 – 5, 157, 161 literature 405; emergence of poetry 405 – 8; historical narrative 409 – 12; masters 412 – 17; of unity 418 – 22 Liu, Xian 337 Liu An 168 Liu Bang 157, 158, 162 – 6, 168, 170, 174, 177 Liu Bosheng 181 Liu Cang 453 Liu Chang 166 Liu Che 169 Liu Fuling 174 Liu Hui 475 Liu Mu 456 Liu Qiang 453 Liu Xiang 381, 421 Liu Xie 405 Liu Xin 505 Liu Xiu 181, 183 – 8, 332; The Book of Huang Shigong 186 – 7 Liu Xu 192 Liu Yan  181 Liu Ying  166 lodge system, astronomy 502 – 4, 503 Longgang tomb no. 6 389 Longshan culture 25 – 6; pottery 428, 429 Long Wall of Qi 114 Lord Ling of Jin 5 – 7 Lord Mu of Lu Asked Zisi see Lu Mugong wen Zisi Lord Zhuang 6 Lou Hu 466 Lubang dahan 267 – 70 Lü Buwei 148 – 50, 353 Lu Mugong wen Zisi 223 Luhan people 124 Lunheng 421 Lunyu 266, 268, 412 – 13 Luoxia Hong 505 Luo Zhenyu 64 Lüshi chunqiu 149, 193, 290 – 1, 294 Lu state 122 Ma Chengyuan 222, 268, 337 Mair,Victor  415 Majiayao pottery 425 – 8, 426 – 7 marketplaces 129 – 30; medicine 465 – 6 Ma Rong 193 Marxist theory 311, 368 masters literature 412 – 17

Index Mathematical Canon of the Zhou Gnomon see Zhoubi suanjing mathematics: of the calendar and the right-angled triangle, differences in emphasis on 487 – 9; key difference in bodies of knowledge and related practices in 483 – 7, 486; manuscripts on 475 – 81; meaning of chapters in The Nine Chapters as key difference in practice of 489 – 90; practices and epistemological values 482 – 3; topics 481 – 2; two types of sources for history of early Chinese 473 – 5 Mathematics see Shu1 McGovern, P. E. 40 medicine 459 – 60, 470 – 1; bureaucracy and 466 – 8; at court 468 – 70; healing at home 461 – 5; market 465 – 6; temple as site of healing and 460 – 1 Mencius 311, 313, 331, 412 – 15 Meng Ao 149 Meng Tian 155 – 6 Mengzi 282 – 3, 291 – 3; on people as the polity’s root 293 – 5 Mesopotamia 77 – 8, 92, 221, 234 microblades 18, 19, 19 – 20 microliths 18, 19, 20 Middle Chinese language 200 – 1 Mid-Former Han empire 168 – 74 Millet, Lord 262 millet agriculture, development of 39 – 42, 40, 304 – 5; stable isotope method and 42 – 9 Mozi, Master 251 – 2 monarchic rule 284 – 7 money see currency moral theology of Xunzi 270 – 3 Morgan, Mo 476 morphology, Old Chinese 210 – 14 Mother spirits 238 – 40 Mozi 251 – 2, 267, 280, 285, 288, 412 – 13 mutilated skeletons 30, 64 – 5

Old Northwest Chinese 208 omens, astral 498 – 507 oracle bones 235 – 6, 370 – 1 “Origin of the Family, Private Property, and the State, The”  48 Panlongcheng site 69 – 70 Pechenkina, E. A. 45 Peiligang Culture 218 Peking Man 2 Peloponnesian War,The 319 penal system and criminal procedure 393 – 5 Perspicuity of Ghosts and Spirits,The see Guishen zhi ming Peterson, Christian 25 phonology, Old Chinese 199 – 201 Ping, Emperor 177 Ping Wang, King 113 Pleistocene 18 – 21,  24 poetry 405 – 8 politics 280; commoners and 293 – 5; development from the mid- to late Western Zhou 92 – 4, 94; intellectual activity and 280 – 2; intellectuals and the state in 290 – 3; monarchic rule and 284 – 7; social structure and 287 – 90; stability in unity 282 – 4 porotic hyperostosis 51 pottery see art pre-Neolithic period 16 – 21, 17, 19 – 20 Prognostication Canon of the Kaiyuan Reign Period see Kaiyuan zhan jing Puett, Michael 238 Pulleybank, E. G. 204 Pulling Book see Yinshu Pythagorean theorem 488 Qi, kingdom of 347 – 9, 348 – 9 Qian Baocong 487, 490 Qieyun System (QYS) 200 – 1; initial consonants 206 – 8; morphology 210 – 14; Old Chinese reconstructions 202 – 3, 202 – 5 Qimin yaoshu 302, 306 Qinlü shibazhong 223 Qin dynasty 146 – 7, 161 – 2, 191 – 2; art 450 – 7, 451 – 2, 454 – 5; banliang 351, 351 – 5; conquest at start of 147 – 51; content of writing in 224; criminal procedure during 395 – 401, 396; development of Chinese characters in 220; empire of 151 – 9; legal system (see legal system, Qin-Han); mathematics in 475 – 81; medicine in 466 – 8; military troops 328; politics in 295; rise of Early Han and 162 – 5, 168; silviculture in 310; warfare 326 Qin state 114, 116, 116, 129 – 30, 140, 142 Qin Statutes in Eighteen Domains 473 Qin Xiang Gong, Lord 117 Qi Xuan Wang, Lord 128

