Routledge Handbook of Early Chinese History 1138775916, 9781138775916

The study of early China has been radically transformed over the past fifty years by archaeological discoveries, includi

742 66 11MB

English Pages 546 [547] Year 2014

Report DMCA / Copyright


Polecaj historie

Routledge Handbook of Early Chinese History
 1138775916, 9781138775916

Table of contents :
List of figures
List of tables
List of maps
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

Citation preview


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).


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



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



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



  3 The Bronze Age before the Zhou dynasty Robert Bagley


  4 The Western Zhou state Li Feng




  5 The age of territorial lords Chen Shen


  6 The Qin dynasty Charles Sanft


  7 The Former Han empire Vincent S. Leung


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



Topical studies197   9 The Old Chinese language Axel Schuessler


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


11 The spirit world Jue Guo


12 Religious thought Ori Tavor


13 Political thought Yuri Pines


14 Food and agriculture Roel Sterckx


15 Warfare Wicky W.K.Tse


16 Currency François Thierry


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




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


19 Literature Stephen Durrant


20 Art Wang Haicheng


21 ‘Medicine’ in early China Miranda Brown


22 Mathematics Karine Chemla


23 Astronomy David Pankenier






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


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


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 


451 454 456


500 501 508



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 


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



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


86 87



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


?–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



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.




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, 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


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


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


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


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.




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


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


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


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).



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.


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.


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.





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



16 17 18


1 13

20 21




14 28 29 31

22 23



30 32

33 27


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



II H23








H14 H18

F34 H17 H27 H29





H24 H16 F24

F40 H33




F32 H13


G1 F50

F51 F46 F45


F30 F31




F52 F49

H36 M2











M3 M1 M7 M4 M10 M9 M5 M6





F23 T0508 T0408











tY et

H7 D3





D4 D1








Exc av ate d


F3 H9






F16 F17

H2 H1


F14 H12









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






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



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).




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


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.


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.


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.


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.




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


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


δ 13C ‰ mean

δ 15N ‰ SD


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




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


0.71 1.1


9.24 8.3


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


δ 13C ‰ mean

δ 15N ‰ SD



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



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






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






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)


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









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)




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







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.


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


δ 13C ‰ mean

δ 15N ‰ SD


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



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






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






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


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.


Henan   Henan





Henan   Henan



Xipo   Xipo





Erlitou   Zhangdeng


  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


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


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


−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


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





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


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


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



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


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.



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.


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.


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.


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.


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.



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


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


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


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


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).


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.


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.




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



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).


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


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.


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.


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


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


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.


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


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


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.


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.


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).


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.


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.



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


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


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


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


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.


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


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


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


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.


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.


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.


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.




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.




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).


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.


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.



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.


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.


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.



Topical studies



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


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


Wáng Lì

Lǐ F-k.

B. ’92

B&S ’14/11



*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?



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


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



*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














*kâr *klwâr

*kai *kuai

*karh *kwarx

*kajs ? ? -ajʔ

*kˤa[r]-s *[k]ˤo[r]ʔ

*kaals *kloolʔ

*kâih *kʷâiʔ/*kôiʔ


*gʷrajʔ ? *m-kˤo[r] *ɡroolʔ ʔ

*gʻlwar *goai

*kaj *maj *kʷajs


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.


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.




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


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



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.


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.




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.


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.


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.


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.


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.




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.


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,


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。


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