Buried Ideas: Legends of Abdication and Ideal Government in Early Chinese Bamboo-Slip Manuscripts 9781438457772

Table of contents :
Dedication
Contents
List of Illustrations
Acknowledgments
1 Introduction
2 History and Historical Legend
3 The Chu-script Bamboo-Slip Manuscripts
4 Advocating Abdication: Tang Yú zhi dao 唐虞之道, “The Way of Tang Yao 唐堯 and Yú Shun 虞舜”
Tang Yú zhi dao 唐虞之道: Translation and Chinese Edition
5 The Zigao 子羔 and the Nature of Early Confucianism
Zigao 子羔: Translation and Chinese Edition
6 Rongchengshi 容成氏: Abdication and Utopian Vision
Rongchengshi 容成氏: Translation and Chinese Edition
7 The Bao xun 保訓: Obtaining the Center to Become King
Bao xun 保訓: Translation and Chinese Edition
8 Afterthoughts
Bibliography
Index

Citation preview

B u r i e d I d e as

SUNY series in Chinese Philosophy and Culture Roger T. Ames, editor

Buried ideas Legends of Abdication and Ideal Government in Early Chinese Bamboo‑Slip Manuscripts

SaraH aLLan

SUNY P R E S S

Published by S t a t e U n i v e r s i t y o f N e w Y o r k P r e ss Albany © 2015 State University of New York All rights reserved Printed in the United States of America No part of this book may be used or reproduced in any manner whatsoever without written permission. No part of this book may be stored in a retrieval system or transmitted in any form or by any means including electronic, electrostatic, magnetic tape, mechanical, photocopying, recording, or otherwise without the prior permission in writing of the publisher. For information, contact State University of New York Press www.sunypress.edu Production and book design, Laurie D. Searl Marketing, Anne M. Valentine Library of Congress Cataloging-in-Publication Data Allan, Sarah. Buried ideas : legends of abdication and ideal government in early Chinese bamboo-slip manuscripts / Sarah Allan. pages cm. — (SUNY series in Chinese philosophy and culture) Includes bibliographical references and index. ISBN 978-1-4384-5777-2 (hardcover : alkaline paper) — ISBN 9781-4384-5779-6 (e-book) 1. China—History—Warring States, 403-221 B.C.—Sources. 2. China—Politics and government—To 221 B.C.—Sources. 3. China—Kings and rulers—Abdication—History—Sources. 4. Ideals (Philosophy)—Political aspects—China—History—To 1500—Sources. 5. Merit (Ethics)—Political aspects--China—History—To 1500—Sources. 6. Political science—China—History—To 1500—Sources. 7. China— Intellectual life—To 221 B.C.—Sources. 8. Manuscripts, Chinese. 9. Chinese language—To 600—Texts. I. Title. DS747.2.A55 2015 931'.03—dc23 2014041449 10

9

8

7

6

5

4

3

2

1

to Nicol Allan

Contents

List of Illustrations ix Acknowledgments xiii Chapter One Introduction 1 Chapter Two History and Historical Legend

9

Chapter Three The Chu-script Bamboo-Slip Manuscripts

25

Chapter Four Advocating Abdication: Tang Yú zhi dao 唐虞之道, “The Way of Tang Yao 唐堯 and Yú Shun 虞舜” 79 Tang Yú zhi dao 唐虞之道: Translation and Chinese Edition

118

Chapter Five The Zigao 子羔 and the Nature of Early Confucianism

135

Zigao 子羔: Translation and Chinese Edition

168

vii

viii Contents

Chapter Six Rongchengshi 容成氏: Abdication and Utopian Vision

181

Rongchengshi 容成氏: Translation and Chinese Edition

223

Chapter Seven The Bao xun 保訓: Obtaining the Center to Become King

263

Bao xun 保訓: Translation and Chinese Edition

304

Chapter Eight Afterthoughts 315 Bibliography 329 Index 357

Illustrations

Figure 1.1. xiv Chu cemetery region, near Ji’nan, Hubei Province: Scenery with unexcavated tumulus. 28 Figure 3.1. Writing brush and holding tube, Baoshan Tomb Two. After Baoshan Chu mu 包山楚墓, ed. Hubeisheng Jingsha Tielu Kaogudui 湖北省荊沙鐵路考古隊 (Beijing: Wenwu, 1991), plate 15.1. Figure 3.2. 29 Bao xun 保訓, Slips 1–11, Tsinghua University collection. From Qinghua Daxue cang Zhanguo zhujian 清華大學藏戰國竹簡, vol. 1, 8. Courtesy of the Chutu Wenxian Yanjiu yu Baohu Zhongxin 出土文獻研究與保護中心, Tsinghua University. Figure 3.3. Oracle bone characters with bamboo-slip component (ce 册).

30

Figure 3.4. 32 Wood tablet (du 牘), Taiyuan you sizhe 泰原有死者, Qin Dynasty, Peking University collection. From Wenwu 2012.6: 71.

ix

x I l l u s t r at i o n s

Figure 3.5. 34 Silk manuscript, Laozi 老子 (jia 甲 version), Mawangdui Tomb One, Changsha, Hunan Province. From Mawangdui Han mu Boshu 馬王 堆漢墓帛書 (Beijing: Wenwu, 1980), vol. 1, frontispiece. Figure 3.6. Plan of Guodian Tomb One. After Wenwu 1997.1: 36.

38

Figure 3.7. Inscribed lacquer wine-cup from Guodian Tomb One.

43

43 Figure 3.8. Inscription on bottom of lacquer wine-cup from Guodian Tomb One. Figure 3.9. 52 Map of Chu cemetery region, near Ji’nan, Hubei Province, showing Guodian and Guojiagang. After Wenwu 1997.1: 35. Figure 3.10. 64 Bamboo slips upon arrival at Tsinghua University, wrapped in plastic. Courtesy of the Chutu Wenxian Yanjiu yu Baohu Zhongxin 出土文獻研究與保護中心, Tsinghua University. 65 Figure 3.11. Bamboo slips when first opened after arrival at Tsinghua University. Courtesy of the Chutu Wenxian Yanjiu yu Baohu Zhongxin 出土文 獻研究與保護中心, Tsinghua University. Figure 3.12. 65 Bamboo slips (Shifa 筮法) with residue of fabric holder. Courtesy of the Chutu Wenxian Yanjiu yu Baohu Zhongxin 出土文獻研究與 保護中心, Tsinghua University. Figure 3.13. 67 Bamboo slips preserved in trays of pure water. Courtesy of the Chutu Wenxian Yanjiu yu Baohu Zhongxin 出土文獻研究與保護中 心, Tsinghua University.



I l l u s t r at i o n s x i

Figure 4.1. 84 Tang Yú zhi dao, detail of mark left on Slips 26, 27, 28, 4, 5, 6, stepped to show the slant at which the slips were skewed after placement in the tomb. After Jingmenshi Bowuguan 荊門市博物 館, ed. Guodian Chu mu zhujian 郭店楚墓竹簡, (Beijing: Wenwu, 1998), 39–41, based on Chūgoku no kiso shiryō kenkyū han 中國 の基礎史料研究班, “Du Guodian Chu mu zhujian zhaji 讀郭店 楚墓竹簡札記” (4), Yuegu 曰古 10 (2007.9): 1–18. 86 Figure 4.2. Tang Yú zhi dao, rearranged, Slips 1–3, 22–28, 4–7. After Jingmenshi Bowuguan, ed. Guodian Chu mu zhujian (Beijing: Wenwu, 1998), 39–41. Figure 4.3. 87 Tang Yú zhi dao, rearranged, Slips 8–10, 12–13, 18–21, 11, 14–17, 29. After Jingmenshi Bowuguan, ed. Guodian Chu mu zhujian (Beijing: Wenwu, 1998), 39–41.

Acknowledgments

This book has taken many years to write during which I have received substantial institutional support, for which I would like to express my gratitude. I owe particular thanks to Dartmouth College, which has long supported my research on bamboo-slip manuscripts, beginning with the first International Conference on the Guodian Laozi held here at Dartmouth in 1998. This includes, most recently, the Dartmouth-Tsinghua International Conference on the Tsinghua University bamboo-slip manuscripts, held at Dartmouth, August 30–September 1, 2013, for which we received assistance from the Leslie Center for the Humanities, Offices of the President, Provost, Dean of the Faculty of Arts and Sciences, and the Department of Asian and Middle Eastern Languages and Literatures, Anne and Steven Tseng, Jennifer and Peter M. Kurz, and Xiao Nan Liao. I am grateful for the receipt of a fellowship from the National Endowment for the Humanities in 2006–07 and am happy to include the required disclaimer: any views, findings, conclusions or recommendations expressed in this book do not necessarily represent those of the National Endowment for the Humanities. I am also grateful for a Chiang Ching-kuo Senior Research Fellowship in 2009–10. Though I was overly optimistic about the time necessary to complete this work in making those applications, without the uninterrupted time these fellowships gave me for research and writing, this book would not have been possible. On the personal level, my debts are too numerous to acknowledge herein. Many are longstanding. The ideas in this book are a development of my earliest research, and I am grateful to my teachers, Wolfram Eberhard and Peter Boodberg, who sparked my interest in the relationship between historical accounts and philosophical thought in early China. Since the xiii

xiv

Buried Ideas

publication of the Guodian manuscripts in 1998, the field has developed exponentially and over the years, I have presented many of the ideas found in these books at various conferences and workshops. This interchange has contributed greatly to my understanding. The complexity of the materials and the speed with which research on Warring States manuscripts has developed are reasons that this book has taken so long to write, but I have been buoyed by the enthusiasm and generosity of very many colleagues. Among my more specific debts, I would like to thank my colleague at Dartmouth, Susan Blader, for proofreading and many helpful comments and suggestions that have much improved the book. I would also like to thank Diana Matias for copy editing an earlier version of this book, Cai Yuqian 蔡雨錢, who has brought various mistakes and infelicities of language to my attention in the course of translating the book into Chinese, and Andrew West for making the font of archaic characters used herein. The mistakes, of course, remain my own. Finally, but most importantly, my husband, Nicol Allan, has always inspired my work and this book is, as always, dedicated to him.

Figure 1.1. Chu cemetery region, near Ji’nan, Hubei Province: Scenery with unexcavated tumulus.

1

Introduction

In the autumn of 1993, tomb robbers operating south of Mount Ji 紀, near Jingmen 荊門 City, in the southern Chinese province of Hubei, had just broken into the head compartment of the outer coffin of a tomb in the village of Guodian 郭店, when they were disturbed and abandoned their effort. The area where the robbers were digging is well known to archaeologists as a cemetery site of an ancient capital of the southern state of Chu 楚 during the middle Warring States period (375–278 BCE) (see Figure 1.1). The attempted robbery was reported to the authorities and archaeologists from Jingmen City Museum quickly organized a salvage excavation.1 The local environmental conditions and the burial customs of the ancient state of Chu are unusually conducive to preservation. This region of Hubei is moist, with heavy, clay-laden soil. The coffins, which included compartmentalized outer coffins made of carefully selected timbers assembled with tenon-and-mortise joints and then lacquered, were placed in deep pits and the tombs were sealed with layers of fine clay.2 Thus, 1. See Sarah Allan and Crispin Williams, eds., The Guodian Laozi: Proceedings of the International Conference, Dartmouth College, May, 1998 (Berkeley: Society for the Study of Early China and Institute of East Asian Studies, University of California, Berkeley, 2000), 23–32, 117–19; Jingmenshi Bowuguan 荊門市博物館, “Jingmen Guodian yihao Chu mu 荊門郭店一號楚墓,” Wenwu 1997.7: 35–53. 2. Alain Thote, “Continuities and Discontinuities: Chu Burials during the Eastern Zhou Period,”in Exploring China’s Past: New Discoveries and Studies in Archaeology and Art, ed. Roderick Whitfield and Wang Tao (London: Saffron Books, 1999), 189–204; Alain Thote, “Burial Practices as Seen in Rulers’ Tombs of the Eastern Zhou Period: Patterns and Regional Traditions,” in Chinese Religion and Society, ed. John Lagerwey and Marc Kalinowski (Hong Kong: Chinese University Press, 2004), vol. 1, 77–82. 1

2

Buried Ideas

archaeologists frequently uncover perishable materials, such as wood, lacquer, bamboo, and silk; and they had excavated thousands of Warring States tombs in the region with many remarkable finds when they opened the Guodian tomb. Nevertheless, on this occasion, they made one of the most spectacular discoveries in the history of Chinese archaeology—philosophical texts brush-written on some eight hundred slips of bamboo. This find has been followed by two other major discoveries of similar materials, although these, unfortunately, were not archaeologically excavated. In 1994, freshly unearthed bamboo-slip manuscripts looted from a Warring States period tomb arrived in Hong Kong, where they were sold in the local antiquities market. They were acquired for the Shanghai Museum by its late director, Ma Chengyuan 馬承源 (1927–2004). This collection includes about 1,200 bamboo slips with writing in the same type of Chu script as that found on the Guodian manuscripts. There are philosophical texts among these manuscripts, but their contents are more varied than those from Guodian Tomb One. The provenance is uncertain, but Ma surmised that they came from a tomb of similar date, probably nearby.3 Then, in the summer of 2008, approximately 2,500 more bamboo slips with writing in Warring States period script were acquired by Tsinghua University.4 These are mainly historical texts, including some that are similar in style to, or versions of, documents found in the Shang shu 尚書 (“Ancient Documents”) and a chronicle of events beginning in the Western Zhou and continuing to the middle Warring States period. The publication of these bamboo-slip manuscripts is on-going. In 1998, transcriptions in modern characters of the Guodian manuscripts, prepared under the auspices of the Jingmen City Museum, were published under the title Guodian Chu mu zhujian 郭店楚墓竹簡, together with annotations and high-quality black and white photographs of the original brush-written

3. Ma Chengyuan 馬承源, ed., Shanghai Bowuguan cang Zhanguo Chu zhushu 上海 博物館藏戰國楚竹書, (Shanghai: Shanghai Guji, 2001), vol. 1, 1–2. See also Ma Chengyuan,“Zhushu Kongzi shi lun jianji shi de youguan ziliao (tiyao) 竹書《孔子 詩論》兼及詩的有關資料(提要),” in Xin chu jianbo yanjiu: Xin chu jianbo guoji xue‑ shu yantaohui wenji 新出簡帛研究: 新出簡帛國際學術研討會文集 (English title: Proceedings of the International Conference on Recently Excavated Manuscripts, Beijing University, August 19–22) ed. Ai Lan 艾蘭 (Sarah Allan) and Xing Wen 邢文 (Beijing: Wenwu, 2004), 1. 4. Qinghua Daxue Chutu Wenxian Yanjiu yu Baohu Zhongxin 清華大學出土文獻 研究與保護中心, “Qinghua Daxue cang Zhanguo zhujian Bao xun shiwen 清華大 學藏戰國竹簡《保訓》釋文,” Wenwu 2009.6: 73–75.

Introduction 3

bamboo slips.5 The Shanghai Museum slips began to be published in 2001. Thus far (March 2015), nine volumes have been published under the title Shanghai Bowuguan cang Zhanguo Chu zhushu 上海博物館藏戰國楚竹書.6 The Shanghai Museum volumes include color enlargements of the brushwritten characters, as well as black and white photographs, annotations, and transcriptions. A team of scholars, working under the direction of Li Xueqin 李學勤, is preparing the Tsinghua University slips for publication. The first five volumes of the manuscripts have already been published under the title Qinghua Daxue cang Zhanguo zhujian 清華大學藏戰國竹簡.7 A total of twelve volumes, divided into about sixty-five manuscripts, is projected. These manuscripts are revolutionizing our understanding of the history of early Chinese thought. Historically, discoveries of ancient manuscripts were rare events to which great significance was attached. The most famous examples are the reputed discovery of manuscripts hidden in the walls of Confucius’s house, which were edited by Kong Anguo 孔安國 (died ca. 100 BCE), and those recovered from a late Warring States tomb at Jizhong 汲冢 (in present-day Henan Province) in the third century CE, the most famous of which is the Zhu shu jinian 竹書紀年 (“Bamboo Annals”). With the advent of modern archaeology, discoveries of ancient manuscripts are no longer so rare. Since 1973, when the first major archaeological excavation of ancient manuscripts was made from a Han Dynasty tomb at Changsha Mawangdui 長沙馬王堆 in Hunan Province, Chinese archaeologists have excavated dozens of ancient tombs that contained ancient manuscripts written on silk, bamboo or wooden slips, and wooden tablets of great importance. Nevertheless, these three groups of Warring States period bamboo-slip manuscripts written in the Chu script have a special significance because they are different from but closely related to the core texts of the Chinese philosophical and historical tradition. Some are early versions of transmitted classical

5. Jingmenshi Bowuguan 荊門市博物館, ed., Guodian Chu mu zhujian 郭店楚墓竹 簡 (Beijing: Wenwu, 1998). The transcriptions and notes were prepared by Peng Hao 彭浩, Liu Zuxin 劉祖信, and Wang Chuanfu 王傳富. Additional annotations by Qiu Xigui 裘錫圭 are included. 6. The transcriptions and annotations in Shanghai Bowuguan cang Zhanguo Chu zhushu were prepared by a group of scholars from the Shanghai Museum and other institutions, including the main editor Ma Chengyuan 馬承源, Pu Maozuo 濮茅左, Li Chaoyuan 李朝遠, Chen Peifen 陳佩芬, Li Ling 李零, Zhang Guangyu 張光裕, and Cao Jinyan曹錦炎. 7. Li Xueqin, ed., Qinghua Daxue cang Zhanguo zhujian 清華大學藏戰國竹簡 (Shanghai: Zhong Xi, 2010-15).

4

Buried Ideas

Chinese texts, but most of them were previously unknown. Moreover, since they were buried before the Qin conquest, unlike transmitted texts, they were not edited in the Han Dynasty. I argued in The Heir and the Sage, published in 1981, long before the discovery of these texts, that transformation of historical legend is the primary means of expressing political theory in Warring States period texts and that the role of these legends in this period is to mediate the conflicting principles of rule by heredity and rule by virtue.8 In light of this earlier study, the four manuscripts that I have selected for analysis in this book are ones that include discussion of the legend that the predynastic ruler, Yao 堯, abdicated to Shun 舜. They include: Tang Yú zhi dao 唐虞之道 (“The Way of Tang [Yao] and Yú [Shun]”) from Guodian Tomb One; Zigao 子羔 (“Zigao”) and Rongchengshi 容成氏 (“Rongchengshi”) in the Shanghai Museum collection; and the Bao xun 保訓 (“Cherished Instruction”) in the Tsinghua University collection.9 These manuscripts reveal a meritocratic challenge to the idea of hereditary succession that is absent in the transmitted texts, presumably because it was no longer possible to make this type of argument after China became unified under the Qin and Han dynasties. Once this challenge is recognized, however, we can see that it served as the background to many well-known arguments concerning dynastic legitimacy and the changing celestial mandate (tian ming 天命). Although these four manuscripts are all strongly meritocratic, they do not form a coherent philosophical group. Indeed, they have no obvious relationship with one another other than their interest in the legend of Yao and Shun and a shared meritocratic bent. They also take different literary forms. Tang Yú zhi dao takes the form of philosophical discourse. Its primary theme is that the Way of Tang Yao and Yú Shun—that is, abdication to a man of virtue—is the ideal form of political succession. Some scholars have taken the manuscript as Confucian; others as belonging to or influenced by the followers of other groups or philosophers, such as the Horizontal and Vertical Strategists (zongheng jia 縱橫家), Mozi 墨 子, Yang Zhu 楊朱, or some combination thereof. I will not attempt to 8. Sarah Allan, The Heir and the Sage: Dynastic Legend in Early China (San Francisco: Chinese Materials Center, 1981). 9. For Tang Yú zhi dao, see Guodian Chu mu zhujian, 39–41, 157–59. I mark the tone of the lineage name Yú 虞 here and elsewhere in this book to distinguish this name from that of Yu 禹. For Zigao and Rongchengshi, see Shanghai Bowuguan cang Zhan‑ guo Chu zhushu, vol. 2, 31–47,183–89 and 91–146, 247–93. For the Bao xun, see Qin‑ ghua Daxue cang Zhanguo zhujian, vol. 1, 8–9, 55–63.