Nao, Lord 239 navy 326 Neolithic period see Chinese Neolithic, the Nine Chapters on Mathematical Procedures,The see Jiuzhang suanshu Norman, Jerry 201 Numinous Efficacy Manifested see Ling xian “Obscure Night” 510 Ockham’s Razor 207 Odes see Shi Oikonomikos 302 Old Chinese (OC) language 199; evolution and dialects 214; grammar 208 – 10; initial consonants 206 – 8; many reconstructions of 202 – 3, 202 – 5; morphology 210 – 14; phonology 199 – 201; tones in 205 – 6

522

Index querns 18 – 19 Qufu city 128 Qu Yuan 419 – 20

Shi (Odes) 262, 302, 314, 322, 380; contents of 406 – 8; Mao Commentary 380 Shiji 7, 127, 146, 139, 154 – 5, 235, 321, 418 – 22, 505, 507, 510 Shijiahe site 30 Shijia site 43 Shi Shen 504 Shu (Documents) 224, 241, 283, 321, 380, 413, 504 Shu1 (Mathematics) 474, 477 – 9; topics covered in 481 – 2 Shuihudi legal texts 386 – 7; tomb no. 11 387 – 8 Shuowenjiezi 398 Sichuan Basin: fortified sites in 30; Sanxingdui pits in 70 – 1 Siku quanshu 217 silviculture 310 Sima Guang 146 – 7, 153 Sima Qian 7 – 8, 64, 66, 76 – 7; on agriculture and food 307, 310; on astronomy 493, 501, 504, 505; compilation of unified history 418; on the Qin dynasty 146, 149, 154 – 5, 157; Records of the Historian 7, 235, 321, 418 – 22; “Treatise on the Celestial Offices” 510 – 12 Sima Tan  418 Si Mu Wu ding 225 Sivin, Nathan 470 skeletal remains: health indicators in Yangshao 49 – 51, 50, 50 – 1; human sacrifice and 75 – 6; mutilated 30, 64 – 5; spirit world and 245 – 6 social order: and development of social complexity 25 – 31, 27 – 8; Eastern Zhou 132 – 9, 134 – 5, 136 – 8; farm and peasant life in 310 – 13; hierarchies 28 – 9, 116 – 18; monetarization of society and 344 – 55, 345 – 6, 348 – 51, 354; politics and 287 – 90; territorial lords and 116 – 18; Western Zhou 100 – 2, 102, 116, 118 Song dynasty 146 Song state 129 – 30 Songs of Chu see Chuci Songs of the South see Chuci spades as currency 341, 341 – 6, 343 – 5, 349 – 50,  350 spirit world 229, 255 – 6; definition and method 229 – 31; discoursing sources 231, 240, 248 – 55; practicising sources 231, 235 – 48; presencing sources 231, 232 – 5; see also religion Spring and Autumn Annals see Chunqiu Spring and Autumn period 109; agronomic developments in 308; astronomy 502 – 4, 503, 506; classification of Zhou and non-Zhou elites in 100; development of Chinese characters in 219; farm and peasant life in 311; as golden age of China’s hereditary aristocracy 287 – 8; number of states in 118; official records 119; spades and knives as currency in 341 – 3; spirit world 241; taxation in 104; urbanization in 126;