Introduction 5

place it within any philosophical school or to match it to any known text, but I will argue that the manuscript reflects ideas that were current in the late fifth and early fourth centuries BCE. Moreover, in light of this manuscript, we can see that Mencius’s theory of a changing mandate of heaven, demonstrated by the movement of the people from one ruler to the next, was articulated in response to contemporary arguments, if not necessarily to this specific text, in favor of political succession by abdication to the most meritorious. The Zigao consists of a series of questions that the disciple, Zigao, addresses to Confucius together with Confucius’s replies. Zigao is known from the Analects (Lun yu 論語), where he appears as one of Confucius’s less accomplished disciples. In the bamboo-slip manuscript, Zigao inquires about the divine birth of the ancestors of the three dynasties: Yu 禹 of the Xia, Xie 契 of the Shang , and Hou Ji 后稷 of the Zhou. Confucius, in response, tells the stories of their miraculous births. In this discussion, Confucius uses the conventional term for a ruler, tian zi 天子, “son-ofheaven,” literally—to mean someone of divine birth as opposed to an ordinary human. He concludes by stating that all of these sons-of-heaven, had they been alive in his time, would have served the sage, Shun. This manuscript is not only radically meritocratic, but it suggests that the followers of Confucius included people with ideas that would have been considered heterodox in later times. Rongchengshi is a long narrative with a vision of high antiquity that is entirely absent from transmitted texts. It begins with an idyllic era, before the time of Yao, in which everyone served according to their ability and the rulers all abdicated to the most worthy person, rather than passing the rule to their own sons. Yao abdicated to Shun, but he himself achieved power by attracting the loyalty of the people through lenient government. The succession of Yu’s son, Qi 啓, rather than the sage Yi 益, whom Yu had chosen to be his successor, and thus the foundation of the Xia Dynasty, is described as a matter of force. In this devolutionary scheme, the evil behavior of the last rulers of the Xia and Shang dynasties, Jie 桀 and Zhòu 紂, is described in graphic detail and the dynastic founders achieve power by attracting the loyalty of the people through their cunning use of good government. This manuscript is the only known text that states that there were rulers before Yao who abdicated and its particular vision of highest antiquity as a period in which all served according to their abilities is unique. It is also the earliest historical narrative that encompasses successive historical periods. However, I will argue that it is best read in the context of Warring States period philosophical texts, that is, as political theory, rather than as a historical record.

6

Buried Ideas

The Bao xun takes the form of a shu 書, “document.” It has a short preface, followed by the words of a xun 訓, “instruction,” left by King Wen 文 of Zhou for his son, Fa 發. In this instruction, King Wen cites two historical models: Shun and the Shang ancestor, Wei 微, who had “obtained the center.” Transmitted texts frequently equate the founding of the Zhou with a breach of hereditary right when Yao gave the rule to Shun, but there are no records in the transmitted tradition that King Wen or other Zhou rulers cited the precedent of Shun to justify the overthrow of the Shang. Moreover, whereas, in other texts, King Wen cites the founding of the Shang, and the evil of their last king, to justify his overthrow, here he discusses the founder of the Shang lineage as a potential model and “obtaining the center” is taken as the means by which rule may be achieved. This, rather than a dynastic cycle, serves as the historical paradigm for obtaining rule, and the legends of Shun and Wei function in this context. The meaning of the center is unclear, but I shall argue that it refers here to the geographical and cosmological center of the world and that this concept was the origin of the later philosophical idea of the “center.” This book has two purposes. One is philosophical—to explore the implications of these recently discovered manuscripts for understanding the development of political philosophy, especially the idea of kingship and the theory of political succession, in early China. To this end, my earlier study, The Heir and the Sage, will serve as a foundation for my analysis, although my concerns herein are not limited to those of that work and I have no over-arching theoretical purpose. The other is textual—to introduce readers, including those without specialist knowledge of excavated texts, to these Chu-script bamboo-slip manuscripts and the problems involved in deciphering them. Thus, I provide a close reading of each manuscript in its entirety, including translation, analysis, and explication, as well as a discussion of historical and philosophical context. Because these bamboo-slip manuscripts are closely related, but not strictly equivalent, to received texts, I also explore the possible implications of these discoveries for the early development of the literature that has been transmitted to us since the Han Dynasty. The following chapter,“The Historical Setting,” discusses the historical background in which the legends of abdication in the predynastic period arose, most importantly the breakdown of the power of the aristocratic lineages. It also discusses the function of ancient historical materials as a means of expressing political theory in pre-Qin transmitted Chinese texts. Finally, this chapter also discusses the significance of the Qin dynasty as a watershed in Chinese textual history.

Introduction 7

Chapter 3, “The Chu-script Bamboo-slip Manuscripts,” provides a general introduction to the three groups of Chu-script bamboo-slip manuscripts, which include the four manuscripts to be discussed in the following chapters. This includes a new hypothesis concerning the circulation of early Chinese manuscripts and the influence of the physical forms of such manuscripts on the development of later literature. It also includes a review of the archaeological information about the occupant of Guodian Tomb One and of the possibility that a woman’s tomb in a nearby village might be the source of the Shanghai Museum collection. Finally, it discusses the process of transcribing the manuscripts for publication and explains the apparatus to my editions of the manuscripts found at the end of each of the following chapters. Chapters 4 through 7 are each devoted to one of the four manuscripts mentioned above. They include an English translation, explication, discussion, and a scholarly edition with the original graphs and citations to different readings at the end of the chapter. Because the manuscripts are different in both literary genre and ideological implication, the issues that they raise are somewhat different and the questions explored in each chapter are consequently somewhat diverse. Thus, in chapter 4, which focuses on Tang Yú zhi dao, I discuss the manner in which historical legend is used to express political philosophy in the Warring States period and explore the relationship of the argument for abdication found in this manuscript to the philosophical positions taken in related transmitted texts, as well as the historical evidence for an actual abdication in the state of Yan. In chapter 5, which focuses on the Zigao, I discuss the nature of early Confucianism, the historical paradigm implicit in taking the rulers of the three dynasties as descendants of women who were supernaturally impregnated, and the possible relationship between the rise of abdication legends and Confucius’s personal reputation. In chapter 6, on Rongchengshi, I focus on its utopian vision of ideal government in high antiquity and the concept of the sage ruler as a cosmic power, who brings harmony to both the natural and human world. Bao xun, discussed in chapter 7, purports to be a“document” and so this chapter discusses the origin and nature of the Shang shu as well as the center as a political, geographic, and cosmological concept. In chapter 8,“Afterthoughts,” I reflect upon the implications of these bamboo-slip manuscripts for the debates over the last century concerning history and textual authenticity, the implications of the paradigm of the cosmic centrality of the ruler, and the relationship between the rise of legends of rule by sage-kings in high antiquity and the reputation of Confucius.

2

History and Historical Legend

Dynastic Cycle and the Conflicting Principles of Rule by Virtue and Rule by Hereditary Right The ideas of a dynastic cycle and a changing mandate of heaven served as the foundation for all Chinese political theory before the fall of the Qing dynasty in the early twentieth century. According to this view of history, each dynasty was founded by a good man and continued hereditarily until a descendent of the founder was so depraved that heaven transferred the mandate to rule to another good man, who then established a new hereditary dynasty. This new dynasty continued in the same pattern, coming to an end with an evil ruler, who was replaced by another good man. The earliest formulation of the idea of a dynastic cycle is found in some of the earliest chapters of the Shang shu 尚書. These chapters are generally considered to date from the Western Zhou Dynasty (ca. 1050–771 BCE).1 There, the Zhou founders seek to legitimize their own rule by citing historical precedent and claim that the founders of the Shang Dynasty (ca. 1600–1050 BCE) overthrew a previous hereditary dynasty, called the Xia. The idea of a dynastic cycle that was thus created embodied the conflicting values of rule by virtue and rule by heredity, so, once it became current, a paradox arose: any hereditary succession could be challenged on the grounds of the ruler’s lack of virtue; and any non-hereditary succession as an illegitimate usurpation of hereditary right. The theory of a changing mandate of heaven attempted to explain this inherent contradiction 1. See chapter 7, 268. 9

10

Buried Ideas

between new rule that was legitimized by the virtue of its founder and rule by hereditary right, thereby limiting its possible manifestations. However, there was always the potential for conflict. A rebel could always claim that the ruler in power was oppressive and immoral; and a new hereditary ruler could always be accused of regicide.2 At its most basic level, the conflict between the heir and sage was one of loyalty to self and family, as opposed to obligation to the state or larger community. This conflict is inherent in any human society, but it took a particular form in the Chinese tradition. A concomitant of the early Chinese system of ancestor worship was that lineages and extended families functioned as religious as well as social units. With the development of philosophical discourse in the Warring States period (475–221 BCE), the inherent contradictions between loyalty to kin and to the state or community were commonly played out as a conflict in which the ethics of filial piety and the demands of kinship solidarity were opposed to loyalty to the state and its ruler. In the case of the ruler, an obligation to the royal lineage was set in opposition to the welfare of the common people. Such tensions are a continuous theme in Chinese literature and history. As I shall discuss below, this conflict between obligation to kin and obligation to the larger community was especially acute after the movement of the Zhou capital eastward in 771 BCE, because of the collapse of centralized rule and of the social class of aristocrats defined by their status within noble lineages. By the fifth century BCE, legends of abdication set in a period before the Xia Dynasty had become prevalent. Accordingly, in the transmitted texts of the Warring States period, history conventionally begins with Yao 堯 as the first ruler. Yao’s son (or sons) was evil, so Yao passed the rule to a good man, Shun 舜, having already appointed him as his minister. Shun similarly rejected his evil sons in favor of the minister who had successfully controlled the flooding, Yu 禹, who then founded the first hereditary dynasty, the Xia. This period—when the legends of Yao and Shun became prominent—is the same as that in which the manuscripts I shall discuss herein were recorded on slips of bamboo. And it is these legends of non-hereditary succession that are the basis upon which I have selected the manuscripts that I will discuss herein. In the transmitted texts of the Warring States period, these legends serve to support rather than challenge the idea of a dynastic cycle because Yao and Shun act as precedents for the founders of the Shang and Zhou dynasties and their evil sons are likened to the last 2. See Sarah Allan, The Heir and the Sage: Dynastic Legend in Early China (San Francisco: Chinese Materials Center, 1981), especially chapter 1.



History and Historical Legend

11

kings of the dynasties that followed. Moreover, Shun and Yu, in their role as ministers before succeeding to the throne, are correlated with the ministers who aided the founders of the later dynasties. In these recently discovered manuscripts, however, we will find that that the legends of abdication of the good to the good bruited about in this period served to promote abdication as an alternative to hereditary rule. This paradigm of abdication is the only alternative to the idea of dynastic cycle found in the Chinese tradition and it did not survive the Qin and Han dynasties as an idea for an alternative form of succession. The Collapse of the Hereditary Aristocracy in the Eastern Zhou Period The Warring States period is well known as one in which “one hundred schools” proposed different political theories and competing visions of an orderly society. Not only was this period full of the political strife to which it owes its name, it was also a period of dramatic social change in which the lineage system of the early Western Zhou period finally collapsed under the pressure of new social developments. The pressures of new economic, technological, political, and social developments—such as increasing urbanization, the use of currency, iron tools and weapons, the appointment of salaried officials, the registration of the population, the rise of mass infantry armies, and the segmentation of hereditary lineages—had entirely undermined the Western Zhou social and political structure.3 The crisis in political legitimacy was thus accompanied by social breakdown. Hsü Cho-yun argued some fifty years ago that a shift in social power began in the seventh century BCE and finally saw the collapse of the old hereditary aristocracy in the Warring States period. Hsü’s argument for dramatic social change was based upon an analysis of the family backgrounds of named figures in the transmitted historical records, especially the Zuozhuan 左傳. He concluded: After 464 B.C.E. most historical figures were selfmade men who rose from obscurity. This trend, together with the decline of the minister class in the late [Spring and Autumn] period, may indicate not only that there was more mobility between classes at the beginning of [Warring States] times, but that the former 3. Mark Edward Lewis, Sanctioned Violence in Early China (Albany: State University of New York Press, 1990), while focusing on sanctioned violence, provides an excellent introduction to many aspects of social and political change in the Warring States period.

12

Buried Ideas

dominant class, the ministers, had already collapsed. The disappearance of old families may be a consequence of the conquest and annexation of many older, smaller states by a handful of newer states. An inspection of the backgrounds of the chancellors of various [Warring States period] states indicates that there were few if any such families. In brief, what happened during the [Warring States] period was the disappearance of the former social stratification, not merely freer mobility between strata.4 Concomitantly, this period saw the rise of the shi 士, a class of “gentlemen” who traced their descent to the noble lineages, but who were not the primary heirs and had little, if any, land. Such gentlemen achieved authority through technical skills, including both military arts and literary culture, and by appointment to official office. The philosophers and their disciples, were, by and large, drawn from this class. This analysis, based on textual records unearthed when Chinese archaeology was still young, has been largely vindicated by material evidence from later excavations. Lothar von Falkenhausen, in his recent book, Chinese Society in the Age of Confucius, synthesizes a vast amount of information on Chinese mortuary practices from archaeological excavations. He argues that, because of the prevalence of lineage segmentation, the social hierarchy of the Western Zhou period quickly began to break down and that there were two major attempts to realign the sumptuary rules, one in the late Western Zhou (ca. 850 BCE) and again in the middle of the Spring and Autumn period; that is, the social changes noted by Hsü Cho-yun were marked somewhat earlier in the burial system. Von Falkenhausen states, “In any segmentary lineage society, descent is the decisive criterion in negotiating social inequality. . . . Continuity of descent from as prestigious as possible an ancestral figure in the distant past—and seniority among those descended from that ancestor—entailed access to privilege and power. Nevertheless, the segmentation of the lineages gradually led to the destruction of their religious and ritual authority.” Although the second “Ritual Restructuring,” which occurred in the middle of the Spring and Autumn period, attempted to bring ritual in line with social realities, the character of the social distinctions had changed too dramatically. The new distinctions were no longer between ranked and unranked members of a lineage, but simply between rulers and the ruled.5 4. Hsü Cho‑yun, Ancient China in Transition: An Analysis of Social Mobility, 722–222 B.C.E. (Stanford: Stanford University Press, 1965), 38. 5. Lothar von Falkenhausen, Chinese Society in the Age of Confucius (1000–250 BC): The Archaeological Evidence (Los Angeles: Cotsen Institute of Archaeology at UCLA, 2006), 70, 395, and elsewhere.



History and Historical Legend

13

Changes in religious beliefs became manifest in mortuary practices in the period of transition from the Spring and Autumn to the Warring States period, that is, in the sixth century BCE. This is reflected both in the construction of the tombs and especially in a change in the artifacts buried within them. Whereas tombs had been primarily repositories of ritual artifacts, they were increasingly transformed into subterranean models of the world of the living. By the middle of the fifth century BCE, these changes had come to be expressed in an increasing emphasis on luxury items and objects associated with daily life, as opposed to ritual paraphernalia.6 Bronze vessels, even when decoratively inlaid with gold and silver, are still essentially ritual artifacts and this is reflected in the ancestral dedications of their inscriptions. Bamboo-slip manuscripts, on the other hand, were artifacts meant for the living that were buried with the dead. This collapse of the old political, social, and religious order from the fifth century onward provided fertile ground for thinking about political theory and ethical problems, and the ideas formulated in this period became the foundation of later Chinese tradition. The theory of a changing mandate of heaven found in the Warring States period had originated in the claim of the early Western Zhou kings that an astronomical phenomenon during the reign of the last king of the Shang Dynasty signified that Shang Di 上帝, the “supreme thearch,” had transferred the right to rule over all-under-heaven (tianxia 天下) from the Shang rulers to their own lineage.7 The Zhou further proclaimed that, in defeating the last Shang king, Zhòu Xin 紂辛, they were simply repeating a pattern that had been

6. Falkenhausen, “Social Ranking in Chu Tombs: The Mortuary Background of the Warring States Manuscript Finds,” Monumenta Serica 51 (2003): 293–325. 7. See Sarah Allan, “On the Identity of Shang Di and the Origin of the Concept of a Celestial Mandate,” Early China 31 (2007): 1–46; Yuri Pines, “Changing Views of tianxia in Pre‑imperial Discourse,” Oriens Extremus 43, no. 1–2 (2003): 101–16, argues that the term tianxia does not appear—or does not become significant—until about the sixth century BCE. I believe, however, that the idea of“all‑under‑heaven” as the appropriate realm of the king is implicit in the use of tian “sky” or “heaven” as a euphemism for Shang Di and of the term “son of heaven” (tianzi 天子) for the kings, both of which frequently occur in the early Western Zhou. The absence of the term in bronze inscriptions is because of the nature of their content. While tianxia certainly had cultural implications and in some contexts might exclude those peoples who were culturally or ethnically distinct, it was nevertheless a political claim of rule over the entire civilized world and the cosmos more generally. That the term is not found extensively in transmitted literature until the end of the Spring and Autumn period reflects the nature of the texts and its growth in importance in the political circumstances of interstate warfare.

14

Buried Ideas

established by the founder of the Shang Dynasty, Cheng Tang 成湯, when he defeated Jie 桀, the evil last ruler of a previous dynasty. I have argued previously that this claim by the Zhou founders originated in the historicization of a Shang myth of an earlier people, the Xia, who were their opposites, dark to their light, moons to their suns, dragons to their birds, and so forth.8 Whatever its origins and purpose, with history cast in this manner, the Zhou became the third dynasty in a cyclical pattern, with the implication that hereditary dynasties, including their own, would eventually be destroyed and replaced by celestial power once they declined to the extent that the rulers oppressed the people and/or engaged in immoral behavior. Thus, the Zhou claim to legitimacy laid the foundation for the paradox embodied in the idea of a changing mandate in which the right to rule was claimed as a hereditary right, but established by and dependent on virtue. However, the pattern of subsequent Zhou history did not conform to the expected model. The Zhou kings declined in power as the network of states, which they had established at the beginning of the dynasty, gained in authority.9 In the later historiography, King Li 厲 (r. 878–828 BCE) is described as evil and is correlated with the bad last kings of the Xia and Shang, but no one was able to claim his mandate. As Zhou power declined, the court was forced to move from the capital near Xi’an in Shaanxi eastward to Luoyang in Henan Province in 771 BCE. Rulers of the most powerful states began to call themselves“kings”(wang 王) but still the royal Zhou lineage did not die out. After the move east, the dynasty lasted for another five hundred years before“all-under-heaven”was again unified by the king of Qin in 221 BCE. Expecting that a new dynasty might arise, Confucius (551–479 BCE) had looked for a future king whom he might serve and searched for signs of a new mandate, but, alas, “the phoenix did not arrive; the river did not bring up a chart”—and he finally accepted that a new mandate of heaven would not be awarded in his own time.10 In the following centuries, as Zhou power continued to diminish and multiple power centers rose, 8. Sarah Allan, “The Myth of the Xia Dynasty,” Journal of the Royal Asiatic Society 2 (1984): 242–56; The Shape of the Turtle: Myth, Art, and Cosmos in Early China (Albany: State University of New York Press, 1991), 57–73. 9. Li Feng. Landscape and Power in Early China: The Crisis and Fall of the Western Zhou, 1045–771 B.C. (Cambridge: Cambridge University Press, 2006), provides an insightful account of the reasons for the decline of Zhou authority after the beginning of the dynasty. 10. Lun yu jishi 論語集釋 (Beijing: Zhonghua, 1990), juan 17, 588 (9.9).



History and Historical Legend

15

which then struggled with one another for supremacy, the explanatory force of the theory that a dynastic cycle in which a man of virtue from one hereditary lineage replaced an oppressive king of a different hereditary lineage became increasingly strained. By the time of Mencius—and the occupants of the tombs in which these manuscripts were buried—the Zhou no longer exercised sufficient political control over the states for the overthrow of the Zhou king to result in the establishment of a new dynasty. This historical and ideological crisis appears to have provided the context for a predynastic period in which sage-kings practiced abdication as a means of political succession. The Question of Monarchism Yuri Pines, in Envisioning Eternal Empire and articles relating to some of the manuscripts discussed herein, has argued for the development of an idea of an omnipotent monarch in this period as the key to later Chinese political theory.11 He sums up his argument as follows:“The Warring States period was an age of rapid proliferation of monarchism as the guiding political principle and the most important ideological construct. Thinkers of various convictions and intellectual affiliations came to an almost unanimous conclusion that preservation of sociopolitical order would be impossible unless ‘all under heaven’ is unified under the aegis of a single omnipotent Monarch.”12 Accordingly, Pines takes arguments for abdication as an alternative means of succession to hereditary transfer in these Warring States period manuscripts to be proof of the supreme ideological importance of the omnipotent monarch, because the theoretical challenge was confined to the manner of choosing the successor, not to the role of the single and uniquely powerful ruler. It is true that philosophers in the Warring States period generally agreed that only unification could put an end to the raging civil wars. Chinese cosmology emphasized the center from at least Shang times, in which Shang Di was identified with the Pole Star who ruled 11. Yuri Pines, Envisioning Eternal Empire: Chinese Political Thought of the Warring States Period (Honolulu: University of Hawaii Press, 2009); “Disputers of Abdication: Zhanguo Egalitarianism and the Sovereign’s Power,” T’oung Pao 91, no. 4–5 (2005): 243–300; “Subversion Unearthed: Criticism of Hereditary Succession in the Newly Discovered Manuscripts,” Oriens Extremus 45 (2005–06): 169–75. Pines’s theory is strongly influenced by Liu Zehua 劉澤華, Zhongguo de wangquan zhuyi 中 國的王權注意 (Shanghai: Shanghai Renmin, 2000). 12. Yuri Pines, “To Rebel Is Justified? The Image of Zhouxin and the Legitimacy of Rebellion in the Chinese Political Tradition,” Oriens Extremus 47 (2008): 8.

16

Buried Ideas

heaven, whereas the king, who modeled himself after Pole Star, ruled all-under-heaven.13 In the cosmological vision of the Warring States period, the centrality of rulership and the significance of the center was neatly expressed by the invention of the cosmograph (shipan 式盤) with its square earth, attached by a pivot at the center to a round, rotating sky.14 Metaphorically and conceptually, centralized rule and a cosmology that focused on the power of the center were integrally related. However, the concept of a single monarch was also a logical response to the historical situation, especially after a series of alliances between states headed by hegemons had failed to bring peace. How else could the fighting between competing states be brought to an end except by submission to a single ruler? Monarchism per se is not unique to China, or even particularly Chinese; the vast majority of ancient societies seem to have had some form of hereditary kingship. The conceptual alternative provided by Pines is aristocratic oligarchy, but this is surely a much rarer political ideal in world history than that of a single king. What is unique to the Chinese tradition is not the idea of omnipotent monarchy or even the cosmic centrality of the king—these are common among ancient civilizations— but the idea of a dynastic cycle, in which one dynasty is founded by merit and then carried on hereditarily until a depraved last king is overthrown by a good man, who then establishes a new hereditary dynasty. From the beginning of the Western Zhou Dynasty on, political thinkers assumed that there could only be one king who stood at the center of the cosmos, was sanctioned by heaven, and faced south as his ministers faced north. Among the manuscripts that I will discuss herein, this idea of kingship and centrality as necessary concomitants of one another is evident in Rongchengshi and, especially, in the Bao xun. However, the unusual feature of the political ideology sown when the Zhou overthrew the Shang was not the central role of the king, but its dualism: namely, that the hereditary principle is inevitably balanced by one of rule by virtue. The heir always requires a sage; the king, a minister; and the evil hereditary descendant of a virtuous ruler must be replaced by a man of virtue. The seeds of this dualism are found in the texts and inscriptions of the early Western Zhou that explain the Zhou overthrow 13. See Sarah Allan, “On the Identity of Shang Di.” I will return to the issue of the importance of the center in chapter 7. 14. See Sarah Allan, “The Great One, Water, and the Laozi: New Light from Guodian,” T’oung Pao 89, no. 4–5 (2003): 237–85.



History and Historical Legend

17

of the Shang in terms of a changing mandate of heaven. This theory was developed later in the context of the historical changes that took place in the Spring and Autumn and Warring States periods and in the philosophical responses to those changes. As these manuscripts make clear, among the more experimental thinkers of this period were some who argued instead for rule by merit as opposed to hereditary right. This was to be accomplished by a good ruler selecting another good man as his successor, rather than by hereditary transfer. The Legend of Yao and Shun and the Function of Historical Legend in Transmitted Texts Legends of a predynastic era in which Yao abdicated to Shun and Shun to Yu, the founder of the Xia Dynasty, become prominent in the transmitted tradition around the fifth century BCE, that is, at the time that the transmitted tradition begins to include philosophical texts. How these legends arose is not clear. As noted above, I hypothesized early on that the legend of Yao and Shun originated with a Shang myth that the supreme thearch, Shang Di, appointed the first ancestor of the Shang Dynasty to rule allunder-heaven.15 However, there is so little transmitted literature between the end of the Shang Dynasty and the fifth century that it is impossible to trace how or when this transformation took place. Gu Jiegang 顧頡剛 proposed that the abdication legends were a creation of the Mohists, who lived in the fifth century and advocated the promotion of the meritorious (shang xian 尚賢).16 We shall see in the course of this work that there is no evidence in the discussions of abdication in these bamboo-slip manuscripts of an association with Mozi or his followers—or any particular philosophical school. Nevertheless, the Zigao implicitly links Confucius and Shun, and I shall argue for a link between the rise of the legend of Confucius as a sage, who was more meritorious than any king, and the legends of Yao’s appointment of Shun as his successor. In the transmitted texts, the rulers and ministers of the predynastic era are correlated with those of the dynastic era as exemplars of the same type. Moreover, the legends of transfer of rule balance the conflicting 15. Sarah Allan,“Sons of Suns: Myth and Totemism in Early China,” BSOAS 44, no. 2 (1981): 290–326; Allan, The Shape of the Turtle, 31ff. 16. Gu Jiegang 顧頡剛, “Shanrang chuanshuo qi yu Mojia kao 禪讓傳説起於墨家 考,” in 古史辨 (Shanghai: Shanghai Guji, 1982 [first published 1926–1941]), vol. 7, 30–109; see also the rejoinder by Yang Kuan 楊寬, 110–17.

18

Buried Ideas

demands of rule by virtue and rule by hereditary right. This is true even in the Mozi. There are five critical transfers of rule from the time of Yao to the formation of the Zhou Dynasty: (1) the non-hereditary transfer from Yao 堯 to Shun 舜; (2) the non-hereditary transfer from Shun to Yu 禹; (3) the hereditary succession of Qi 啟 and the establishment of the Xia Dynasty; (4) the non-hereditary transfer from Jie 桀 to Cheng Tang 成湯 and the establishment of the hereditary Shang Dynasty; and (5) the non-hereditary transfer from Zhòu Xin 紂辛 to Wu Wang 武王 and the establishment of the hereditary Zhou Dynasty. At each of these critical junctures, themes of heredity and virtue are expressed and mediated by the sequence of events. In accepting the rule from Yao, Shun is not only juxtaposed to the evil, but rightful, heir Dan Zhu丹朱; he is also contrasted to Xu You 許由, who in his purity refused the rule that was not his by hereditary right. Similarly, the Shang founder Tang 湯 is contrasted with the pure figures, Wu Guang 務光 and Bian Sui 卞隨, while the Zhou founder Wu Wang is juxtaposed to Bo Yi 伯夷 and Shu Qi 叔齊, who starved themselves to death rather than eat the grain of Zhou after Wu had assumed a throne to which he had no hereditary right. Shun also stands as a mediatory figure between the principles of kinship obligation and hereditary right in terms of his personal characteristics because, even though he displaced Yao’s son by assuming the throne, he was a model of filial piety in his conduct toward his own father, Gu Sou 瞽瞍. Gu Sou, having remarried, tried to murder Shun on several occasions, in league with Shun’s half-brother, Xiang 象. Shun nevertheless remained filial toward his father and fraternal toward his brother, showing no resentment of their evil deeds. This story is at the heart of Tang Yú zhi dao and provides the twist that is developed by the manuscript to an unusual purpose; that is, to advocate abdication to a man of merit as the best means of loving one’s kin. Although the legends of dynastic change have certain features that remain constant—what I call their “deep structure,”they could be interpreted variously, thus giving rise to the various transformations of the legends found in individual texts. Accordingly, the particular citations of legend found in particular texts serve to express philosophical arguments. Because the changes were systematic and themes repeated, with legend figures of different eras often correlated with one another, it was possible to predict from a given philosopher’s account of any particular legend about transfer of rule how he was likely to interpret transfers set in other eras. Thus, a philosopher who described the transfer of rule from Yao to Shun as forcible usurpation would describe the transfer of rule from Shun to Yu in a similar



History and Historical Legend

19

manner, and not take the position that Shun or Yu abdicated the rule. Nor would such a philosopher take the position that the transfer was affected by the movement of the people from one ruler to another. If, like Mozi, a philosopher stressed Shun’s low social status—his working in the fields before Yao abdicated to him—he would also take the view that Tang raised up Yi Yin 伊尹 from the kitchen and that Wen Wang raised up Taigong 太公 from the butcher’s market or riverbank.17 However, if he stated, as in the Mencius, that Shun went to farm at Li Shan in a deliberate attempt to transform the people who had quarreled over the boundaries of their fields, he would also hold the opinion that Yi Yin and Taigong Wang were eremitic gentlemen waiting in retirement for a true king. If Shun usurped the rule from Yao, then Wu Wang also violated hereditary right in overthrowing Zhòu Xin. While individual transmitted texts are not absolutely consistent in this respect—any more than they are entirely consistent as philosophical works—they are generally so. While these facts remain constant—no one ever claimed, for example, that Shun was really Yao’s son—the story of how Shun came to succeed Yao and the meaning of the legend undergo regular transformations. There are three major forms in the received tradition: (1) Yao gave the rule to Shun. This is the form of the legend found in Tang Yú zhi dao. In the received tradition, Mozi is its representative text, but, as I will discuss in more detail in the following chapter, Tang Yú zhi dao opposes the core Mohist doctrine of jian ai 兼愛 (“love without distinction”). (2) Yao could not have bestowed the rule on Shun. It was not in his power to do so, because only heaven can bestow the rule and it expresses its changing command in the changing allegiance of the people. This transformation is found in the Mencius and the Xunzi. This difference distinguishes them from Tang Yú zhi dao and, I will argue, it was formulated to oppose the argument for abdication made by Tang Yú zhi dao and other contemporary texts. (3) Shun forced or inveigled the rule from Yao. This last transformation is found in the Han Feizi, and I will argue below that it serves there to oppose the idea of abdication as a means of succession. This transformation is also found in the Guben zhushu ji nian 古本竹書紀年, but both the Han Feizi and Zhushu jinian also attribute 17. In Sarah Allan, “The Identities of Taigong Wang in Zhou and Han Literature,” Monumenta Serica 30 (1972–73): 57–99, I argue that Taigong’s different roles as butcher, fisherman, etc., reflect regional legends that function in a similar manner to portray Taigong as a “founding minister.”To be reprinted as an appendix in the new edition of the The Heir and the Sage, SUNY, forthcoming.

20

Buried Ideas

subsequent transfers of rule in the ancient period to the use of force and violence. Since the transformations of the legends that include force seem more realistic to many scholars than abdication, they take their realism as evidence of historicity. In light of the textual tradition as a whole, however, they may be understood as a set of transformations of the legends associated with philosophers such as Han Feizi; that is, they are no more valid per se than the forms of the legends found in philosophical texts.18 In the transmitted tradition, the legends of abdication in the predynastic period are always tied to legends of dynastic foundation so that they serve to support, rather than challenge, the idea of hereditary dynasties founded by a changing mandate of heaven. Moreover, even though some figures, such as the Yellow Emperor, who supposedly lived before the time of Yao, occasionally occur in pre-Qin transmitted texts, they do not occur as exemplars in the correlated sets of dynastic rulers and their ministers of the predynastic and dynastic periods found in the philosophical texts. In contrast, as will be seen in the course of this book, the bamboo-slip manuscripts reveal a more fluid idea of the predynastic past than found in the transmitted tradition and a greater variety of political stances. Indeed, the very idea of hereditary dynasties as an ideal form of government was challenged by at least a few thinkers in the Warring States period. These thinkers proposed an alternative form of succession: abdication to the most meritorious. This theory was a product of the collapse of the hereditary aristocracy and the rise of a new literati class in the Warring States period, as well as of the lack of a central authority. Such a theory was inherently subversive to any hereditary dynasty and represented too radical a challenge to be tolerated once centralized rule was established in the Qin and Han; it is, therefore, not found in the transmitted texts. Nevertheless, in light of these manuscripts, we can now see that many of the arguments concerning kingship in the transmitted tradition were counterarguments to the idea of abdication. The Qin Watershed The Qin dynasty (221–206 BCE), in which most of the known world was finally unified under a single ruler after a longer period of warfare, was a watershed in the history of Chinese texts as well as a political one. The 18. The systematic association of legend transformations according to philosophical school is discussed in The Heir and the Sage; see especially, 123–40 (ch. 6). I believe that, rather than a record of events, the Zhushu jinian was a Legalist account of ancient history.



History and Historical Legend

21

destruction of texts by the First Emperor (Shi Huangdi 始皇帝) is one of the most famous events in Chinese history. According to the Shi ji 史記 (“Grand Scribe’s Records”), after the First Emperor of Qin had unified allunder-heaven, he was distraught because scholars continued to cite the models of ancient sage kings to criticize his appointments and policies. He consulted his minister, the Legalist, Li Si 李斯 (ca. 280–208 BCE), who advised him that such scholars “used the past to harm the present”; he warned the emperor against the practice of private study, because such study enabled people to construct arguments based upon their personal learning to criticize governmental laws and edicts. Such people “took personal learning as good and used it to criticize what their superiors had established.”19 Warring States philosophers consistently transformed historical legend to present an ideal vision of society that corresponded to their own ideas of meritorious rule. This rhetorical technique, which used the authority of the past to present a critique of current rule as well as a theoretical ideal, was inherently problematic to the new Qin Dynasty. As Mark Lewis wrote in Writing and Authority in Early China,“[T]extual production in the Warring States period. . . created groups separate from the state who nevertheless claimed for themselves the authority of monarchs or ministers. Over the centuries they created philosophical texts that articulated their claims to power in the guise of sages, histories of a past that served both as a ground for criticisms of government and a realm in which writing became judicial authority or even kingship, and new types of poetry that created authoritative voices not tied to political office. In all these ways texts created a vision of kingship and nobility that paralleled or negated existing state forms.”20 In light of the radical challenge to hereditary authority implicit in the legend of Yao’s abdication to Shun and the foundation legends of the dynastic era in many transmitted texts, as well as in the manuscripts that I will discuss in the following chapters, the Qin concern that arguments based upon historical precedent were a threat to their power was clearly warranted. To protect his newly unified empire, the First Emperor accordingly issued an edict that all works in personal libraries should be surrendered to the government or local commandery within thirty days and burnt. More 19. Shi ji 史記 (Beijing: Zhonghua, 1973 [1959]), juan 6, 254–55; juan 87, 2546–47. 20. Mark Lewis, Writing and Authority in Early China (Albany: State University of New York Press, 1999), 363. The Guodian manuscripts were first published in 1998. Although Lewis’s book was published in 1999, he appears to have completed it before their publication as he makes no reference to them.

22

Buried Ideas

specifically, the Songs (shi 詩) and Documents (shu 書) and the“words of the one hundred schools”(baijia zhi yan 百家之言) were banned for discussion as well as personal possession. Exceptions included histories that glorified the state of Qin and practical works on subjects such as medicine, divination, agriculture, and tree planting. The destructive effects of this edict on textual transmission were compounded when the Qin capital, including the imperial library, was razed in 207 BCE. Historians have usually portrayed the “burning of the books,” which took place in 213 BCE, as a move against the followers of Confucius, with whom the Songs and the Documents were particularly associated. Precisely what was destroyed under this edict and the degree to which it was effective has been the subject of much revisionist research in recent years. This research has pointed out that the edict was aimed at the“hundred schools” rather than the followers of Confucius per se, and particularly at personal libraries, exempting those of court academicians.21 With the discovery of these bamboo-slip manuscripts, we are beginning to have a clearer idea than previously possible of the range and types of literature that were in circulation in the fourth century BCE. We do not know whether the manuscripts that I will discuss in the following pages were deliberately destroyed (or even widely circulated). Nevertheless, they were manuscripts in private libraries composed by scholars of the“hundred schools”and they were potentially subversive to dynastic rule. Thus, they fit the description of the edict. They also give us a better idea of the type of literature being informally circulated and what was at stake politically when the edict was issued. A more insidious but no less significant form of destruction of the ancient literary tradition that has received less scholarly attention took place after the formation of the Han Dynasty when the texts were rewritten in a standardized script. In the Western Zhou Dynasty, bronze inscriptions from the different states do not show much regional variation in script style. However, during the Eastern Zhou, with the decline of centralized power, the scripts of the different states began to show more divergence, particularly those of the eastern states. Literacy also became more widespread and handwriting—performed with a brush on wood, bamboo, silk, 21. Jens Østergård Petersen. “Which Books Did the First Emperor of Ch’in Burn? The Meaning of Pai Chia in Early Chinese Sources,” Monumenta Serica 43 (1995): 1–52; Martin Kern, The Stele Inscriptions of Ch’in Shih‑huang: Text and Ritual in Early Chinese Imperial Representation (New Haven: American Oriental Society, 2000), 183–96; Michael Nylan, The Five “Confucian” Classics (New Haven: Yale University Press, 2001), 30–31.



History and Historical Legend

23

and stone—served many more informal purposes. Thus, regional script styles and unconventional graphic variants evolved.22 After the Han had replaced the Qin, scholars sought to restore the destroyed texts, but the process of restoration was inevitably compromised because the script had been standardized, first by the Qin and then by the Han, on the basis of Qin forms. In the Han Dynasty, the history of the development of the script from oracle bone and bronze inscriptions of earlier periods, now familiar to us from archaeological excavation, was not so well understood. Furthermore, after the standardization of the script in the Qin and Han, the number of variant graphs was drastically reduced. As regional variants fell out of use, they gradually became unrecognizable. The manuscripts that survived from the pre-Qin period were called “old script” (gu wen 古文) texts, as opposed to “modern script” (jin wen 今文) ones, written in contemporary, that is, Han, character forms. These terms, which were initially simply descriptive, came to be associated with different Han Dynasty editions of the classics and took on political overtones, the repercussions of which have lasted up to modern times.23 However, all texts that survived into the Han Dynasty were transcribed in the new standardized script; that is, when “old script” texts were discovered, they were not transmitted in their original old script form, but rewritten in the Han script. Historically, much attention has been paid to the problem of the production of forgeries that claimed the authenticity of old script editions of ancient texts. However, putting the question of forgery aside, the process of transcribing ancient texts in the Han script inevitably resulted in rewriting them to a certain extent. Some of this would have been because of unfamiliarity with old script forms or the same type of ambiguities in the Warring States period scripts that plague scholars attempting to transcribe them today. Moreover, when transcribing a text in the Han script, if a scholar made changes for the sake of clarity or consistency—or to stress some point in the text that he thought was implicit—he may not have 22. See Qiu Xigui 裘錫圭, Chinese Writing, trans. Gilbert L. Mattos and Jerry Norman (Berkeley: Society for the Study of Early China and Institute of East Asian Studies, University of California, Berkeley, 2000), 78–112, for discussion of the evolution of the script. 23. These usages of the terms jinwen and guwen and their ideological implications are laid out succinctly in Wang Baoxuan 王葆玹,“The Study of Ancient and Modern Text Classics: Dispute and Implications,” Contemporary Chinese Thought 36, no. 4 (Summer 2005): 58–81. See also Hans van Ess, “The Old Text/New Text Controversy: Has the 20th Century Got It Wrong?” T’oung Pao 80 (1994), 146–70.

24

Buried Ideas

considered that he was altering the text. And, when the manuscript at hand was inconsistent with his preconceived ideas about what it should say, he may sometimes have assumed that the text had been corrupted and altered it accordingly. That is, many—or indeed most—alterations were probably not intentional forgeries but editorial changes or misguided attempts to restore an original text or meaning. Indeed, much of the commentarial tradition for the next two thousand years has been devoted to correcting—or attempting to correct—problems that have resulted from this process. The Han rewriting of old script manuscripts would also have involved a process of compilation and selection. Indeed, as I shall discuss in the following chapter, the “reconstruction” of ancient texts in the Han, as well as the production of new works, was likely to have been, in many cases, the construction of long texts from an assembly of shorter manuscripts. It is these long texts that have, by and large, been handed down to us. In compiling the long texts from short manuscripts, there would also have been alterations for the sake of literary and stylistic consistency.

3

The Chu-script Bamboo-Slip Manuscripts

Many excavated Chu tombs include inventories, written on bamboo slips, of the artifacts buried with the deceased and some have bambooslip manuscripts, including such texts as “daybook” almanacs (ri shu 日書) and other divinatory texts.1 Fragmentary versions of texts similar in kind to those found in Guodian Tomb One and in the Shanghai Museum and Tsinghua University collections have also been found.2 These three groups of texts are unique, however, because of the combination of three factors: their content, their early date, and the unusual quality of their preservation. 1. Summaries of manuscript discoveries include Xilin Zhaoyi (Nishibayashi Shô­ichi) 西林昭一 and Chen Songchang 陳松長, Xin Zhongguo chutu shuji 新中國出土書蹟 (Beijing: Wenwu, 2009), 53–140; Pian Yuqian 駢宇騫, Jianbo wenxian gaishu 簡帛文 獻概述 (Taibei: Wanjuanlou, 2005), 5–38; Hu Pingsheng 胡平生and Li Tianhong 李 天虹, Changjiang liuyu chutu jiandu yu yanjiu 長江流域出土簡牘與研究 (Wuhan: Hubei jiaoyu, 2004), 56–217; Zhang Xiancheng 張顯成, Jianbo wenxianxue tonglun 簡帛文獻學通論 (Beijing: Zhonghua, 2003), 26–83. For summaries of pre‑Han excavated manuscripts in various mediums, see Zhao Chao 趙超, Jiandu boshu faxian yu yanjiu 簡牘帛書發現與研究 (Fuzhou: Fujian Renmin, 2005), 13–34; Enno Giele, Database of Early Chinese Manuscripts, http://www.lib.uchicago.edu/early china/ res/databases/decm, updated 1 September 2000 (no longer online), listed 158 sites (sites.htm) and 287 bamboo‑slip manuscripts (including tomb inventories). A slightly earlier printed version of this database is found in Enno Giele, “Early Chinese Manuscripts: Including Addenda and Corrigenda to New Sources of Early Chinese History,” Early China 23–24 (1998–99): 247–338. 2. Fragments of Warring States period philosophical texts were also found at Xinyang, Henan Province, but their state of preservation was insufficient for

25

26

Buried Ideas

Their content is closely related to that found in the core philosophical and historical texts of the Chinese classical tradition. They were buried around 300 BCE, which corresponds to the approximate date of the death of Confucius’s most famous follower, Mencius, so they were written—or at least written down—in the most fertile period of that tradition. Because they predated the Qin watershed, they are our first glimpse of what Chinese texts were like before they were rewritten in modern script of the Han Dynasty. Moreover, the slips and the ink calligraphy are remarkably well preserved, so it is possible to reconstruct the texts with relatively few gaps. One of the many surprising aspects of these corpuses of Warring States period bamboo-slip manuscripts is that most of the texts recorded on them are very short, resembling the chapters or sections of chapters in transmitted texts (zhang 章 or pian 篇), rather than the multi-chapter works that have been passed down since the Han Dynasty. One possible explanation is that the mourners selected sections of such works for the purpose of burial, the chapters acting as samples that stood for the whole. However, the bamboo scrolls in each of the three corpuses were written by many different calligraphers and the slips that made up the scrolls were of different lengths and had endings of various shapes. This lack of uniformity suggests that each group of manuscripts represents a collection made by the deceased person for his own use and then buried with him, rather than manuscripts commissioned specially after his death for burial.3 Moreover, if single manuscripts were selections taken from long texts, we would expect each bamboo-slip scroll to include one or more parts of the same text. However, many scrolls include more than one manuscript and these manuscripts often have different rhetorical styles and content.4 Concomitantly, although transmitted early Chinese literature generally consists of long, multi-chapter works, one of the peculiarities of reconstruction of the texts; see Henansheng Wenwu Yanjiusuo 河南省文物研究 所, ed., Xinyang Chu mu 信陽楚墓 (Beijing: Wenwu, 1986). Still unpublished manuscripts were also found in Tomb 36, Shibancun 石板村, Cili 慈利, in 1987; see Zhang Chunlong 張春龍, “Cili Chu jian gaishu 慈利楚簡概述,” in Xin chu jianbo yanjiu: Xin chu jianbo guoji xueshu yantaohui wenji 新出簡帛研究: 新出簡帛國際學 術研討會文集, ed. Ai Lan 艾蘭 (Sarah Allan) and Xing Wen 邢文 (Beijing, Wenwu, 2004), 4–11. 3. Scholars generally assume that the manuscripts in the Shanghai Museum and Tsinghua University collections were from single tombs though this cannot be proved. 4. For discussion of the relationship between Zigao and the two other manuscripts with which it was bound, see chapter 5.