Records of the Historian see Shiji Records of Rites 210, 212 – 13, 222, 229, 273 – 6, 338, 381 Regulation of the Marketplace see Xunzi religion 261, 276 – 7; cosmic nature of ritual and early imperial 273 – 6; decline of the Zhou and rise of innovation in 264 – 6; food, farming, and 313 – 15; and morality in the Warring States period 266 – 70; in the Shang and Western Zhou 262 – 4; Xunzi’s moral theology and 270 – 3; see also cosmology; spirit world Remnants of Ch’in Law 386 Remnants of Han Law 386 Rerum Rusticarum Libri Tres 302 rice 304 – 5 rickett 313 Rites of Zhou see Zhouli ritual 4; cosmic nature of 273 – 6 Rong people 124 Sagart, Laurent 206, 207, 211 Sanft, Charles 151 Sanxingdui pits 70 – 1 Schaberg, David 265, 269, 410 science see astronomy; mathematics sea-cowries 336 – 9,  337 Sena, David 373 Seven States, warring by the 139 – 40 Shaanxi sites 43 – 4, 87, 147; pottery 430, 430 Shandong province 129 Shangjun shu 294, 332 Shang dynasty 7, 84 – 5; agronomic developments in 308; astronomy in 496 – 7, 506; currency 336, 338, 339; development of Chinese characters in 218 – 19; founding and expansion of the Zhou dynasty and 90 – 2; king lists and royal legitimation 76 – 9; language 199, 209, 214; medicine in 460; military troops 321 – 2, 327; oracle bones 235 – 7; predynastic Zhou and 73 – 4; religion in 262 – 4; royal lineage and ancestor cult 238 – 9; spirit world of 235 – 48; war with the Zhou 322 – 3; women in 370 – 2;Yanshi Erlitou excavations and 68 – 9; Zhengzhou excavations and 66 – 8 Shang Yang 147 – 8 Shan Rong people 124 Shao Tuo  243 Shaughnessy, Edward L. 91, 406 shellfish 20 – 1 Shen, Chen 126, 128, 133, 139 Shennong deity of agriculture 314 – 15 shi aristocrats 117, 290 – 1 Shi Dan 176

523

Index warfare in 323, 325, 330; women in 376 – 7, 380; writers 89 stable isotope perspective on Yanghsao diets 42 – 9, 45, 46, 47 Statutes and Edicts of Year Two see Ernian lüling Statutes and Ordinances of the Second Year see Ernian lüling Strategies of the Warring States see Zhanguo ce Suanjing shi shu 475 Suanshu shu 473, 475 – 9, 476; practices and epistemological values 482 – 3; topics covered in 481 – 2 Sui state 249 – 50 Suixing jing 504 Sumerian King List (SKL) 77 – 8 Sun, Xidan 274, 275 – 6 Sun,Yirang 266 – 7 Su Qin 128 swords/sabres 326

village life, early 21 – 5, 23 violence, intra-societal 30 Waley, Arthur 263 Wang, Qingzheng 271 – 3 Wangchenggang site 34 Wangcheng state 119 Wang Chong 253 – 5, 421 Wang Fenshi 246 – 7 Wang Guowei 64 Wang Hai, Lord 239 Wang Li 205, 206 Wang Mang: art and 455; as leader 174 – 7, 181 – 2, 185, 189, 191, 313; mathematics and 474; monetary reforms 358 – 62, 359 – 62 Wang Mingke 188 Wang Xi 468 Wang Yirong  63 Wang Zhengjun 177 Wan Zhang 413 warfare 319 – 20; cavalries used in 325; chariots used in 323 – 4, 327; as the core of history of early China 320 – 6; infantries in 324 – 5; motivations for 329 – 32; preparing for 326 – 9; as self-defense 331; weapons 326 Warring State Bamboo Books Collection of Chu State from Shanghai Museum 112, 114 Warring States period 113, 122, 139, 191; agronomic developments in 307 – 8; astronomy 502 – 7; battles 139 – 42, 141; beginning of 96; content of writing in 223 – 34; currency 344 – 9, 345 – 6, 348 – 51, 354; development of Chinese characters in 219; as Eastern Zhou sub-period 109 – 10, 112; end of 151; excavated tombs 89; factories 130, 132; farm and peasant life in 310 – 11; farming in 306; intellectual activity in 281 – 2, 295 – 6; medicine in 461, 468 – 70; monarchic rule and 284 – 7; morality and religious thought in 266 – 70; people as the polity’s root in 293 – 5; political thought on unity in 282 – 3; regional Zhou kings 97; Shennong deity of agriculture and 314; social mobility in 289 – 90; spirit world 241 – 3; Tsinghua University collection of manuscripts from 226; unification of intellectual world during 193; urbanization in 128; volume of countries created in 104; warfare 323, 325 – 6; women in 380 – 1; writers 73 – 4 Way of Tang and Yu, the see Tang Yu zhi dao Wei Dan 456 Wei state 114, 139 – 40, 151 Weiyu deng zhuang sizhong 392 Wen, Emperor 166, 168, 507 Western Zhou dynasty 84, 161; art 434 – 50, 451 – 2, 454, 454 – 5, 456; breakdown of royal order and transition to territorial states 103 – 5, 108; bronze inscriptions 222 – 3; currency