Th e C h u - s c r i p t Ba m b o o - S l i p M a n u s c r i p t s

27

these transmitted texts is that they are generally made up of bits and pieces—small sections of text that are linked only by their genre or by a master and associated disciples and related persons. These small sections of text that make up the longer works are, by and large, arranged without any clear connecting logic. Moreover, there are no cross-references or other indications that the author of a particular section of text was aware of what else was included in the long work. The form of these unearthed manuscripts thus provides a key to understanding the process of text formation in the Warring States period and the contribution made by Han period editors to the formation of the classical texts that have been transmitted to us. This chapter will briefly review each of the three corpuses of bambooslip manuscripts, including the information about the sources of the manuscripts to the extent that it is available, as well as the contents. It will also discuss methods of deciphering bamboo-slip manuscripts written in the Chu script. First, however, I will present a tentative hypothesis concerning the development of Chinese literature in the Eastern Zhou Dynasty in light of the physical mediums used to record early Chinese texts. This hypothesis provides an explanation for some of the characteristic features of transmitted texts as well as the form of these newly discovered manuscripts. The Formation of Multi-chapter Texts: A Hypothesis As Li Ling has observed,“Most ancient writings from the early period were compiled from ‘short passages’ (that is, fragmentary passages and sentences) according to the occasion, so, when they circulated, they often lacked a unified structure, and hence there was a great chance that they were rearranged and additions and alterations made to them. There was no stability in how these textual units were divided up or combined with each other, and what got lost or survived was subject to chance.”5 Accordingly, we may suppose that early Chinese texts initially circulated as small units; that is, the multi-chapter texts that were transmitted from the Han Dynasty were not composed as whole works, but were collections of shorter manuscripts arranged in sequence. 5. Li Ling 李零, Jianbo gushu yu xueshu yuanliu 簡帛古書與學術源流 (Beijing: Sanlian, 2008), 214. My translation is based on that of Matthias L. Richter, The Embodied Text: Establishing Textual Identity in Early Chinese Manuscripts (Leiden: Brill, 2013), 66. See also William G. Boltz, “The Composite Nature of Early Chinese Texts,” in Text and Ritual in Early China, ed. Martin Kern (Seattle: University of Washington Press, 2005), 50–78.

28

Buried Ideas

My hypothesis is that the key to this process is the physical materials that were commonly used for writing and circulating manuscripts. These include wood (or bamboo) tablets (du 牘), bamboo- (or wood-) slip scrolls (ce 册, pian 篇), and silk scrolls (juan 卷) (see Figures 3.2. 3.3, and 3.4 below). People initially recorded relatively short units of text on tablets and bamboo slips and circulated them in these forms. These units might be collected together because of a perceived relationship. However, although such collections of similar units might be circulated together, they tended to be open-ended and the passages did not have a definite sequence. Thus, they were not “texts” in the sense that they were repeatedly transmitted in the same form. However, when long texts compiled from these short manuscripts were written on long silk scrolls, they acquired a set, stable form. When this process began is uncertain, but it was on-going in the fourth century BCE when these three groups of manuscripts were buried. Moreover, it came to fruition during the Han when earlier manuscripts were rewritten on silk in the modern script and deposited in the imperial library. Bamboo-slip Scrolls The manuscripts from Guodian Tomb One and those in the collections of the Shanghai Museum and Tsinghua University were all written using a brush and ink on “slips”—or “strips,” as they are sometimes called—of bamboo. When they were buried, the slips were bound together with two or three cords and rolled up as scrolls. By the time they were excavated, the cords had deteriorated. However, some slips had notches to hold the cords or blank spaces and it appears that most bamboo-slip scrolls were

Figure 3.1. Writing brush and holding tube, Baoshan Tomb Two.

Figure 3.2. Bao xun 保訓, Slips 1–11, Tsinghua University collection.

30

Buried Ideas

bound before writing, although scribes sometimes wrote on slips that were later bound together. Such scrolls are similar to silk scrolls except that they could not hold as much writing because of their bulk. Wood, bamboo, and silk are all highly perishable mediums. Therefore, the archaeological record provides little evidence with which to establish when they were first used for writing. Especially important is that conditions in the core cultural region along the Yellow River are such that perishable materials are not easily preserved there. Nevertheless, there is indirect evidence from the divinations inscribed on bone and shell (the so-called oracle-bones) that bamboo slips were already used for writing in the Shang Dynasty (ca. 1600–1050 BCE). For example, the Shang oracle-bone inscription graph, transcribed with the modern character dian 典, includes a depiction of bamboo slips tied together. To this may be added two hands holding the manuscript, a mouth (口), or an altar (示):

Figure 3.3. The dian-rite involves written promises offered to the ancestors in prayer for assistance (or to remove a curse), so the bamboo slips may have been placed on an altar and the text, possibly, read aloud. We do not know, however, whether the Shang used bamboo slips to record texts that were not associated with ancestral rites and divination. There is also considerable evidence in the inscriptions on bronze vessels of the Western Zhou Dynasty (ca. 1050–771 BCE) for the use of bamboo slips in official records. When the king made official appointments, he or a surrogate read out the appointment from a bamboo-slip document, a copy of which was conferred upon the appointed subject at the end of the ceremony. This ceremony is called, ce ming 册命. The graph for ce is a pictograph of bamboo slips tied together, as in the oracle-bone form of dian shown above. As I will discuss in more detail in chapter 7, these speeches must have been composed in writing beforehand so that they could be delivered orally in court. This practice of composing speeches in writing for oral delivery provides a context for the origin of shu 書,“documents,” such as those found in the transmitted Shang shu 尚 書, which can be considered the first literary compositions. Significantly, such documents were originally discrete texts. Only later were they brought together.



Th e C h u - s c r i p t Ba m b o o - S l i p M a n u s c r i p t s

31

Wood Tablets Tablets (du) provide a wider flat surface than slips (jian). The primary difference in definition is that tablets are usually wide enough to hold more than one line of vertical writing, whereas slips usually have only a single line of writing or at most two. Moreover, tablets are independent, whereas bamboo slips are tied together with cords to form a longer scroll. Bamboo is usually used for slips, but sometimes wood was used, presumably in imitation of bamboo. Wood is usually used for tablets (du), but sometimes wide sections of bamboo were used. Du were not standardized in size and the amount of text they might contain was variable; nevertheless, the amount of writing on tablets was limited by their size. Their function may have been similar to notepads. They could be used to record a short section of text and might even be scraped and reused, but they were not suitable for more ambitious compositions.6 The term du does not appear in oracle bone or bronze inscriptions. However, inventories of burial artifacts in Warring States tombs are written on both tablets and bamboo slips. Administrative and other documentations are also commonly found on du. To my knowledge, no literary compositions, such as the philosophical and historical manuscripts under consideration herein, have yet been found on du from the Warring States period. However, the Qin dynasty manuscripts recently acquired by Peking University include six wood and three bamboo tablets that may serve as examples of their usage in that period and thus possibly earlier. In each case, a single tablet includes a complete short text.One of the wood tablets measures 23 cm. x 4.7 cm. and has 165 characters written in eight vertical lines. Given the title, Taiyuan you sizhe 泰原有死者 (“A Dead Person on the Great Plain”) based upon its content, it records a short story, in which a corpse comes to life again. It is similar in style and length to the Zhiguai 志怪 stories in the later transmitted tradition (see Figure 3.4). Three other tablets—two wood and one bamboo—record songs (shi ge 詩歌). Another four tablets—two wood and two bamboo—are financial accounts.7 6. See Pian Yuqian 駢宇騫and Duan Shu’an 段書安, Ershi shiji chutu jianbo zongshu 二十世紀出土簡帛綜述 (Beijing: Wenwu, 2006), 32-146, for discussion of writing on tablets, slips, and silk cloth, and the relationship of writing materials and content. 7. Beijing Daxue Chutu Wenxian Yanjiusuo 北京大學出土文獻研究所, “Beijing Daxue cang Qin jiandu gaishu 北京大學藏秦簡牘概述,” Wenwu 2012.6: 65–73; Li Ling 李零, “Beida Qin du Taiyuan you sizhe jianjie 北大秦牘《泰源有死者》簡介,” Wenwu 2012.6: 81–84.

32

Figure 3.4. Wood tablet (du), Taiyuan you sizhe 泰原 有死者, Qin Dynasty, Beijing University collection.



Th e C h u - s c r i p t Ba m b o o - S l i p M a n u s c r i p t s

33

The other wood tablet, given the title Jiujiu shu 九九術 (“The technique of nine nines”), was probably used for memorizing the nine times nine tables. It provides direct evidence of the role of du vis-à-vis jian. A slightly different version is found within a bamboo-slip scroll in this same collection, assigned the title Suan shu 算書 (“Calculation document”). The Suan shu is divided into a Jia pian 甲篇 and a Yi pian 乙篇, each with distinctive content. The Jia pian is further divided into four parts, the second of which corresponds to the wood tablet Jiujiu shu. Moreover, a Qin-script wood tablet excavated from Well One (J 1) at Liye 里耶 in Hunan Province is identical to the Jiujiu shu in the Peking University collection.8 This suggests that the text circulated independently in the tablet form and was also placed within a longer bamboo-slip manuscript. Here, we should note that many of the Liye manuscripts are primarily administrative documents, recorded on tablets rather than on bamboo-slip scrolls. This presumably reflects the independent function of each tablet. Silk Scrolls The impetus for the formation of the long pre-Qin texts transmitted since the Han Dynasty was the use of long silk scrolls for copying more than one short manuscript. As I discussed in the previous chapter, the“reconstruction” of ancient texts in the Han Dynasty included copying them in the“new”Han script on long silk scrolls, and this was undoubtedly a major factor in the formation of long texts with stable contents (see Figure 3.5). Silk was woven in China from the Neolithic period, but because so little is preserved, we do not know when it began to be used as a writing medium. Although silk could be woven into long scrolls, rectangular pieces of silk were also used for writing. These rectangles were also used for drawings. For example, the so-called Chu silk manuscript from Zidanku 子彈庫, near Changsha 長沙 in modern Hunan province, which is roughly contemporaneous with Chuscript bamboo-slip manuscripts mentioned above, was only 47 cm. long and 38 cm. wide; it contained fewer than one thousand characters with figures drawn around the sides. A number of such silk rectangles were found in the early Han tombs excavated at Changsha Mawangdui 長沙馬王堆. The evidence for the Warring States period is more limited. Nevertheless, we may reasonably assume that long silk scrolls, as well as the shorter rectangular pieces of silk, were already in use in the Warring States period, if not so 8. Han Wei 韓巍.“Beida Qin jian zhong de suanxue wenxian 北大秦簡中的算學文 獻,” Wenwu 2012.6: 85–91 [87].

34

Buried Ideas

Figure 3.5. Silk manuscript, Laozi 老子 (jia 甲 version), Mawangdui Tomb One, Changsha, Hunan Province.



Th e C h u - s c r i p t Ba m b o o - S l i p M a n u s c r i p t s

35

commonly circulated as the bamboo-slip scrolls, and that there were some long texts, which were recorded on scrolls. The Development of Stable Texts In the Warring States period, bamboo-slip scrolls were the most common medium for circulating manuscripts. Bamboo grows prolifically and, even though the slips had to be prepared and bound together, it is unlikely that the prepared slips were ever particularly valuable as commodities. Nevertheless, because bamboo slips are bulky, only a limited number can be bound together in a single scroll. Silk scrolls were lighter and could contain much more writing. They would also have been more expensive. Whether intended or not, the effect of writing manuscripts on silk scrolls was to formally set the text, clearly defining its scope, wording, and the sequence of its parts. Bamboo-slip scrolls were relatively short, so long texts would need many scrolls. Moreover, it was (at least theoretically) possible to unbind them and make corrections by substituting slips and adding or removing sections. Thus, the use of silk was important in establishing conventional “texts”—written material that was repeatedly transmitted together in the same form, as opposed to collections of similar but discrete textual materials that might be changed or reordered. The ancient texts that reached a complete form earliest are notably compilations that had a naturally defined scope or sequence, most importantly, the Zhou yi 周易 (“Zhou changes”), with its sixty-four hexagrams, and the Chunqiu 春秋 (“Spring and Autumn Annals”), with its short sections placed in chronological order. By the late Spring and Autumn and Warring States periods, people had begun to collect texts of the same genre; for example, shi 詩 (“songs”), shu (“documents”), and li 禮 (“rites”). Disciples also wrote down the words of their masters (zi 子). These were probably circulated among a particular cohort and collected together. At first, such collections had no definitive order. The contents of the collections would have been varied and open-ended; that is, sections could be added or removed. One person’s collection of songs or of the words of Confucius was not necessarily the same as that of another person. As such collections began to be copied repeatedly in the same sequence and circulated, longer works were formed.9 Here it is important to remember that the intellectual world of the Han period in which there was centralized political authority was very 9. For example, Mozi’s collection of shu—or collections, since they were divided into Xia shu 夏書, Shang shu 商書, and Zhou shu 周書—overlapped with those of Confucius, but were probably not identical, as I will discuss in chapter 6.

36

Buried Ideas

different from that of the Warring States period. The pre-Qin texts that we know today were first circulated in a lively intellectual milieu which transcended any particular court. Indeed, the recent evidence of these Chuscript manuscripts suggests that literacy and scholarship in the Warring States period were much more widespread than previously thought. There were not only numerous competing thinkers with schools of students, but competing disciples who claimed the authority of their master with different ideas. Such a milieu was unlike that of the Han Dynasty, because, without a central court and court-appointed scholars, there was no means of establishing a standard version for any particular text. By the late Warring States period, longer texts made up of short sections had already begun to be compiled and repeatedly copied in the same form; that is, the idea of a “text” as opposed to a collection of similar materials, had taken form. However, it was primarily in the Han Dynasty, when texts were collected and copied onto silk scrolls in the “modern” (clerical) script that they took on the standard form of multi-chapter texts with a definite sequence and limited content that we know today. The setting of the imperial court was particularly important in this process because when written material was presented to the court, its form—that is, its scope, sequence, and particular language—gained a certain status. The “reconstruction” of lost texts (which often consisted of joining together collections of earlier, shorter ones and rewriting them in sequence in a standardized script) and the bureaucratization of scholarship in the Han Dynasty contributed to the formation of more stable texts, but the survival of large numbers of short manuscripts provided a rich resource for both alternative versions of texts and new ones. This model of textual development, which is similar to that offered by Mark Csikszentmihalyi for the Lun yu, assumes broader literacy and greater intellectual exchange than has conventionally been supposed and it is more complex than a simple accretional model. It also accords with the evidence of these new discoveries and explains some of the most perplexing aspects of the transmitted tradition.10 While this hypothesis for the development of the early Chinese textual tradition cannot be proven, it makes sense in terms of the three groups of Warring States manuscripts that I will describe below. It also 10. Mark Csikszentmihalyi, Material Virtue: Ethics and the Body in Early China (Leiden: Brill, 2004), 30–31. For an accretional model, see E. Bruce Brooks and A. Taeko Brooks, The Original Analects: Sayings of Confucius and His Followers (New York: Columbia University Press, 1998).



Th e C h u - s c r i p t Ba m b o o - S l i p M a n u s c r i p t s

37

helps in understanding the problems of textual authenticity that have plagued the Chinese textual tradition for the last two thousand years. Even after definitive versions of multi-chapter texts were recorded on silk and deposited in the imperial library, there were still short“old script” bamboo-slip and silk manuscripts in circulation and in the imperial collection. Since the contents of the newly established definitive texts were based upon earlier materials, the status of these ancient texts vis-à-vis the reconstructed—or, more properly, newly constructed—texts was uncertain. Moreover, the relative antiquity of material found on bamboo-slip manuscripts written in the old script in comparison with that recorded in the recently established “new script” versions of a text was impossible to determine with certainty. Now let us turn to the three groups of manuscripts and their sources. Guodian Tomb One Some ten thousand Chu tombs have been excavated, more than four thousand in Hubei Province alone, so there is a vast body of evidence concerning Chu burial practices and sumptuary rules. From this large body of evidence, archaeologists are able to place tombs within a context of historical development and to rank them according to the social status of the owner. The physical features of Guodian Tomb One and the artifacts found therein provide information about the identity of the occupant—presumably the owner of the manuscripts—and the date when they were buried, that is, the terminus ante quem of their possible composition. Except for the presence of philosophical texts written on bamboo, Guodian Tomb One conforms to a pattern found in other Chu tombs, and archaeologists agree that the occupant was a member of the lower aristocracy in the late-middle Warring States period.11 The tomb was an earthen “vertical shaft tomb,” which contained one wooden outer coffin or burial chamber (guo 槨) and one inner coffin (guan 棺) (see Figure 3.6). The opening of the pit was slightly larger than the base

11. Falkenhausen,“Social Ranking in Chu Tombs: The Mortuary Background of the Warring States Manuscript Finds,” Monumenta Serica 51 (2003): 442–43, which is reprised in chapter 7 of his Chinese Society in the Age of Confucius (1000–250 BC): The Archaeological Evidence (Los Angeles: Cotsen Institute of Archaeology at UCLA, 2006), provides an excellent synthesis of the various ranking schemes and historical development of Chu burial practices over the Spring and Autumn and Warring States period. Both works provide meticulous references for the Chinese sources and I have made liberal use of them in my discussion below.

38

Figure 3.6. Plan of Guodian Tomb One.



Th e C h u - s c r i p t Ba m b o o - S l i p M a n u s c r i p t s

39

and it was sealed with grey kaolin mud. A single sloping ramp entered the pit on the eastern side. The outer coffin was divided into head, side, and inner compartments. The corpse, placed in the inner coffin, rested on a carved wooden sleeping platform with its head to the east. These are all features that are distinctive characteristics of Chu tombs. The orientation is particularly noteworthy because Qin and Central Plains tombs are normally oriented to the north or west, so it is evidence that the tomb dates to the period before the Qin overthrow.12 The archaeologists dated the tomb fairly precisely to a period of around twenty years before the invasion of Qin and the movement of the Chu capital from Ying 郢 to Chen 陳 (now Huaiyang 淮陽, Henan Province) in 278 BCE, that is, to around 300 BCE.13 This dating was based on two primary factors: (1) The tomb and burial style are similar to those of 12. See Li Boqian,“A Brief Account of the Origin and Development of Chu Culture,” and Liu Zuxin, “An Overview of Tomb Number One at Jingmen Guodian,” in The Guodian Laozi: Proceedings of the International Conference, Dartmouth College, May, 1998, ed. Sarah Allan and Crispin Williams (Berkeley: Society for the Study of Early China and Institute of East Asian Studies, University of California, Berkeley, 2000), 11–12 and 30–31. For the characteristics of Chu tombs, see Guo Dewei 郭德維, Chuxi muzang yanjiu 楚系墓葬研究 (Hankou: Hubei Jiaoyu, 1995), 30; Xu Shaohua, “Chu Culture: An Archaeological Overview,” in Defining Chu: Image and Reality in Ancient China, ed. Constance Cook and John S. Major (Honolulu: University of Hawaii Press, 1999), 21–32; Alain Thote, “Continuities and Discontinuities: Chu Burials during the Eastern Zhou Period,” in Exploring China’s Past: New Discoveries and Studies in Archaeology and Art, ed. Roderick Whitfield and Wang Tao (London: Saffron Books, 1999), 189–204; and Alain Thote,“Burial Practices as Seen in Rulers’ Tombs of the Eastern Zhou Period: Patterns and Regional Traditions,” in Chinese Religion and Society, ed. John Lagerwey and Marc Kalinowski (Hong Kong: Chinese University Press, 2004), vol. 1, 65–107. 13. Archaeologists have agreed on the approximate date of the tomb. However, this dating has been disputed by the philosopher Wang Baoxuan 王葆玹 in a number of articles, and some scholars have followed his lead. For an English translation of one of Wang’s articles, see “A Discussion of the Composition Dates of the Various Guodian Chu Slip Texts and their Background,” in Contemporary Chinese Thought 32, no. 1 (Fall 2000): 18–42 (originally published in Zhongguo zhexue 20 [1999]: 366–89). See also his articles, “Guodian Chu jian de shidai ji qi yu Zisi xuepai de guanxi 郭店楚簡的時代及其與子思學派的關係,” in Guodian Chu jian guoji xueshu yantaohui lunwenji 郭店楚簡國際學術研討會論文集, ed. Wuhan Daxue Zhongguo Wenhua Yanjiuyuan 武漢大學中國文化研究院 (Wuhan: Hubei Renmin, 2000), 644–49; and “Zai lun Guodian zhushu zhi shidai ji qi wenhua beijing wenti 再論 郭店竹書之時代及其文化背景問題,” Kakuten Sōkan jukyō kenkyū 郭店楚簡儒教 研究 6 (Tokyo: Kyuko, 2003): 25–31. Wang rejects the dates of both the Baoshan tomb and Guodian Tomb One—and indeed the entire corpus of archaeological

40

Buried Ideas

other Chu burials—a style that changed substantially after the Qin capture of Ying. Considering the development of tomb structure, it falls toward the end, but does not belong to the latest period in the stylistic sequence of Chu tombs established by archaeologists, so it should be about twenty years before the Qin invasion; (2) Comparison of the artifacts with similar ones found in other Chu tombs for which dates have been established. For example, the Guodian tomb also includes a set of seventeen lacquer cups with square ears which are very similar in shape and decoration to ones found in Tomb 245 at Yutaishan 雨臺山 and Tomb One at Baoshan. These tombs are dated to the middle Warring States period and provide comparative evidence for the dating of the Guodian Tomb. Guodian Tomb One also includes artifacts that are similar to ones from Baoshan Tomb Two. These include a square lacquer-painted bronze mirror (a type of mirror found only in Chu) that is so similar to one from Baoshan Tomb Two that they might have been cast with the same mold. The divinations recorded on bamboo slips concerning the occupant’s last illness mentioned above suggest that the tomb was closed sometime between 323–316 BCE (see Fig. 3.3).14 This dating does not, of course, establish the date of the composition of the manuscripts; it simply tells us when they were entombed. Besides bamboo-slip manuscripts, Guodian Tomb One contained almost four hundred artifacts. That these are primarily objects that the literature on thousands of Chu tombs and calendrics. Arguing that Chu people did not strictly follow Qin law until 227 BCE, he proposes a date of 278–227 BCE for Guodian Tomb One. Unfortunately, this dating appears to be based upon a preconceived scheme of philosophical and textual development, rather than on an impartial analysis of the excavated materials. Since thousands of Chu and Qin tombs have been excavated, the entire construct of mortuary evidence would have to be overturned for his dates to be accepted and he has not offered any alternative analysis of the archaeological evidence. For summaries of the archaeological evidence, related material, and textual evidence that support a dating before 278 BCE, see Xu Shaohua 徐少華,“Guodian Yihao Chu mu niandai zhelun 郭店一號楚墓年代折論,” Jiang Han kaogu 2005.1: 68–72; Zhang Changping 張 昌平, “Zhanguo wanqi Chu mu de duandai yu Guodian M1 Chu mu de niandai yanjiu 戰國晚期楚墓的緞帶與郭店M1楚墓的年代研究,” in Shutsudo shiryō to kanji bunkaken 出土資料と漢子文化圈, ed. Yanaka Shin’ichi 谷中信一 (Tokyo: Kyūko Shoin 汲古書院, 2011), 31–49. 14. For the comparison between the tombs, see Li Boqian, “A Brief Account of the Origins and Development of Chu Culture,” and Liu Zuxin, “An Overview,” in The Guodian Laozi, 9–23, 24–32, and discussion on 118–20. See also Hubeisheng Jingsha Tielu Kaogudui 湖北省荊沙鐵路考古隊, ed., Baoshan Chu mu 包山楚墓 (Beijing: Wenwu, 1991).



Th e C h u - s c r i p t Ba m b o o - S l i p M a n u s c r i p t s

41

deceased might have used in his lifetime, or might use in daily life in another world, is a reflection of the changes in mortuary practices noted above. The inclusion of bamboo-slip philosophical manuscripts should also, I believe, be understood in this context. The only artifacts specially made for ritual use are: a bronze ewer, a bronze basin, a bronze eared-cup, a pottery ding 鼎 (food vessel), and a pottery he 盉 (pouring vessel).15 On the other hand, there were many objects of daily life, including seventeen lacquer eared-cups, a lacquer container, two lacquered qin 琴 (“zithers”), wooden combs, and various implements, including an iron link or chain (lian 鏈). Four wood figurines to serve the deceased after death were also included among the burial objects. The tomb also contained a number of weapons, including a bronze sword, a halberd, a pike, a bow, a quiver, arrows, and horse and chariot fittings. In the Warring States period, all members of the shi class received military training. In Chu tombs, these weapons served as indicators of male social status; they did not signify that the individual had any particular role in the military. The inclusion of two walking staffs in the tomb suggests that the occupant was elderly at the time of his death. The staffs were topped with doves or pigeons (jiu 鳩). These birds suck rather than sip liquids, so they were considered auspicious for the elderly, because they are vulnerable to choking. The Hou Han shu 後漢書 (“History of the Later Han Dynasty”) prescribes staffs with doves (jiu 鳩) on the handles as suitable for elderly men over seventy years of age.16 While the prescribed age may not date back to the Warring States period, we may reasonably assume that the presence of the staffs indicates that the deceased was an elderly man. Because the Guodian tomb was excavated after a robbery attempt, we cannot be absolutely certain that no slips were stolen. However, that seems unlikely for two reasons: (1) As Liu Zuxin 劉祖信, who excavated the tomb, has pointed out, the bamboo slips were on the other side of the coffin from the hole made by the tomb robbers and the water that had filled the tomb would have made the removal of anything more than a few small artifacts difficult; (2) The contents of the manuscripts do not have the gaps in continuity one would expect if slips were missing. 17 This 15. See Jingmenshi Bowuguan,“Jingmen Guodian yihao Chu mu,” Wenwu 1997.7: 48. 16. Hou Han shu 後漢書 (Beijing: Zhonghua, 1973), vol. 6, 3124; Allan and Williams, The Guodian Laozi, 123–24. 17. Allan and Williams, The Guodian Laozi, 117–18.

42

Buried Ideas

is evident from those texts for which there are corresponding versions in the transmitted tradition or from other excavations; for example, three of the scrolls in the tomb contain material now found in the Laozi Daodejing 老子道德經. In the Guodian manuscripts, the sequence of chapters is very different from the transmitted text and only about one-third of the chapters are included, but each chapter is continuous without missing slips. A version of the Zi yi 緇衣, which was transmitted as a chapter of the Li ji 禮記, was found in Guodian Tomb One and corresponds to one in the Shanghai Museum collection.18 Two more manuscripts have corresponding unearthed texts. One, Wu xing 五行 (“Five Conducts”), corresponds to a manuscript found in 1973 in Tomb Three at Mawangdui, near Changsha in Hunan Province. The other, Xing zi ming chu 性自命出 (“Human nature derives from the natural order”), corresponds to a manuscript in the Shanghai Museum collection, Xing qing lun 性情論 (“On human nature and natural endowment”).19 Again, although there are some differences between the Guodian manuscripts and other versions, there is no evidence of missing slips in the Guodian manuscripts. The Identity of the Tomb Occupant The identity of the tomb occupant is unknown, but one eared lacquer winecup has an inscription on its base that provides a clue. It is also the only artifact in the tomb, besides the bamboo strips, with an inscription. The excavation report transcribes the inscription as dong gong zhi bei 東宮 之杯 (“cup of the Eastern Palace”) (see Figures 3.6 and 3.8). The Eastern Palace was the residence of the crown prince (see Figures 3.7 and 3.8). A direct transcription of the Chu graph for the word bei 杯, “cup,” is: 不; that for the word shi 師, “teacher” or “master,” is: 帀. In Chu script, 不 and 帀 are written in a very similar manner and Li Xueqin has suggested that the transcription of the final graph of the inscription on the lacquer cup should be shi, meaning teacher. Thus, he interprets the inscription

18. Jingmenshi Bowuguan 荊門市博物館, ed., Guodian Chu mu zhujian 郭店楚墓 竹簡, (Beijing: Wenwu, 1998), 15–20, 127–38; Ma Chengyuan 馬承源, ed., Shang‑ hai Bowuguan cang Zhanguo Chu zhushu 上海博物館藏戰國楚竹書 (Shanghai: Shanghai Guji, 2001), vol. 1, 43–66, 169–214. 19. These texts have generated a vast literature. Recent studies in English include: Mark Csikszentmihalyi, Material Virtue; Dirk Meyer, Philosophy on Bamboo, 71– 176; Kenneth Holloway, The Quest for Ecstatic Morality in Early China (Oxford: Oxford University Press, 2013).



Th e C h u - s c r i p t Ba m b o o - S l i p M a n u s c r i p t s

Figure 3.7. Inscribed lacquer wine-cup from Guodian Tomb One.

Figure 3.8. Inscription on bottom of lacquer wine-cup from Guodian Tomb One.

43

44

Buried Ideas

as “teacher of the Eastern Palace,” and the deceased as the tutor of the crown prince.20 Li Ling 李零 has challenged Li Xueqin’s revised reading, supporting the original analysis of bei, “cup.”21 The inscription is written very casually and it is difficult to come to any firm conclusion. Von Falkenhausen, who agrees with Li Ling that the graph should be bei, has further challenged the identification of the deceased as a tutor to the crown prince on the grounds that, according to the sumptuary rules of the period, the tomb and its contents are inadequate for someone of this rank. He observes that the tutors of the Western Zhou period (taishi 太師 or taifu 太傅) were royal ministers and that, in the Zhou Li 周禮, the tutors of princes of the royal lineage had the rank of Middle Magnate (zhong daifu 中大夫). Different archaeologists have established systems for ranking Chu tombs that are not entirely identical, but they agree that Guodian Tomb One has the characteristics of a tomb for a member of the lower rank of the aristocracy, which would correspond to the shi 士 or “gentleman” class in the textual tradition.22 Thus, Von Falkenhausen argues, the occupant did not have the requisite rank to be a grand tutor of the crown prince.

20. This suggestion was first made by Li Xueqin at the 1998 Dartmouth conference, reported in The Guodian Laozi, ed. Allan and Williams, 123–24. See also Li Xueqin, “Jingmen Guodian Chu jian zhong de Zisizi 荊門郭店楚簡中的《子思子》,” Wenwu tiandi 1998.2, reprinted in Guodian Chu jian yanjiu 郭店楚簡研究, Zhong‑ guo zhexue 20 (1999): 75–80 and translated as “The Zisizi in the Jingmen Guodian Chu Slips,” Contemporary Chinese Thought 32, no. 2 (2000–01): 61–67; Li Xueqin, “Guanyu ‘donggong zhi shi’ de taolun 關於‘東宮之師’的討論,” Jianbo yanjiu 2001: 44–6. 21. Li Ling 李零, “Guodian Chu jian yanjiu zhong de liangge wenti—Meiguo Damusi Xueyuan Guodian Chu jian Laozi guoji xueshu taolunhui ganxiang 郭店楚 簡研究中的兩個問題—美國達慕思學院郭店楚簡《老子》國際學術討論會感想,” in Guodian Chu jian guoji yantaohui lunwenji ji 郭店楚簡國際研討會論文集, ed. Wuhan Daxue Zhongguo Wenhua Yanjiuyuan 武漢大學中國文化研究院 (Wuhan: Hubei Renmin, 2000), 47–52. 22. Lothar von Falkenhausen,“Social Ranking in Chu Tombs,”518, includes a comparative chart of different schemes, including his own. In his scheme, the tomb rank of Guodian Tomb One is designated as: N/O, that is, two ranks distinguished in the Spring and Autumn period that merged in the Warring States period. In the corresponding Chinese schemes, it is that of an “Upper Gentleman” or simply “Gentleman.” While Falkenhausen refrains from drawing correspondences between his tomb ranking and the traditional titles, they are implicit in his argument about the Guodian tomb.



Th e C h u - s c r i p t Ba m b o o - S l i p M a n u s c r i p t s

45

In my view, the graph on the bottom of the cup is more easily read as shi 師 than bei, but shi here does not necessarily mean teacher. Master craftsman were also called shi and, at least in later times, their names were sometimes recorded on the base of lacquer vessels. Moreover,“dong gong 東宫”found on the base of Han dynasty lacquer wine cups from Changsha Wangcheng Gushan 長沙望城谷山 has been identified as indicating the place of manufacture. Thus, it is probably the signature of the maker rather than a mark of ownership.23 Nevertheless, it would still be significant in identifying the owner of the manuscripts, because the Eastern Palace was that of the crown prince. A total of seventeen cups were found in Guodian Tomb One, but this is the only one that bears an inscription. This suggests that it had a particular significance for the occupant. The bamboo-slip manuscripts of philosophical texts are also not conventional mortuary items, but rather ones that must have been chosen for inclusion in the tomb because of their special significance for the occupant. However, not all bamboo-slip manuscripts found in Warring States and Han period tombs play this role. Some, such as tomb inventories, divination texts, and daybooks or almanacs (ri shu 日書), were presumably placed in tombs for ritual reasons or because they were deemed auspicious. However, others, such as administrative and legal documents sometimes found in the tombs of high officials, were personal items, used by the tomb occupant during their lifetime.24 The Guodian manuscripts belong to this latter category and thus reflect the identity of the deceased. From their content, we can see that the person was being identified as a thinker concerned with philosophical issues. The most obvious role for such a person was that of a teacher. And, even if he was not the Grand Tutor, at some

23. See Allan and Williams, The Guodian Laozi, 124. For the inscription of artisan names on Qin and Han lacquers, see Anthony J. Barbieri‑Low, Artisans in Early Imperial China (Seattle: University of Washington Press, 2007), 110–14; Nie Fei 聶 菲, Hunan Chu Han qimuqi yanjiu 湖南楚漢漆木器研究 (Changsha: Yue Lu shushe, 2013), 188–90. 24. Zhang Xiancheng 張顯成, Jianbo wenxianxue tonglun 簡帛文獻學通論 (Beijing: Zhonghua, 2003), 56–70; Pian Yuqian 駢宇騫, Jianbo wenxian gaishu 簡帛文獻概 述 (Taibei: Wanjuanlou, 2005), 241–499 classifies excavated manuscripts by type and summarizes their contents; Nishibayashi Shô­ichi and Chen Songchang, Xin Zhongguo chutu shuji, 53–65, provides succinct summaries of the Warring States period bamboo‑slip manuscripts found in the Chu region. Baoshan Tomb Two, which was used for dating Guodian Tomb One, includes divination and inventory slips, as well as administrative documents from the life of the deceased.

46

Buried Ideas

point in his life, he must have served the crown prince of Chu, for which he was rewarded with a lacquer cup from his palace. The Guodian Laozi and Textual History The editors of Guodian Chu mu zhujian divided the bamboo slips from Guodian Tomb One into sixteen texts.25 The three scrolls of bamboo slips with text that corresponds to material now found in the Laozi Daodejing are given the single title, Laozi, but are divided into three groups corresponding to the scrolls in which they were bound, Laozi Jia 甲 (A), Yi 乙 (B), and Bing 丙 (C). Altogether thirty chapters (zhang 章) are represented. Since the calligraphy reflects the hands of at least two scribes and the length and shape of the endings of the bamboo slips in each of the three groups differ, we may reasonably assume that the three scrolls were transmitted separately, though it is possible that two of the scrolls (A and B) were taken as related to one another. The relationship of the Guodian Laozi material to the transmitted Laozi Daodejing has been much debated. Some scholars take the chapters found therein as selections from an eighty-one chapter text similar to the transmitted one, but with some differences in wording; others as representing a stage before the formation of the current text.26 I take these three scrolls as important evidence in understanding the development of Chinese literature from small units of text to long multi-chapter ones, as discussed above. Within each scroll, there are several smaller sections that read continuously from slip to slip; that is, the writing begins at the top of one slip and continues on the top of another slip and so on. At the 25. Jingmenshi Bowuguan, ed., Guodian Chu mu zhujian. A revised version of the transcriptions found in the official publication, which includes twenty‑seven unplaced slip fragments, was published as: Wuhan Daxue Jianbo Yanjiu Zhongxin 武 漢大學 簡帛 研究中心 and Jingmenshi Bowuguan, Chudi chutu Zhanguo jiance heji 楚地出土戰國簡冊合集, vol. 1 (Beijing: Wenwu, 2011). 26. This idea was first discussed at the Dartmouth conference held soon after the initial publication of the Guodian manuscripts, see Allan and Williams, The Guo‑ dian Laozi, 142–46, and it has since been developed by some of the participants. See Rudolf G. Wagner,“The Impact of Conceptions of Rhetoric and Style upon the Formation of Early Laozi Editions: Evidence from Guodian, Mawangdui and the Wang Bi Laozi,” Transactions of the International Conference of Orientalists (Kokusai tōhō gakusha kaigi kiyō 國際東方學者會議紀要) 44 (1999): 32–56; William G. Boltz,“The Fourth‑Century B.C. Guodiann Manuscripts from Chuu and the Composition of the Laotzyy,” Journal of the American Oriental Society, 119, no. 4 (Oct.‑Dec. 1999): 590–608. See also Harold D. Roth, “Some Methodological Issues in the Study of the Guodian Laozi Parallels,” in The Guodian Laozi, ed. Allan and Williams, 71–88.



Th e C h u - s c r i p t Ba m b o o - S l i p M a n u s c r i p t s

47

end of the section, if room on the slip remains, it is left blank. Sometimes such units are also marked by punctuation. Each continuous section of text includes material that corresponds to one or more chapters found in the transmitted eighty-one chapter Daodejing, but in a different order. Some of the chapters are only partial and there are some differences in wording between the transmitted Daodejing and the Guodian bamboo slips, but the correspondence between the two at the level of individual chapters is clear. The sequence of the sections cannot be determined but, within the sections, the sequence of chapters is obvious—and it is usually radically different from that found in the current, eighty-one chapter, Daodejing. There is not even a division into a Daojing and Dejing; chapters from both are mixed together. For example, the Guodian Laozi A scroll, which includes the largest number of chapters, divides into five sections. The Daodejing chapter numbers that correspond to them are: (a) 19, 66, 46, 30, 15, 64, 37, 63, 2, 32; (b) 25, 5; (c) 16; (d) 64, 56, 57; and (e) 55, 44, 40, 9. This radical difference in the chapter sequence makes it difficult to understand how the bamboo-slip manuscript could have been copied from a long text, such as the current one, in which the chapters had a definite sequential order. It would be very odd to skip around in this manner when selecting the chapters, especially since there are no obvious connections in content within the sections. This suggests that the chapters originally circulated independently and were then collected together, but did not yet have a standard sequence. The repetition of identical passages within the same or different texts is another indication that long texts were formed by collecting smaller textual units that had circulated independently. Taken together, the A and B scrolls of the Guodian Laozi do not include repeated passages, but one passage in the A group (corresponding to lines 10–18 of chapter 64 of the received text) is also found in the C group of slips. Many scholars have argued on this basis that the Laozi A and B scrolls should be grouped together as a single manuscript and that Laozi C was a separate one. Since the Laozi A scroll and the Laozi C scroll were written by different hands, the conjecture that they were not considered parts of a single long text is probably correct. However, transmitted texts sometimes repeat the same passage in different places. For example, part of the third passage in the first chapter of the Mencius is also found in the seventh passage of that same chapter.27 These repetitions would be a natural consequence of the formation of long texts from a collection of 27. Mengzi yizhu 孟子譯注 (Beijing: Zhonghua, [1960] 2000), juan 1, 5, 17 (1.3, 1.7). See William G. Boltz,“The Composite Nature,” for further examples.

48

Buried Ideas

shorter ones on tablets and slip-scrolls gathered by different people. That is, if the same passage were incorporated in different manuscripts, when these manuscripts were collected together, it would occur twice unless it was edited out. This also explains why the same passages are sometimes found in different works. We may hypothesize, then, that these small sections consisting of several chapters of the Daodejing represent an intermediary stage in which the sequences within small groups of chapters were supplied by the person who wrote the chapters down on the slip-scrolls or by another person who dictated the material to him. These sequences in the Guodian Laozi scrolls left no trace in the later history of the text, as they were not incorporated into the eighty-one chapter version, which was the only one transmitted in later times. This situation is not unique. For example, as William Boltz and Rudolf Wagner observed early on, the two bamboo-slip manuscripts of the Zi yi, one from Guodian Tomb One and one from the Shanghai Museum collection, correspond to one another and to a section of the transmitted Li ji. The correspondence is at the level of short sections that are in a different sequence from the transmitted text, as in the case of the Guodian Laozi material and the Daodejing. Thus, they too may have circulated as independent units in an earlier period.28 The individual chapters of the Guodian Laozi may thus be regarded as the “building blocks” of the later Daodejing, to borrow the term used by Boltz and Wagner. As such, they would have first circulated independently, possibly through oral transmission. In the second stage in the development of the text, one or more chapters were probably written on tablets or bamboo slips. In the third stage, that of the Guodian Laozi, several such sections, constituting one or more chapters, were copied onto bamboo slips and tied together with cords to form a single scroll. This occurred more than once in different contexts, which accounts for the difference in the chapter sequences found in the Guodian Laozi scrolls and the transmitted Laozi, as well as the repetition of material. Further evidence for the independent circulation of the units may be found in the C scroll, which includes fourteen slips that are not found in the current Daodejing. In Guodian Chu mu zhujian 郭店楚墓竹简, they are 28. See William G. Boltz,“The Composite Nature,”56–57; Wagner,“The Importance of Context Structures on Paleography, Translation and Analysis,” 278–300. For a recent comparative study, see Wen Xing, “New Light on the Li ji 禮記: The Li ji and the Related Warring States Period Bamboo Manuscripts from Guodian and the Shanghai Museum Collection,” Early China 37 (2014), 519-550.