Tables for Major Events of the Spring and Autumn see “Chunqiu Dashi Biao” Tai Jen 379 Taiyi Engendered Water see Taiyi sheng shui Taiyi sheng shui 223 Tang, Lan 337 Tang Yu zhi dao 223 Taosi site 28, 32, 34, 218; astronomy and 495 temples 460 – 1 Ten Mathematical Canons see Suanjing shi shu territorial lords: age of transformation to 108 – 9; commercialization and production and 128 – 32, 131; dates of 113 – 16, 115; life in peace and war 132 – 9, 134 – 5, 136 – 8; people and social hierarchy of 116 – 18, 116; sources on 109 – 13; and their states 118 – 25, 120 – 1, 123 – 4; toward unification of 139 – 42, 141; urbanization and population in cities of 125 – 8, 127 Thirty Years’ War  103 Thucydides 319, 409 Tian Shan 126, 139 Tibeto-Burman language 200, 204, 206 – 9, 212, 214 tombs see graves and tombs tones, Old Chinese 205 – 6 “Treatise on the Celestial Offices” 510 – 12 Treatise on Jupiter see Suixing jing Ulysses 463 unity, literature of 418 – 22 urbanization and population in Eastern Zhou states 125 – 8,  127 urns see vessels and urns Varro 302 vessels and urns: Bronze Age 433 – 50, 454, 456; late Neolithic 425 – 33, 426 – 7, 429 – 32

524

Index 336 – 41, 340; development of Chinese characters in 218 – 19; “Five Cities” 98 – 9; founding and expansion 90 – 2, 91; government and bureaucracy 98 – 100; kings 84 – 8, 85; maps 86, 87; organization of 95 – 8, 97; political development from the mid- to late 92 – 5, 94; regional rulers 96 – 8; religion in 262 – 4; rise of religious innovation with decline of 264 – 6; social order 100 – 2, 102, 116, 118; sources 88 – 90; time and space of 84 – 8, 85, 86 – 7; women in 372 – 6 wheat agriculture, development of 41 – 2 women 367, 381 – 2; in the Eastern Zhou dynasty 376 – 81; education of 381; Neolithic period 367 – 70; Shang paleographic texts on 370 – 2; as soldiers 328, 375 – 6; in the Western Zhou dynasty 372 – 6 writing, early Chinese see early Chinese writing Writings on Mathematical Procedures see Suanshu shu Wu, Emperor of Han 168 – 74, 308, 326, 355; astronomy and 505; infantries of 326, 328; literature under 421 – 2; motivations for warfare and 331 – 2; wuzhu coins 356 – 7, 356 – 7 Wu, King 90 – 2, 122, 373 Wu Ding, King 65 – 6, 71, 321, 371 – 2; astronomy and 496 – 7 Wu Guang 156, 157, 162 Wu Zhenfeng 337 Wuxing 223 wuzhu 356 – 7, 356 – 7; exchange value of 357 – 8

Xunzi 129, 252 – 3, 270 – 5, 415 – 16; on coinage 358; on the intellectual and the state 292 – 3; on monarchic rule 285 – 7; on ritual as universal social regulator 290 Xu Shaohua 125 Xu Shen 226 Yahuazhai site 43 – 4 Yang, Bojun 264, 265, 269, 330, 331 Yang, Kuan 109 Yang,Yong 463, 464 Yangshao culture: animal bones and 41; astronomy and 494; dietary variation in 44 – 8, 53, 54; grave sites 43 – 4; Henan site 44; indicators of human health in skeletal collections of 49 – 51, 50, 50 – 1; male-female differences in 48 – 9; millet agriculture and 40, 40, 40 – 1; Shaanxi sites 43 – 4; shift toward wheat and barley agriculture in 52; stable isotope perspective on diets in 42 – 9, 45, 46, 47 Yangtze region 76, 87; ceramic artifacts and cultural interactions 32 – 3; graves 31, 31 – 2; rice in 304 – 5; unplanned archaeology in 72 – 3 Yang Xiong 420 – 1 Yan Le 158 Yanshi Erlitou site 68 – 9 Yan Shigu 2 – 3 Yan state 126 – 7, 127, 130, 132, 140 Yantie lun 174, 175 – 6, 358 Yanxiadu city 128, 130, 132 Yan Yan  275 Yan Ying  250 Yates, Robin 386 Yellow Turban Rebellion 190 – 1 yeren 327 Yi (Changes) 483 Yi, Zeng Lord 111 Ying Zheng 142 Yin Ji 375 – 6 Yinqueshan city 129 Yinshu 463 Yi state 124 Yongcheng city 129 Yong Rui 411 – 12 You Wang, King 108, 124 Yu, Lord 125 Yuan Guohua 270 Yuanmou Man 2 Yuan Wang, King 113 Yuelu Academy 392 – 3 Yuhuazhai site 46