Th e C h u - s c r i p t Ba m b o o - S l i p M a n u s c r i p t s

49

separated out as a single manuscript and given the title, Tai Yi sheng shui 太一生水 (“The Great One Produced Water”). However, these fourteen slips divide naturally into two shorter sections: Slips 1–8 and Slips 9–14. Slips 1–8 are a coherently composed cosmology that begins with Tai Yi producing water. The rhetorical style of this cosmology is quite different from anything found in the current Daodejing and from the other six slips. These slips (9–14) make another coherent group, which begins with the line: Tian dao gui ruo 天道貴弱, (“The way of heaven is to value weakness”). This section is not found in the current Daodejing, but is similar in content and style to its chapters. I have argued elsewhere that they should be considered part of the Laozi material rather than grouped with Tai Yi sheng shui. If this is accepted, then it is evidence of Warring States period Laozi-type material that was not included in the transmitted text.29 The chapter sequence of the Mawangdui silk scrolls, recorded just before and after the foundation of the Han Dynasty, is essentially that of the transmitted Daodejing, except for the different order of the two parts, now called the Daojing and Dejing. Thus, we may surmise that there was a fourth stage, in which there were two texts that included Laozi material. Within each of these two, the passages had an established sequence, but their order with relation to each other was not firmly established until, finally, in the Han period, the form of the text we now have was established. Other Manuscripts from Guodian Tomb One Most of the manuscripts found in Guodian Tomb One were previously unknown and do not have a corresponding version in the transmitted tradition or among other excavated texts. In contrast to the Laozi material, four groups of aphorisms, given the single title, Yu cong 語叢 (“Thicket of sayings”), by the editors, are divided as four separate manuscripts. These aphorisms were possibly teaching materials, like the “thousand character classic,” memorized by later generations of students—further evidence that the occupant of the tomb was a teacher—but they are not joined in a sequence to be memorized as a single long text. Although Confucius is not mentioned as their source, some of the statements have corresponding material in the Analects. 29. See Sarah Allan, “The Great One, Water, and the Laozi: New Light from Guodian,” T’oung Pao 89, 4–5 (2003): 237–85. Cf. Dirk Meyer, Philosophy on Bamboo, 209–26, esp. 213, who argues that this section is a politico‑philosophical discussion that forms an integral part of Tai Yi sheng shui.

50

Buried Ideas

The other texts were also previously unknown. Within the context of the early Chinese textual tradition, they are not only philosophical in character, but quite abstract. Only one of these, Lu Mu Gong wen Zisi 魯 穆公問子思, names a particular philosopher or provides a geographical setting. In this manuscript, Duke Mu of Lu, Confucius’s own state, converses with Confucius’s grandson and Mencius’s alleged teacher, Zisi子 思 (483–402 BCE). Li Xueqin has proposed that five manuscripts might be grouped with Lu Mu Gong wen Zisi and understood as a single text, the Zisizi 子 思子, a lost text listed in the bibliographic chapter (Yiwen zhi 藝文志) of the Han shu 漢書. These five are: the Zi yi, Wu xing, and three previously unknown manuscripts, Cheng zhi wen zhi 成之聞之 (also called, Cheng zhi, “Bringing things to conclusion”), Zun de yi 尊德義 (“Honoring virtue and rightness”), and Liu de 六德 (“Six virtues”). This attribution has been very influential and much scholarship has been devoted to reassessing the role of Zisi, about whom little was previously known, in the development of early Confucianism.30 In contrast, Mark Csikszentmihalyi has argued that details of several historical figures have been conflated to form an evolving and unreliable biography of Zisi.31 In any case, the other texts in this group are stylistically diverse and do not mention Zisi (or any other contemporaneous person). Moreover, contrary to the pattern of textual development I have proposed herein, the theory that the manuscripts belong to a lost Zisizi presupposes that the basic form of texts in the Warring States period was long multi-chapter works. Without this presupposition, there would be no reason to link the texts together. The two manuscripts, Tang Yú zhi dao (“The Way of Tang [Yao] and Yú [Shun]”), which will be the subject of chapter 4, and Zhong xin zhi dao 忠 信之道 (“The Way of Fidelity and Trustworthiness”), are distinctive philosophically as well as materially. The calligraphic style of the two manuscripts is similar and the slips are also similar in length and in the shape 30. Li Xueqin, “The Confucian Texts from Guodian Tomb Number One: Their Date and Significance,”in The Guodian Laozi, ed. Allan and Williams, 107–11; see also 179– 83, 254–55. An example of a recent work that develops this thesis is Liang Tao 梁濤, Guodian zhujian yu Si‑Meng xuepai 郭店竹簡與思孟學派 (Beijing: Zhongguo Renmin Daxue, 2008). Among Western scholars, Kenneth W. Holloway, Guodian: The Newly Discovered Seeds of Chinese Religious and Political Philosophy (Oxford: Oxford University Press, 2009), does not take these manuscripts as the Zisizi, but he begins with the assumption that they are a philosophically coherent group of texts. 31. Mark Csikszentmihalyi, Material Virtue, “The Identification of the Wuxing and the Zisi Myth,” 86–103.



Th e C h u - s c r i p t Ba m b o o - S l i p M a n u s c r i p t s

51

of their endings. Li Xueqin suggested that these two manuscripts might have been composed by someone belonging to a “School of Vertical and Horizontal Strategists” (zongheng xuepai 縱橫學派). The existence of such a philosophical school has been challenged and this hypothesis has not been widely accepted. As I will discuss in the next chapter, Tang Yú zhi dao cannot be easily categorized using traditional philosophical categories and the argument for abdication as the ideal means of political succession found therein is not found in any transmitted text. The Shanghai Museum Collection The Chu-script bamboo-slip manuscripts now being published as Shang‑ hai Bowuguan cang Zhanguo Chu zhushu 上海博物館藏戰國楚竹書 were purchased in three batches, the first of which appeared in Hong Kong in the spring of 1994 and the last in the autumn of that same year.32 Zhang Guangyu 張光裕 of the Chinese University of Hong Kong, who assisted Ma Chengyuan 馬承源 in the purchasing of the bamboo slips for the Shanghai Museum, describes the slips when he first saw them as still fresh—soft and stuck together, like “wet noodles.”The bamboo slips in the first group were still caked in mud and dried quickly if they were not kept damp, so they were probably freshly dug up. The second and third bundles had been washed before the purchase.33 The illegal excavation, transportation, and treatment of the bamboo slips in the Shanghai Museum collection, including the washing of many of them, undoubtedly resulted in damage to the manuscripts and loss of slips. It is also clear that some slips or slip fragments were dispersed before the collection was acquired by the Shanghai Museum. For example, as we shall see in chapter 5, the Art Museum of the Chinese University of Hong Kong has a fragment that can be joined with a slip in the Shanghai Museum Zigao.34 However, we do not know how many slips have been lost. This has made the difficult problem of sequencing the slips in previously

32. Shanghai Bowuguan cang Zhangguo Chu zhushu, vol. 1, 1. See also Ma Chengyuan,“Zhushu Kongzi shi lun jian ji Shi de youguan ziliao 竹書《孔子詩論》兼及詩的 有關資料 (tiyao 提要),” in Xin chu jianbo yanjiu, ed. Ai Lan and Xing Wen, 1. 33. Personal conversation, 25 March 2007. 34. Chen Songchang 陳松長, Xianggang Zhongwen Daxue Wenwuguan cang jiandu 香港中文大學文物館藏簡牘, Slip 3, as cited in Shanghai Bowuguan cang Zhanguo Chu zhushu, vol. 2, 194. (I have not attempted to identify other fragments and slips that can be matched to those in the Shanghai Museum.)

52

Buried Ideas

Figure 3.9. Map of Chu cemetery region, near Ji’nan, Hubei Province, showing Guodian and Guojiagang.



Th e C h u - s c r i p t Ba m b o o - S l i p M a n u s c r i p t s

53

unknown texts particularly complicated. Nevertheless, the calligraphy on the slips in this collection, like that on those from Guodian Tomb One, is remarkably clear after conservation, and the slips are mostly intact. Guojiagang 郭家崗 Tomb One The Chu script of the Shanghai Museum bamboo-slip manuscripts is very similar to that of the Guodian corpus, and two texts are included in both corpuses—the Zi yi and Xing zi ming chu or Xingqinglun, as it is titled in the Shanghai Museum collection. There is no title written on the manuscript. The different titles for what is evidently the same text reflect the different naming traditions adopted by the editors of the two collections. Generally, if the manuscript does not have a title written on it, the titles of the Guodian manuscripts are based on the first few characters of the text; those of the Shanghai manuscripts, on their content. As Martin Kern has noted, that the manuscript versions of both texts are so similar lends weight to the supposition that the Shanghai manuscripts came from a tomb near Guodian.35 Ma Chengyuan conjectured that the slips purchased by the Shanghai Museum under his direction were from the same area as Guodian Tomb One, that is, the cemetery region on the outskirts of the ancient Chu capital at Ying.36 Looted tombs are usually discovered later on because of the evidence of digging left by the looters. Informal speculation that the Shanghai slips came from a woman’s tomb at Guojiagang, the neighboring village to Guodian, has been rife since its excavation in 1994 following a series of tomb robberies. Many scholars dismiss the possibility that it could be the source of the Shanghai Museum manuscripts because it is a woman’s tomb, but others argue that this is the only likely tomb. An online search of crime reports in the local newspapers yielded a certain amount of information about the robbery at Guojiagang. In Autumn 1993 and Spring 1994, there was a series of robberies, carried out by a gang of local tomb robbers, of three tombs in this village. The tombs were excavated by archaeologists in May 1994. Of the three tombs, two were so small that it is unlikely that they included more than a corpse and a few artifacts. However, the third tomb, now called Guojiagang Tomb One, contained the preserved corpse of a woman and many 35. Martin Kern,“The Odes in Excavated Manuscripts,”in Text and Ritual, 186, n. 25. 36. Shanghai Bowuguan cang Zhanguo Chu zhushu, vol. 1, 2.

54

Buried Ideas

textiles. The state of preservation of this tomb is reminiscent of Han Tomb Three at Mawangdui, near Changsha in Hunan Province, where the corpse of a woman was found with soft tissue still intact. Her many layers of clothing, quilts, lacquer grave goods, etc., were also well preserved. The corpse and textiles in Guojiagang Tomb One, in contrast, had deteriorated during the interval between the opening of the tomb and the excavation. Nevertheless, the corpse and textiles were well enough preserved at the time of excavation to surmise that they would have been similarly intact if they had not been exposed to the opening of the tomb.37 Thus, if bamboo-slip manuscripts had been buried in this tomb, they are likely to have been well preserved. The robbers—a group of local young men—were arrested after the robbery was finally reported and items from the robberies were found in their homes. In 1995, they were sentenced and three of the leaders were executed. The news reports of the trial include testimony about items from the tomb that were found in local houses and about the robbery itself, with vivid descriptions of the robbers’ discovery of the corpse, including an attempt by the leader of the gang to hoist it up, although no one else dared touch it. There is no mention of bamboo slips. However, the court reports concentrate on the artifacts that were discovered in the houses of villagers that served as the main evidence in the trial and were sufficient to convict the accused. If any artifacts had already been sold, they are not mentioned.38 There are several reasons why this tomb might be the source of the Shanghai Museum slips: (1) The first discovery of well-preserved

37. For the Mawangdui tomb, see Hunansheng Bowuguan 湖南省博物館, Chang‑ sha Mawangdui yihao Han mu 長沙馬王堆一號漢墓 (Beijing: Wenwu, 1973). For the Guojiagang corpse, see “Zhanguo nü shi canzao dushou ‘Zhongguo di yi gu shi an’ jiemi 戰國女屍慘遭毒手‘中國第一古屍案’解密,”http://big5.xinhuanet.com/ gate/big5/news.xinhuanet.com/legal/ 2005/11/02 (original source: Chutian jinbao 楚天金報), accessed Jan. 18, 2007. 38. Hubeisheng Jingmenshi Zhongji Renmin Fayuan 湖北省荊門市中級人民法 院 (1995), Fayuan anli ku 法院案例庫, http://big5.lawyee.com/Case/Case_Display. asp?RID=23979 (accessed Feb. 5, 2007). More than three tombs were excavated by the archaeologists, but the court accusations do not include the numbers assigned by the archaeologists. The court report lists photographs of Guojiagang Tombs One, Two, and Four and, since the accusations seem to involve only three tombs, One, Two, and Four are probably the ones that were robbed. A Tomb Three must have been excavated, but I have not found any report on it.



Th e C h u - s c r i p t Ba m b o o - S l i p M a n u s c r i p t s

55

bamboo-slip manuscripts from the Warring States period was from the nearby Guodian Tomb One. The excavation was prompted by a robbery attempt that was foiled and reported to the authorities. Thus, we know that tomb robbers were active in the area; (2) The date of the Guojiagang tomb robberies was shortly before the freshly dug-up bamboo slips appeared in Hong Kong; (3) Although tomb robbing is difficult to prevent even with draconian penalties, the remains of illegal excavations are difficult to disguise, especially in a densely populated heritage preservation area like this one. No other tombs robbed around this time that are likely to have been the source of the Shanghai Museum slips have come to light (or at least been reported); (4) The woman’s tomb (called Guojiagang Tomb One since its excavation) was extraordinarily well preserved. Bamboo slips are highly perishable, but the conditions were clearly sufficient for manuscripts to have been preserved. Female Literacy in the Warring States Period The primary reason that archaeologists have dismissed Guojiagang Tomb One as a likely source for the Chu-script manuscripts in the Shanghai Museum is that the tomb was that of a woman. However, we do not know much about literacy in general in the Warring States period and even less about female literacy. There is some archaeological evidence that at least limited literacy was not narrowly confined to males of the official class. Constance Cook has pointed to archaeological evidence for a “vast array of textual types, everything from inventory lists, covenants, administrative records, curses, prayers, sacrificial records, almanacs, and philosophical tracts to exercise manuals, divination manuals, [and] medical manuals” as evidence of at least widespread functional literacy.39 For example, in the fifth century BCE, thousands of lineage members in the state of Jin entered into covenants by depositing brush-written pledges on stone tables. As many as five thousand such tablets have been excavated at Houma, Shanxi Province, and another twelve thousand in Wenxian, Henan Province (not all bearing names). According to Crispin Williams, who estimates that

39. Constance A. Cook, “Education and the Way of the Former Kings,” in Writing and Literacy in Early China: Studies from the Columbia Early China Seminar, ed. Li Feng and David Prager Branner (Seattle: University of Washington Press, 2011), 334.

56

Buried Ideas

at least ten thousand names were recorded at Wenxian, “An individualized tablet was prepared for each covenantor, giving the covenantor’s name and the text of the particular covenant type. The number of covenantors participating in each covenant ranged from dozens to thousands.”40 Although it seems unlikely that so much of the population was even functionally literate, the personalization of the tablets including writing of individual names indicates that the idea of writing as a record had entered the popular imagination and that a number of these people would have been familiar at least with the characters in their own names. Moreover, a small number of the tablets include pledges about hanging up public notices. In these the person swears not to hang false notices and to report the posting of false notices to a certain lord.41 The display of false public notices only makes sense if literacy was sufficiently widespread for forgers to arise with the intention of spreading false information with written words; in other words, at least some common people must have had a level of basic literacy. There is little direct evidence of women writing or reading in this period. However, in ancient China, there were never prohibitions against women learning to read and write. There is also some indirect archaeological evidence for limited literacy among elite women. Tomb 621 from Jiudian 九店, in Jiangling, Hubei Province, includes some excavated Warring States period bamboo-slip manuscripts. Among them is a manuscript concerned with sacrificial food offerings. It is in very poor condition, as are most of 40. Crispin Williams, “Ten Thousand Names: Rank and Lineage Affiliation in the Wenxian Covenant Texts,” Asiatische Studien/Études Asiatiques 83, no. 4 (2009): 961, 968. See also, Susan Roosevelt Weld, Covenant in Jin’s Walled Cities: The Discoveries at Houma and Wenxian, PhD dissertation, Harvard University (1990), and“The Covenant Texts from Houma and Wenxian,” in New Sources of Chinese History, ed. Edward L. Shaughnessy (Berkeley: The Society for the Study of Early China and The Institute of East Asian Studies, University of California, Berkeley, 1997), 125–60; Crispin L. Williams, Interpreting the Wenxian Covenant Texts: Methodological Proce‑ dure and Selected Analysis, PhD dissertation, School of Oriental and African Studies, University of London (2004), 373–449. 41. Hao Benxing 郝本性 and Wei Kebin 魏克彬 (Crispin Williams), “Xuan shu jie 懸書解,” Kaoguxue yanjiu 考古學研究 6 (2006): 461–64. Crispin Williams, “Scribal Variation and the Meaning of the Houma and Wenxian Covenant Texts’ Imprecation Ma yi fei shi 麻夷非是,” Early China 37 (2014): 101–79, argues on the basis of the large number of participants (thousands) in the ceremonies that the tablets were probably written in advance using names supplied to the scribes by officials. Thus, the covenantors themselves would not have been literate, but those in charge of them would have been.



Th e C h u - s c r i p t Ba m b o o - S l i p M a n u s c r i p t s

57

the materials from this tomb, but the title (from the last four graphs of the manuscript) Jizi nü xun 季子女訓 (“Jizi’s Instruction for Women”) suggests that the intended audience was female.42 While it would be surprising to find that a woman in the Warring States period was so learned that her mourners defined her through her collection of manuscripts, it is within the realm of possibility. Elite women in the Han Dynasty were often educated and their education might go beyond literature specially written for women to include philosophy, history, and traditional classics. For example, according to the Shi ji, the Empress Dowager Dou 竇 (d. 135 BCE) was fond of the words of Huang Di 黃帝 and Laozi 老子 and she insisted that the emperor and crown prince read their works.43 According to the Han shu, in 74 BCE, when Emperor Zhao 昭 died and his teenage consort became empress dowager and nominal head of the government, the court advisors considered it suitable for her to study the classics and be tutored in the Shang shu.44 A number of elite women became accomplished scholars, among them Empress Xu 許 (died 8 BCE), who was “good at history and writing.”The most famous is, of course, Ban Zhao 班昭 (CE 45–116), the sister of Ban Gu 班固, who completed her elder brother’s history of the former Han Dynasty after his death. She also wrote the Nü jie 女誡 (“Precepts for Women”), a commentary to Liu Xiang’s 劉向 Lie nü zhuan 列女傳 (“Biographies of Exemplary Women”).45 As Lisa Raphals has pointed out, many of the “exemplary women” in the Lie nü zhuan were known for their rhetorical skills or “persuasions” and such skills implied learning.46 Whereas mortuary artifacts in Western Zhou tombs are largely indicators of social and official rank, by the Warring States period the artifacts buried with the dead reflected the breakdown of earlier social structures 42. Hubeisheng Wenwu Kaogu Yanjiusuo 湖北省文物考古研究所 and Beijing Daxue Zhongwenxi 北京大學中文系, Jiudian Chu jian 九店楚簡 (Beijing: Zhonghua, 2000), 142, 145. 43. Shi ji (Beijing: Zhonghua, 1973 [1959]), juan 49, 1975. 44. Han shu (Beijing: Zhonghua, 1979 [1962]), juan 75, 3155; Anne Behnke Kinney, Representations of Childhood and Youth in Early China (Stanford: Stanford University Press, 2004), 132–37. 45. See Michael Nylan,“Golden Spindles and Axes: Elite Women in the Achaemenid and Han Empires,” in Early China/Ancient Greece, ed. Steven Shankman and Stephen W. Durrant (Albany: State University of New York Press, 2000), 256–62. 46. Lisa Raphals, “Arguments by Women in Early Chinese Texts,” Nan nü 3, no. 2 (2001): 157–95 (163ff).