Xenophon 302, 319 Xi (Shuihudi tomb no. 11) 387 Xia Fuji 265 Xiang Liang 158 Xiang Yu 158, 163 – 4, 418 Xianyun people 323 Xiao, King 93, 147 – 8 Xiao Gong, Lord 114 Xiao He 165 Xiaotangshan shrine 455, 455 Xiaotun site, oracle bones at 64 Xiao Yi, King 496 Xia Shang Zhou Chronology Project 61 Xin dynasty 177 Xingan tomb 71 – 2 Xing zi ming chu 223 Xi Nian 114 Xiongnu confederation 170 – 1, 185 – 8,  192 Xipo site 44; animal bones at 41 Xi Rong 119 Xishan site 44 Xizhufeng site 26, 28 Xuan, Emperor 175 Xuan, King 103, 376 Xun, Master 252 – 3 Xun Kuang 129, 358

Zeng Houyi, tomb of 111 – 12 Zhang Cang 165 Zhang Changping 112 Zhang Heng 422, 509 – 10 Zhangjiashan legal texts 386 – 7

525

Index Zhangjiashan tomb no. 247 389 – 92 Zhanguo ce 126, 411, 418 Zhang Qian 171 Zhang Zhi 456 Zhao, King 93, 222 – 3; women under rule of 375 – 6 Zhao Chongguo 329 Zhao Dun 5 – 7 Zhao Gao 155 – 8, 161 Zhao Guo 309 Zhao She 126, 139 Zhao Shuang 488 Zhao state 140 Zhao Yang  121 Zhao Yi 456 – 7 Zheng, King 149 – 50 Zheng state 130, 139 Zheng Wu Gong, Lord 119 Zheng Xuan 193, 489 Zhengzhou site as local antecedent for Anyang civilization 66 – 8 Zhen Luan 475 Zhongguo ren 3 – 4 Zhongli state 125 Zhou dynasty: agronomic developments in 308; currency 336 – 7; Early Bronze Age before the 61 – 79; Eastern (see Eastern Zhou dynasty); king list and royal legitimation 78 – 9; language development 209; military troops 327; and predynastic Zhou as statesmen or barbarians 73 – 4; rise of religious innovation with decline of 264 – 6; spirit world 240 – 1, 241 – 2; war with the Shang 322 – 3; Western (see Western Zhou dynasty)

Zhou Gong, Lord 122 Zhou Ren 469 Zhou Wen  157 Zhoubi 474 – 5, 477, 490; practices and epistemological values 482 – 3; right-angled triangle and the mathematics of the calendar in 487 – 9; topics covered in 482 Zhoubi suanjing 508 Zhouli 129, 241, 273 – 5, 302, 468, 489 – 90 Zhuangxiang, King 148 Zhuangzi 412 – 13, 414 – 15 Zhu Kezhen 303 – 4 Zhushu Jinian 126, 338 Zhu Shunlong 109, 139 Zhu Yiwen  475 Zichan 265 – 6,  267 Zi Chu 142 Zigong 264, 268 – 9 Ziying, King 158 Zouyan shu: significance of 387, 391 – 3; on appeals and amnesty 398; on interrogation procedure 397; trial of Tuzi shan 400 – 1 Zuo Bo 456 Zuo Commentary see Zuo Zhuan Zuo Qiuming 409 Zuo Zhuan: on chariot warfare 323 – 4; on citybuilding 125; on class divisions 117; on filial piety 374; on infantry and cavalry 325; on Lord Jing Gong 112; on the Luhun Rong people 124; on marriage and women’s behavior 380 – 1; on nobles defining rights 288; on people as polity’s root 293; on preparation and thinking about warfare 330; on ritual 290; significance of 110; on state names 118; on widows 377 – 8

526