58

Buried Ideas

and often included personal belongings. My understanding of these Chu-script manuscripts is that they reflect the identity of the occupant as seen by his or her mourners. The mix of manuscripts in the Shanghai Museum collection does not have anything particularly “feminine” about it and there are no texts specially written for women. But, their contents are more eclectic than those of Guodian Tomb One and they do not have any obvious association with a particular occupation. There is also nothing in their contents that would preclude them from being the library of a learned female. And, indeed, if a Warring States period woman were to have had the elite level of education implicit in a collection of such manuscripts, this would very likely have been the most significant aspect of her life and character. It would not be surprising, therefore, if her manuscripts were buried with her. Unfortunately, however, because the manuscripts were looted, their source must remain a matter of speculation; this serves as a good example of the cultural loss inherent when artifacts are excavated illegally for the market. The Contents of the Manuscripts in the Shanghai Museum Collection There are about 1,200 bamboo slips in the Shanghai Museum collection, as opposed to the eight hundred from Guodian Tomb One. Publication is still on-going, but the slips published in the first nine volumes of Shanghai Bowuguan cang Zhanguo Chu zhushu have been assigned fifty-nine titles, with slip numbers ranging from three to sixty-five. At least five of these titles have two versions.47 The level of abstraction in the Guodian manuscripts is such that there is little evidence of place. The only manuscript from Guodian that mentions a particular state is Lu Mu Gong wen Zisi (“Duke Mu of Lu asks Zisi”), which is set in Confucius’s own state of Lu. Many of the manuscripts in the Shanghai Museum collection, on the other hand, are historical vignettes similar to materials found in such texts as the Guo yu 國語 (“Conversations of the States”), Zuo zhuan 左傳, and Zhanguo ce 戰國策. The contents are very cosmopolitan, including four manuscripts set in Lu 魯,48 four in Qi 47. Shanghai Bowuguan cang Zhanguo Chu zhushu, Tianzi jian zhou 天子建州, vol. 6, 125–46, 307–36; Zhengzi jia sang 鄭子家喪, vol. 7, 125–56, 169–88; Jun ren zhe he bi an zai 君人者何必安哉, vol. 7, 51–74, 189–218; Fan wu liu xing 凡物流形, vol. 7, 109–46, 219–300; Cheng Wang wei Chengpu zhi xing 成王為城濮之行, vol. 9, 15–28, 141–154. 48. Texts set in Lu include: Lu bang da Han 魯邦大漢, Shanghai Bowuguan cang Zhanguo Chu zhushu, vol. 2, 49–56, 201–10; Cao Mo zhi chen 曹沫之陳, vol. 4, 89– 156, 239–85; Ji Gengzi wen yu Kongzi 季 庚子問於孔子, vol. 5, 41–65, 193–236; Kongzi jian Ji Huanzi 孔子見季桓子, vol. 6, 31–59, 193–236.



Th e C h u - s c r i p t Ba m b o o - S l i p M a n u s c r i p t s

59

齊,49 one in Jin 晉,50 one in Zheng 鄭,51 and one in Wu 吳.52 However, the largest number by far is from Chu (thirteen).53 These record conversations of Chu kings from Zhuang Wang 莊王 (613–591 BCE) to Hui Wang 惠王 (488–432 BCE). Taken together, they appear to create a legendary past for the Chu kings in the tradition of the founding kings of the dynasties centered in the Central Plains. Although these historical vignettes may have been collected because of a perceived relationship in their contents, there is no evidence of either a material or a literary relationship; that is, that they ever constituted a single “text” in the sense that they were repeatedly copied and transmitted as a single work. Nevertheless, that a tomb might include numerous manuscripts of this type is evidence of the existence of many such short manuscripts available for collection and incorporation in texts such as the Guo yu, Zuo zhuan, Zhanguo ce, and other works. The other materials are more miscellaneous, but Confucius has a strong personal presence, appearing in conversation with both disciples and rulers.54 The Shanghai Museum manuscripts were the first excavated 49. Texts set in Qi include: Shanghai Bowuguan cang Zhanguo Chu zhushu, Jing jian nei zhi 兢建内之, vol. 5, 15–27, 163–77; Bao Shuya yu Xi Peng zhi jian 鮑叔牙與隰朋 之諫, vol. 5, 29–39, 179–91; Jing Gong nüe 競公瘧, vol. 6, 15–30, 157–91; Shi Liu wen yu Fuzi 史蒥問於夫子, vol. 9, 213–26, 269–88. 50. Text set in Jin: Shanghai Bowuguan cang Zhanguo Chu zhushu, Gucheng jiafu 姑成 家父, vol. 5, 67–78, 237–49. 51. Text set in Zheng: Shanghai Bowuguan cang Zhanguo Chu zhushu, Zhengzi jia sang 鄭子家喪 (jia and yi), vol. 5, 67–78, 237–49. 52. Text set in Wu: Shanghai Bowuguan cang Zhanguo Chu zhushu, Wu ming 吳命, vol. 7, 133–44, 301–25. 53. Texts set in Chu include: Shanghai Bowuguan cang Zhanguo Chu zhushu, Zhao Wang hui shi 昭王毁室, vol. 4, 30–42, 179–90; Jian Tai Wang bo han 柬大王泊旱, vol. 4, 43–67, 191–215; Zhuang Wang ji cheng, and Shen Gong chen Ling Wang 莊王既成 申公臣靈王, vol. 6, 61–74, 239–52; Ping Wang wen Zheng Shou 平王問鄭壽, vol. 6, 85–92, 255–72; Ping Wang yu Wangzi Mu 平王與王子木, vol. 6, 85–92, 265–72; Jun ren zhe he bi an zai 君人者何必安哉, vol. 7, 51–73; Ming 命, vol. 8, 55–68, 189–202; Wang ju 王居, vol. 8, 69–78, 203–14; Zhi shu nai yan 志書乃言, 79–88, 215–26; Cheng Wang wei Chengpu zhi xing 成王為城濮之行, vol. 9, 15–28, 141–54 (2 versions); Ling Wang sui Shen 靈王遂申, vol. 9, 29–36; 155–64; Chen Gong zhi bing 陳公治兵, vol. 9, 37–58, 165–88; Bang ren bu cheng 邦人不稱, vol. 9, 97–112, 237–68. 54. Shanghai Bowuguan cang Zhanguo Chu zhushu, Kongzi shi lun 孔子詩論, vol. 1, 13– 41, 119–68; Zi yi 緇衣, vol. 1, 43–68, 169–213; Min zhi fumu 民之父母, vol. 2, 15–30, 149–80; Zigao 子羔, vol. 2, 31–47, 181–99; 魯邦大旱 Lu bang da han, vol. 2, 49–56, 201–10; Xiang bang zhi dao 相邦之道, vol. 4, 83–88, 231–38; Ji Gengzi wen yu Kongzi 季 庚子問於孔子, vol. 5, 41–65, 193–235; Junzi wei li 君子為禮, vol. 5, 79–96, 251–61; Dizi wen 弟子問, vol. 5, 97–123, 265–83; Kongzi jian Ji Huanzi 孔子見季桓子, vol. 6, 31–59, 193–236; Zi dao e 子道餓, vol. 8, 13–20, 117–36; Yan Yuan wen yu Kongzi 顏淵問於孔 子, vol. 8, 21–36, 137–66; Shi Liu wen yu Fuzi 史蒥問於夫子, vol. 9, 213–226, 269–88.

60

Buried Ideas

texts to record the name of Confucius in the pre-Han period. Thus, when the Kongzi shi lun 孔子詩論 (“Confucius’s discourse on the Songs”) was released prior to publication in August 2000 at the “International Conference on Recently Discovered Chinese Manuscripts,” scholars were at first skeptical of the transcription, Kongzi 孔子 for the joined character written as: . Without supporting evidence, it might be read as 卜子; Bushang 卜商 was the personal name of Confucius’s disciple, Zixia 子夏, and the upper part of the graph looks more like 卜 than 乚. With the release of further materials in which Kongzi clearly referring to Confucius is written in this manner, including the Zigao, the transcription is no longer in doubt.55 Confucius also appears in these manuscripts with other appellations in various contexts, as Fuzi 夫子, Zi 子, and Zhong Ni 仲尼. This attention to Confucius contrasts with the Guodian manuscripts, which are considered by many scholars to be “Confucian” in their philosophical slant, but never mention the master. The use of Confucius’s name in the Shanghai Museum manuscripts does not necessarily mean that these manuscripts are “Confucian” in the sense that the ideas in them correspond to those found in transmitted texts associated with Confucius and his disciples. As we shall see in chapter 5, the Zigao purports to be a conversation between Confucius and his disciple, Zigao, but is quite different philosophically from the transmitted texts attributed to Confucius and his followers. The Tsinghua collection, on the other hand, includes manuscripts that correspond to, or are similar in form to, texts found in the five classics traditionally associated with Confucius, but there are no references to Confucius himself in the manuscripts published thus far. Another noteworthy aspect of the Shanghai Museum collection of manuscripts is that many manuscripts focus on songs, music, and performance. Especially important in this regard is Kongzi shi lun 孔子詩論, one of the texts attributed to Confucius, and Cai feng qu mu 采風曲目, but citations from and references to the shi are also found in other manuscripts.56 From the discussion of the shi in Kongzi shi lun, it is clear that a text similar to, if not entirely identical with, the transmitted Shi jing 詩經, already existed. This manuscript also illustrates how the discovery of these manuscripts 55. For a recent study of interchanges that support this reading, see Xuan Jiancong 禤健聰,“Shangbo Chu jian Kongzi shishuo 上博楚簡孔子試說,” Guwenzi yanjiu 29 (2012): 541–46. (The proceedings of this conference are published in Ai Lan and Xing Wen, eds., Xin chu jianbo yanjiu.) 56. See Martin Kern, “The Odes in Excavated Manuscripts,” 149–93, for discussion of shi in excavated texts.



Th e C h u - s c r i p t Ba m b o o - S l i p M a n u s c r i p t s

61

sheds light on the transmitted tradition of the last two thousand years in unexpected ways. The four sections of the Shi jing are traditionally called: “Guo feng 國風,”“Da ya 大雅,”“Xiao ya 小雅,” and “Song 頌,” but in Kongzi shi lun, they are “Bang feng 邦風,”“Da xia 大夏,”“Xiao xia 小夏,” and “Rong 容.”57 The“Bang feng”(“Airs of the regions”) presumably became the“Guo feng” (“Airs of the states”) because of the taboo against using the name of the Han founder, Liu Bang 劉邦, but this information was entirely lost to later generations. As a result, scholars have speculated for centuries about why the regional references,“Zhou nan 周南”and“Shao nan 邵南”(“South of Zhou” and “South of Shao”) would have been included under the rubric of “states.” Moreover, ya 雅, often translated as “elegantiae” following the meaning of the Chinese character, is written in the manuscript as xia 夏 and song 頌, as rong 容. These graphs may be understood as phonetic interchanges, but they cast doubt on the historical discussions of the significance of the rubric guo feng 國風. The collection also includes two previously unknown shi and four poems (or songs) written in poetic forms associated with the Chu ci 楚辭.58 Like the vignettes of the Chu court, the Chu ci–style poems suggest an association between the owner of the tomb and the state of Chu. If one accepts the traditional association of the Changes with Confucius, then the inclusion of a Zhou yi 周易 (“Zhou Changes”) could also be considered one of the manuscripts that relate to Confucius. However, the role of the Zhou yi, which also appears in the Mawangdui Tomb Three corpus, in tomb burials is not clear. It could have been understood as a philosophical text and included because it was a personal possession of the deceased, but it could also have been understood as analogous to other divination texts such as the Ri shu 日書, “day books,” which were commonly included in tombs. Each hexagram line (gua 卦) in the Shanghai Museum manuscript is written independently on one or two slips, so the sequence of hexagrams cannot be reconstructed. It has only thirty-four hexagrams but we know that it is incomplete because a fragment of one slip is in the collection of the Art Museum of the Chinese University of Hong Kong.59 Because some material was dispersed, it is impossible to 57. Shanghai Bowuguan cang Zhanguo Chu zhushu, vol. 1, 11–42, 119–68. 58. Yi shi 逸詩, Shanghai Bowuguan cang Zhanguo Chu zhushu, vol. 4, 23–30, 171–78; Li Song 李頌, vol. 8, 89–94, 227–46; Lan fu 蘭賦, vol. 8, 95–102, 247–68; Youhuang jiangqi 有皇將起, vol. 8, 103–10, 264–84; Liu piao 鶹鷅, vol. 8, 111–16; 285–91. 59. Shanghai Bowuguan cang Zhanguo Chu zhushu, vol. 3, 133.

62

Buried Ideas

judge whether the original was complete. There were also blank slips of the same shape and length, the function of which is unclear. Besides the Zhou yi and Zi yi, there are two other manuscripts that overlap with chapters of transmitted texts. These are possibly different versions of the same original manuscript. They are Wu Wang jian zuo 武 王踐阼 (“King Wu ascends the throne”), which corresponds to material in a chapter of the transmitted Da Dai Li ji 大戴禮記, and Min zhi fu mu 民之父母 (“Father and mother of the people”), which is found in the “Kongzi xian ju 孔子閒居” chapter of the Li ji 禮記.60 All the other texts were previously unknown and appear to be independent of one another. The Shanghai Museuam collection also includes a very fragmentary manuscript which includes discussion of the governments of Yao, Shun, and Yu, Ju zhi wang tianxia 舉治王天下.61 The editors divide it into five sections (wu pian 五篇). The first two, “Gu Gong jian Taigong Wang 古公見太公望” (“The Old Duke visits the Grand Duke Expected”) and “Wen Wang fang zhi yu Shangfu juzhi 文王訪於尚父舉治” (“King Wen enquires about raising and governing from Shangfu”) are set before the foundation of the Zhou Dynasty and concern the founding minister, Taigong Wang, also known as Shangfu. The other three are about the kingship of Yao, Shun, and Yu (“Yao wang tianxia 堯王天下,”“Shun wang tianxia 舜王天下,” and “Yu wang tianxia 禹王天下.’) This manuscript is fragmentary and was published too recently for a detailed analysis to be included herein, but each section is independent, so although the rulership of Yao, Shun, and Yu are described, the manuscript does not discuss how they transferred the rule. The Tsinghua University Collection The Tsinghua University collection of bamboo-slip manuscripts, now being published as Qinghua Daxue cang Zhanguo zhujian 清華大學藏戰國竹簡 with Li Xueqin 李學勤 as the principal editor, consists of about 2,500 slips, some of which are fragmentary and will be joined with other partial slips.62 The site from which the bamboo slips in the Tsinghua University collection 60. Ibid., vol. 7, 13–29, 147–68 and vol. 2, 15–30, 149–80. Matthias L. Richter, The Embodied Text, is a study of this manuscript. 61. Shanghai Bowuguan cang Zhanguo Chu zhushu, vol. 9, 59–76, 189–236. 62. Li Xueqin 李學勤, ed., Qinghua Daxue cang Zhanguo zhujian 清華大學藏戰國竹 簡 (Shanghai: Zhong Xi, 2010–).



Th e C h u - s c r i p t Ba m b o o - S l i p M a n u s c r i p t s

63

were taken is entirely unknown. The slips were for sale in Hong Kong from at least 2006 before they were acquired by Tsinghua University in July 2008, but when and where they were dug up remains a mystery.63 The first volume of the manuscripts was published in December 2010 and, as of May 2015, four further volumes have been published at yearly intervals. Although Li Xueqin cautiously designated them the “Tsinghua University bamboo slips”rather than the“Tsinghua University Chu-script bamboo slips” when they were first acquired, the script is similar to that of the Guodian and Shanghai Museum manuscripts. The calligraphy of the Tsinghua manuscripts is, generally speaking, more carefully executed, but the graphic forms are very similar. The content of some of the manuscripts, such as the Chu ju 楚居, which records the rulers of Chu and the movements of the Chu capital from ancient times on, also suggests a Chu origin. Radiocarbon (accelerated mass spectrometry) dating of a blank slip fragment produced a reading of 305 BCE+/- 30 years.64 This corresponds almost precisely to the date that local archaeologists determined for the Guodian tomb. Thus, the manuscripts were probably written on the slips in the same time period and are possibly from the same region. The bamboo slips were originally preserved in water-logged conditions. The Guodian and Shanghai Museum collections have undergone some deterioration after a drying technique used to preserve them. Thus, the Tsinghua slips are currently kept immersed in water while other preservation methods are investigated. Because the photographs were taken of the wet slips after separation and mounting on glass slides, they are even clearer than those in the other two groups. Qinghua Daxue cang Zhan‑ guo zhujian also includes photographs of the reverse sides of the slips. The reverse sides of the Tsinghua slips have yielded two surprises. One is that many slip-manuscripts have numbers that correspond to the correct sequence of the slips. This type of numbering had not been found previously in pre-Han manuscripts.65 The other is that many manuscripts have traces of a slanting line cut across the slips on the back side. These oblique 63. Zhang Guangyu 張光裕, “Qinghuajian de faxian ji qi yiyi 清華簡的發現及其意 義,”Dartmouth‑Tsinghua International Conference on the Tsinghua Bamboo‑slip Manu‑ scripts: The Fourth International Conference on Excavated Chinese Manuscripts, August 29, 2013. 64. Li Xueqin 李學勤,“Lun Qinghua jian Bao xun de jige wenti 論清華簡《保訓》 的幾個問題,” Wenwu 2009.6: 76. 65. Richter, The Embodied Text, 5. (Richter appears to have completed his book before this information became public).

64

Buried Ideas

Figure 3.10. Bamboo slips upon arrival at Tsinghua University, wrapped in plastic. lines are useful for lining up the slips in reconstructing their sequence, but they do not always correspond to the correct order. Sometimes slips appear to have been replaced, or they are otherwise out of order. Thus, they were probably not made after the manuscript was complete in order to enable the slip sequence to be reconstructed if the cords broke. More likely, they are related to the process used to cut the bamboo stalks in sections or strips before binding them together.66 66. Jia Lianxiang 賈連翔 (Qinghua Daxue Chutu Wenxian Yanjiu yu Baohu Zhongxin 清華大學出土文獻研究與保護中心), “Shi jiezhu shuzi jianmo fangfa fenxi Qinghua Daxue cang Zhanguo zhujian jianbei huahen xianxiang 試借助數字 建模方法分析《 清華大學藏戰國竹簡》簡背劃痕現象,” in Chutu wenxian yu Zhongguo gudai wenming guoji xueshu yantaohui 出土文獻與中國古代文明國 際學術研討會, Huiyi lunwenji 會議論文集, Beijing, June 17-18, 2013, 356–68. This type of slanted line cut across the reverse side of the slips in a manuscript has also been found on the Qin and Han slips now in the collection of Peking University; see Han Wei 韓巍, “Xi Han zhushu Laozi jianbei huahen de chubu fenxi 西漢竹書 老子簡背劃痕的初步分析,”in Beijing Daxue cang Xi Han zhushu 北京大學藏西漢竹 書, ed. Beijing Daxue Chutu Wenxian Yanjiusuo 北京大學出土文獻研究所 (Shanghai: Shanghai Guji, 2012), 227–335.



Th e C h u - s c r i p t Ba m b o o - S l i p M a n u s c r i p t s

Figure 3.11. Bamboo slips when first opened after arrival at Tsinghua University.

Figure 3.12. Bamboo slips (Shifa 筮法) with residue of fabric holder.

65

66

Buried Ideas

The scholars who produced the transcriptions of the manuscripts found at Guodian and in the Shanghai Museum collection paid relatively little attention to the reverse sides except for titles, probably because the numbering was not clearly visible and they worked from photographs that did not include the reverse sides. Their fragile condition after drying and the manner in which the slips are conserved and mounted now makes it difficult to examine them. However, faint numbers have been discovered on the reverse sides of at least two bamboo-slip manuscripts from Guodian Tomb One and it is possible that further examination of the manuscripts in both collections will reveal other evidence.67 Since only five volumes of the Tsinghua manuscripts have been published thus far, the full scope of their contents is not yet clear. Li Xueqin and his team have characterized the manuscripts as primarily historical in content and noted that many of the manuscripts in this collection are related to the traditional “classics” (jing 經). This concentration resembles the focus on philosophical writings in the Guodian tomb, as opposed to the more eclectic contents of the Shanghai Museum collection, possibly reflecting the official status of the deceased during his life. The association with the transmitted classics is a complex one as most of the manuscripts published thus far do not have a corresponding transmitted text. Thus, the similarity is primarily one of text type or genre. Most striking is the large number of shu (“documents”) or shu-like manuscripts in this collection. Thirteen such manuscripts have already been published.68 These include manuscripts that correspond to, partially correspond to, or are similar in style to the chapters of the Shang shu (“Ancient Documents”) and Yi Zhou shu 逸周書 (“Remaining Zhou Documents”). As I will discuss in chapter 7 with relation to the Bao xun 保訓, the discovery of these document-like manuscripts raises a broader question concerning the origin of shu and the significance of “documents” in the Warring States period. The Xinian 繫年 (“Chronicles”), with 138 slips, is the longest bamboo-slip manuscript yet found and published in a single volume (Two). 67. Guan Qiongmei官瓊梅 ,“Guodian Chu jian beimian xin faxian de ziji 郭店楚簡 背面新發現的字跡, Zhongguo wenwu bao 中國文物報, 2013-5-8: 8; Huang Jie 黃傑, “Xinjian youguan Guodian jian Zun de yi deng pian bianlian de zhongyao xinxi 新 見有關郭店簡《尊德義》等篇編聯的重要信息, http://www.bsm.org.cn/show_article.php?id=1857, accessed Jan. 1, 2014. 68. Qinghua Daxue cang Zhanguo zhujian, vol. 1: (Yin zhi 尹至, Yin gao 尹誥, Cheng wu 程寤, Bao xun 保訓, Zhou Wu Wang you ji Zhou Gong suo zi yi dai wang zhi zhi 周武王有疾周公所自以代王之志 (corresponds to Jin teng 金騰), Huang men 皇門, Zhai Gong zhi gu ming 祭公之顧命; vol. 3: Yue ming 説命: shang 上, zhong 中, xia 下; vol. 5: Hou fu 厚父,Feng Xu zhi ming 封許之命, and Ming xun 命訓.



Th e C h u - s c r i p t Ba m b o o - S l i p M a n u s c r i p t s

67

Figure 3.13. Bamboo slips preserved in trays of pure water. Prof. Liu Guozhong, Tsinghua University (left), and the author. The Xinian has no direct relationship to the transmitted classic, the Chunqiu 春秋 (Spring and Autumn Annals), but it does record chronological events—from the beginning of the Zhou Dynasty to the reign of King Dao 悼 of Chu (ca. 400 BCE). Some of the contents relate to material found in the Zuo zhuan and other transmitted historical texts. The manuscript is divided into twenty-three sections of continuous text, each marked at the end by the “tadpole” punctuation also found at the end of two groups of chapters in Guodian Laozi A. The sections of text have been arranged in chronological order and they have a certain logic—the first four sections are overviews of the formation and decline of the Western Zhou and the following sections are arranged by state. Nevertheless, each section can be read independently. According to the hypothesis outlined above, these would already have been collected together, either from long texts that are no longer extant or from many short ones, to form the present manuscript.69 69. Li Xueqin 李學勤,“You Qinghua jian Xinian lun Jinian de tili 由清華簡《繫年》 論紀年的體列,” in Li Xueqin, Chushi Qinghuajian 初識清華簡 (Shanghai: Zhongxi, 2013), 160–62, attributes the different literary forms to the different historical traditions of Chu (for the Xinian) and Wei (for the Jinian).

68

Buried Ideas

The Tsinghua collection also includes manuscripts that are similar in style to songs found in the Shi jing although they are not included in the transmitted text. Those published thus far are ones with historical references. They include: the Shi ye 耆夜 , which records a shi placed in the context of a feast given by King Wu of Zhou after a successful conquest;70 Zhou Gong zhi qinwu 周公之琴舞 (“The Dance of the Duke of Zhou with zither accompaniment”);71 and Rui Liangfu bi 芮良夫毖 (“The Bi of Rui Liangfu”),72 set at the end of the Western Zhou period. There are also a number of miscellaneous works which purport to record history, including, for example, the historiola-like Chi jiu zhi ji Tang zhi wu 赤鳩之集湯之屋 (“When Red Pigeons Gathered on Tang’s House”).73 Volume Four of this collection contains divination texts, including two manuscripts related to the Changes (Yi 易). One is the Shifa 筮 法 (“Divination Methods”), which discusses the principles and methods of divination, uses trigrams (but not hexagrams), and has titles similar to ones associated with the Guicang 歸藏, traditionally thought to be a Shang equivalent to the Zhou yi. The other is given the title Bie gua 別卦. It also includes a mathematical chart (suan biao 算表), which served the practical purpose of a calculator for multiplication. Arranged in matrix form, the entries where each row and column meet are the results of multiplying those numbers, and by combining multiplication with addition, any whole or half integer between 0.5 and 99.5 can be calculated. The Authenticity of the Shanghai and Tsinghua Bamboo-slip Manuscripts Major archaeological finds in China are inevitably followed by imitations, and Guodian Tomb One was no exception. Because the bamboo-slip manuscripts in the Shanghai Museum and Tsinghua University collections were looted and purchased after they had been sent to Hong Kong, questions about their authenticity have naturally arisen.74 However, the most 70. Qinghua Daxue cang Zhanguo zhujian, vol. 1, 10–13, 149–56. 71. Ibid., vol. 3, 8–11, 53–68. 72. Ibid., 12–15, 69–90. 73. I translate and discuss this manuscript,“When Red Pigeons Gathered on Tang’s House: A Warring States Period Tale of Shamanic Possession and Building Construction Set at the Turn of the Xia and Shang Dynasties,”Journal of the Royal Asiatic Society, 25, 3(2015): 419–438. 74. Hu Pingsheng 胡平生, “Jianbo bianwei tonglun 簡帛辨偽通論,” http://cccp.



Th e C h u - s c r i p t Ba m b o o - S l i p M a n u s c r i p t s

69

experienced Chinese specialists in paleography and excavated texts agree that both groups of manuscripts are authentic. There are many reasons to accept these manuscripts as genuine, other than the simple absence of telltale signs of fakery, such as mistakes in writing the Chu script, linguistic errors, etc.75 The reasons include: their physical condition, including a high degree of water saturation due to submersion in water over the centuries (“like wet noodles”) and their propensity to deterioration when not conserved properly, for example, ink fading when exposed to light, mold growth, etc.76 Another reason is that most of the manuscripts in both collections have no corresponding texts in the transmitted tradition or among other excavated materials. Fakes almost always imitate known materials because it is so difficult to create original works. Another is the consistent style of the Chu script, which was not well known before the discovery of these manuscripts. Here, it is noteworthy that the Shanghai Museum collection was acquired before the Guodian manuscripts had been researched or published, yet the graphic forms, many of which were previously unknown, are closely related to those in the Guodian manuscripts. Another, particularly clear evidence of their authenticity, is that many undeciphered characters in archaeologically excavated materials have been interpreted on the basis of the variants found in these unprovenanced manuscripts. Furthermore, the handwriting in each corpus is by a large number of scribes, so there would need to have been many skilled paleographers with calligraphic skills and a knowledge of Chu script involved. The study of these three groups of “Chu-script bamboo-slip manuscripts” has already become a recognized academic specialization, with thousands of articles, books, and academic theses published on every aspect of these manuscripts, including their paleography, grammar, phonology, and uchicago.edu/ archive/2008_IFBSD/Hu_Pingsheng_2008_IFBSD.pdf, reviews the faked bamboo‑slip manuscripts that have appeared since the discovery of Guodian Tomb One. He also lays out the standard criteria that paleographers use for determining fakes. 75. When faked manuscripts have been acquired by Chinese institutions, close inspection has generally revealed that they were not authentic, so they have seldom been published. An important exception is the “Zuo zhuan 左傳” acquired by Zhejiang University, see Xing Wen 邢文, “Zheda cangjian bianwei 浙大藏簡辨 偽,” Guangming ribao 光明日報, 2012-5-28. In this case, it is noteworthy that even though these manuscripts were published, specialists were dubious of their authenticity and they did not generate research articles. 76. Hu Pingsheng, “Jianbo bianwei tonglun” mentions that he knew of no fakes that had replicated the limp, water‑saturated conditions of genuine manuscripts. (His article was written in 2008).

70

Buried Ideas

their implications for understanding early Chinese history, philosophy, textual development, and so on. It is not possible that faked manuscripts could hold up such an extensive and varied academic edifice. Particularly significant is the very specific research on grammar, phonology, and the development of the script that has developed on the basis of these finds.77 As many scholars who have worked with these materials have remarked, it is impossible to imagine anyone with the range of knowledge of transmitted texts, paleographic skills, and creative imagination necessary to create them. Indeed, the most compelling reason for the acceptance of these manuscripts as authentic is perhaps the complexity of their interrelationships with a wide variety of early transmitted texts and inscribed materials, including bronze and oracle bone inscriptions, at all levels from the individual character to the development of thought. This complexity is particularly evident when attempting to decipher the texts graph by graph, word by word, sentence by sentence. As will become clear from the following chapters, these manuscripts often say things that were previously unimagined, but rather than overturning what we already know, they raise new questions and give us a new perspective from which to reexamine the transmitted tradition. Preparing the Official Publications The bamboo-slip manuscripts from the three groups have all been prepared for publication, including transcription of the Chu-script graphs into modern characters, by teams of scholars with paleographic expertise, as well as a profound knowledge of transmitted literature. The Chinese use the archaeological term, zhengli 整理, often translated as“processing,” for this procedure. Although there have been some differences in their procedures, they follow a general pattern. After the slips are cleaned, separated, and undergo initial conservation, they are photographed and the scholars work from these photographs. The black and white photography of the Guodian slips is very high quality and the best that could be done at the time. The later publications include color enlargements. 77. Some examples of detailed studies of the writing and grammar found in Chu manuscripts include: Li Mingxiao 李明曉, Zhanguo Chu jian yufa yanjiu 戰國楚簡 語法研究 (Wuhan: Wuhan Daxue, 2010); Wu Jianwei 吳建偉, Zhanguo Chu yinxi ji Chu wenzi goujian xitong yanjiu 戰國楚音系及楚文字構件系統研究 (Ji’nan: Qilu shushe, 2006); Xiao Yi 蕭毅, Chu jian wenzi yanjiu 楚簡文字研究 (Wuhan: Wuhan Daxue, 2010); Zhang Yujin 張玉金, Chutu Zhanguo wenxian xuci yanjiu 出土戰國文 獻虛詞研究 (Beijing: Renmin, 2006).



Th e C h u - s c r i p t Ba m b o o - S l i p M a n u s c r i p t s

71

More recently, infrared photographs have been taken to expose writing that is no longer visible to the naked eye.78 Although the bamboo slips were originally tied together with two or three cords and then rolled as scrolls, the bamboo slips considered herein were all disordered when the processing began, as noted above. Once the photography is complete, the next step is an initial sorting on the basis of material evidence into groups that are likely to constitute scrolls. The clues include the length of the slips, the shape of the slip-ends, the distance between the placement of the ties, calligraphic style, and handwriting. Content is another indicator, especially for smaller fragments. The teams of scholars responsible for preparing the manuscripts from Guodian Tomb One and those in the Shanghai Museum and Tsinghua University collections for publication include some of the most erudite Sinologists of our time and each publication is testimony to the profound learning they bring to the task. Nevertheless, the official publications cannot but be preliminary attempts at presenting the material in a coherent form, and, with the advent of the Internet, every publication immediately results in a flurry of online articles suggesting different readings, often including alternative sequences for the slips as well as different transcriptions for particular graphs. These online articles are soon followed by printed articles in specialist and other academic journals. The establishment of a sequence for the slips is fundamental for any interpretation. It depends on such criteria as logic, grammar, and the structure of the manuscript, as well as a profound knowledge of the language and content of contemporaneous texts. Because the reading of particular graphs may depend upon decisions about slip sequence and vice versa, their transcription requires an interpretive strategy based upon a sense of the text as a whole. In other words, the transcriptions and modern-character readings of the Chu graphs and the sequencing of the slips are integrally related. For the manuscripts that I will discuss herein, the sequences of the slips in three of the manuscripts—Tang Yú zhi dao, Zigao, and Rongchengshi 容成氏—have been much debated and those that I use in my editions differ from those in the official publications. The Bao xun, unlike many other manuscripts in the Tsinghua University collection, did not have numbering on the reverse side, but the published sequence has not been generally challenged. 78. Chen Wei 陳偉 et al., Chudi chutu Zhanguo jiance [shisi zhong] 楚地出土戰國簡 冊 [十四種] (Beijing: Jingji kexue, 2009) makes use of these in providing new transcriptions for the Guodian manuscripts.

72

Buried Ideas

Transmission of Texts Scholars have generally assumed that texts were either memorized and transmitted by oral recitation or written and copied. Although there are exceptions, such as the Laozi (老子), songs, and aphorisms, most of the manuscripts in these three corpuses do not have the mnemonic devices, such as rhyme, associated with orally transmitted texts. Indeed, they appear to reflect a literary culture in which ideas were exchanged in written form, as well as from teacher to student. This is true of all of the four manuscripts that I will discuss in the following pages. Indeed, I will argue that even shu, which claim to be speeches of the ancient kings, originated as literary compositions. Nevertheless, there are relatively few visual errors in these Chuscript manuscripts of the type that normally result from misreading a written manuscript; for example, writing a different graph that is similar in appearance to the intended one and missing graphs due to eye-skip. Yet, there are a great many unconventional graphic variants in these manuscripts that can be understood as ad hoc phonetic loans or as taking a sound-carrying element or“phonophore”to represent the word and adding a semantic element in an ad hoc manner. Moreover, the scribes sometimes write the same word with a different graph in the same manuscript. This suggests that the scribe was writing as he was listening to an orally delivered reading, rather than writing from memory or copying a written text. This impression of orality is particularly evident in the few examples where the same text appears in two manuscripts. For example, the Zi yi 緇 衣 (“Black Robes”), which was transmitted as a chapter of the Li ji 礼記 (“Record of Rites”), was excavated from Guodian Tomb One and another, very similar version is included in the Shanghai Museum collection.79 Thus, it provides an opportunity to compare scribal differences in writing the same words. As Dirk Meyer observes,“[I]n many instances, the manuscripts use different forms for presumably the same graph, here and there a signific is either added to the character or is left out entirely, or different graphs or words are chosen. Many of these differences conform to the criteria for phonetic similarity of loan characters and phonetic compounds in Old Chinese.”80 In order to explain this, Meyer draws an analogy with 79. Jingmenshi Bowuguan 荊門市博物館, ed., Guodian Chu mu zhujian 郭店楚墓 竹簡 (Beijing: Wenwu, 1998), 15–20, 127–38; Ma Chengyuan 馬承源, ed., Shanghai Bowuguan cang Zhanguo Chu zhushu 上海博物館藏戰國楚竹書 (Shanghai: Shanghai Guji, 2001), vol. 1, 43–66, 169–214. 80. Dirk Meyer, Philosophy on Bamboo: Text and the Production of Meaning in Early China (Leiden: Brill, 2012), 171.



Th e C h u - s c r i p t Ba m b o o - S l i p M a n u s c r i p t s

73

European practice in the Middle Ages and formulates the hypothesis that people read aloud to themselves rather than silently, thus “when producing a new copy, the scribe did not work from the graph he saw, but would write the sound he heard.”81 A simpler solution, which, I believe, better accounts for the degree of graphic variation based upon sound, is that texts were transmitted by oral dictation; that is, someone read a written manuscript aloud to one or more scribes who wrote it down on prepared bamboo slips. Because the scribes wrote what they heard as someone read the text out to them, there was a high level of graphic variation based upon phonophores, as well as graphic interchange of homophones or near-homophones. This process of transmission would also explain why the same word is sometimes written differently in the same manuscript; that is, the scribe who wrote quickly may not always have had time to consider that the same word was being used. Such dictation would have been a relatively efficient way of transmitting many copies of a text and it might have taken place in a workshop with professional scribes doing the recording. But, it might also have taken place when a teacher read a manuscript or his own composition out to his students to record. Or, groups of friends might have written down a manuscript read out by one of them. However, such dictation is not likely to have been the exclusive means of transmitting texts; I assume that manuscripts would also have been copied from other written manuscripts. Moreover, those texts that were traditionally transmitted by song or recitation would have first been recorded from their performances. Deciphering Chu Graphs Since modern Chinese characters are the descendants of the Qin script as revised in the Han Dynasty, excavated texts from the Qin and Han periods are relatively easy to read. However, the script found on these Warring States period manuscripts includes many graphs that are difficult to identify. Some of the unusual graphic forms are given in dictionaries and collections of “old script” graphic forms (such as the Shuowen jiezi 說文解 字, the Hanjian 汗簡, and Guwen sisheng yun 古文四聲韻), but others were entirely unknown before the advent of modern archaeology. Fortunately, other types of writing on bamboo slips, such as tomb inventories, divination texts, and so on, had already been found in some of the many Warring States period tombs excavated before 1993. The cache of bamboo-slip texts 81. Ibid., 150; see also, 172, 196–98, esp. n. 32, 332.

74

Buried Ideas

found at nearby Baoshan Tomb Two, which included well-preserved divinatory texts and administrative documents, as discussed above, was particularly important in establishing Chu graphic forms.82 These and other finds provide a foundation for deciphering the Chu script on these three groups of manuscripts. Since the discovery of these Chu-script manuscripts in recent years, collections of Chu graphic forms as well as specialist studies of the Chu script based upon them have begun to appear and constitute an invaluable resource.83 The problem of deciphering the Chu graphs in these manuscripts is above all a practical one—of determining how the brush-written graphs should be parsed and directly transcribed in the standard-script style (a process called liding 釐訂 in Chinese) and then read as modern characters in a manner that is convincing analytically and makes sense in context. More precisely, it is one of deciding on the word signified by each particular graph and the modern character that signifies that word. In analyzing Chinese characters, scholars have traditionally assumed that there is a standard character that represents a particular word, with other graphic forms that represent that word designated as variants (yiti zi 異體字) of the standard character or else as loan graphs (jiajie zi 假借字). This assumption is useful in tracing the development of Chinese writing through the lens of its later evolution. Moreover, the designation of a standard form as that which represents a particular word provides a necessary point of reference in any analysis. However, as all Sinologists know, classical dictionaries are full of characters that are glossed with one another and interchanged in the ancient texts. We may reasonably assume that such characters represented the same spoken word.

82. Constance A. Cook, Death in Ancient China: One Man’s Journey (Leiden: Brill, 2006), includes a translation of the divination and inventory slips from Baoshan Tomb Two and of the divination slips found in Wangshan 望山 Tomb One. 83. Important collections of Chu‑script graphs include: Li Shoukui 李守奎, ed., Chu Wenzi bian 楚文字編 (Shanghai: Huadong Shifan Daxue, 2003); Zhang Guangyu 張光裕 and Yuan Guohua 袁國華, eds., Guodian Chu jian yanjiu (Wenzi bian) 郭店楚簡研究·文字編 (Taibei: Yiwen, 1999); Rao Zongyi 饒宗頤 and Xu Zaiguo 徐在國, eds., Shangbo cang Zhanguo Chu zhushu zihui 上博藏戰國楚竹書 字匯 (Hefei: Anhui Daxue, 2012); Wu Jianwei 吳建偉, Zhanguo Chu yinxi ji Chu wenzi goujian xitong yanjiu 戰國楚寅襲及楚文字構件系統研究 (Ji’nan: Qi Lu, 2006). Qinghua cang Zhanguo zhujian also includes tables with images of the Chu graphs at the end of each volume. He Linyi 何琳儀, Zhanguo guwen zidian 戰國 古文字典 (Beijing: Zhonghua, 1998) is also a very useful source of Warring States period graphic forms.



Th e C h u - s c r i p t Ba m b o o - S l i p M a n u s c r i p t s

75

Warring States period Chu script, as found on these manuscripts, represents a stage in the development of early Chinese writing in which there was a high level of graphic multivalence. William Boltz argued early on that Chinese writing, as well as the Egyptian, Mesopotamian, and Mayan, had a pattern of development in three stages. At first, one graph signified one word. Then, rather than continuing to develop new graphs, in the second, transitional stage, existing graphs might stand for more than one word. This multivalence was usually, but not always, based upon sound. Finally, the ambiguity was resolved by the addition of determinatives drawn from the stock of graphic elements already in existence. This formulation is, of course, idealized—the stages cannot really be separated. However it does help in understanding the variation in the use of determinatives in these manuscripts; that is, determinatives were often added in a nonstandardized manner.84 As Imre Galambos has argued, it is more useful to think of the words in pre-Qin excavated manuscripts as having a “conventional” rather than a “standard” graphic form.85 Nevertheless, in order to understand and convey the meaning of the text of a Chu manuscript to modern readers, we must transcribe the Chu graphs in those modern characters that are the standard forms for signifying the words represented by the Chu graphs. In other words, we cannot simply rely on direct transcriptions of the Chu graphs, we must decide what Classical Chinese words we think these Chu graphs represent and transcribe the text in the appropriate modern characters in order to read and understand it appropriately. In my modern character citations of the Chu manuscripts analyzed herein, I underline those characters that are represented in the original manuscript by a different Chu graph. The original graph and/or a direct transcription may be found in the apparatus to the edition at the end of the chapter. When confronted with a graph that does not have a transmitted equivalent or does not make sense as a direct reading in its context, there are essentially four methods of decipherment. In my editions of these four manuscripts, they are indicated by the symbols, (1) ~ (alternate graphic form), (2) < (shared graphic element), (3) : (phonetic interchange), and (4) 〈〉 (writing error). These symbols are meant to roughly indicate 84. William G. Boltz, The Origin and Development of the Chinese Writing System (New Haven: American Oriental Society, 1994). 85. Imre Galambos, Orthography of Early Chinese Writing: Evidence from Newly Ex‑ cavated Manuscripts (Budapest: Department of East Asian Studies, Eötvös Loránd University, 2006).

76

Buried Ideas

the primary means used to decipher the graph. They are not intended as descriptive of the relationship or as an analytic tool. Moreover, the means by which graphs are deciphered are not mutually exclusive as more than one means may be used. For example, a part of a graph may be taken as a phonophore and then interpreted on the hypothesis of phonetic interchange. Moreover, some graphic forms are conventionally used in the Chu script but can also be regarded as examples of phonetic interchange. Thus, the symbol ~ is sometimes used as a catchall where the reading is based upon complex considerations. The first method of decipherment is to assume that the brush-written graph is an alternative graphic form (~). Thus it uses paleographic criteria. Although such graphs are generally classified as yiti zi (“variants”), these alternative forms may be the conventional ones in the Chu script. This method uses paleographic resources such as those indicated above to compare the bamboo-slip graph with guwen graphic forms found in the transmitted tradition and with the graphic forms found in other excavated materials, such as oracle bones, bronze inscriptions, and, especially, contemporaneous bamboo-slip manuscripts. The second method (