Statecraft and Classical Learning: The Rituals of Zhou in East Asian History 9789004177499

Citation preview

Statecraft and Classical Learning

Studies in the History of Chinese Texts Edited by

Martin Kern, Princeton University Li Ling, Peking University Ding Xiang Warner, Cornell University

VOLUME 1

Statecraft and Classical Learning The Rituals of Zhou in East Asian History

Edited by

Benjamin A. Elman Martin Kern

LEIDEN • BOSTON 2010

This volume received support from the “New Perspectives in Chinese Culture and Society” program, which is made possible by a grant from the Chiang Ching-kuo Foundation for International Scholarly Exchange to the American Council of Learned Societies. On the cover: Nie Chongyi 倞ⲯ佑 (10th century), Xinding Sanli tu 㕘⭂ᶱ䥖⚾ (AD 962). Song woodblock edition from 1175, repr. Shanghai: Shanghai guji chubanshe, 1985. This book is printed on acid-free paper. Library of Congress Cataloging-in-Publication Data Statecraft and classical learning : the Rituals of Zhou in East Asian history / edited by Benjamin A. Elman, Martin Kern. p. cm. — (Studies in the history of Chinese texts ; 1) Includes bibliographical references and index. ISBN 978-90-04-17749-9 (hardback : alk. paper) 1. Zhou li. 2. China—Politics and government. 3. East Asia—Politics and government. I. Elman, Benjamin A., 1946- II. Kern, Martin. III. Title: Rituals of Zhou in East Asian history. IV. Series. PL2468.Z7S73 2010 299’.51282—dc22 2009045591

ISSN 1877-9425 ISBN 978 90 04 17749 9 Copyright 2010 by Koninklijke Brill NV, Leiden, The Netherlands. Koninklijke Brill NV incorporates the imprints Brill, Hotei Publishing, IDC Publishers, Martinus Nijhoff Publishers and VSP. All rights reserved. No part of this publication may be reproduced, translated, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without prior written permission from the publisher. Authorization to photocopy items for internal or personal use is granted by Koninklijke Brill NV provided that the appropriate fees are paid directly to The Copyright Clearance Center, 222 Rosewood Drive, Suite 910, Danvers, MA 01923, USA. Fees are subject to change. printed in the netherlands

CONTENTS Introduction……………………………………………………………1 Benjamin A. Elman and Martin Kern I. Early China 1. The Zhouli as Constitutional Text…………………………………33 David Schaberg 2. Offices of Writing and Reading in the Rituals of Zhou……………64 Martin Kern 3. The Many Dukes of Zhou in Early Sources……………………….94 Michael Nylan 4. Centering the Realm: Wang Mang, the Zhouli, and Early Chinese Statecraft…………………………………………129 Michael Puett 5. Zheng Xuan’s Commentary on the Zhouli.………………………155 Andrew H. Plaks II. Medieval China 6. The Role of the Zhouli in Seventh- and Eighth-Century Civil Administrative Traditions…………………………………181 David McMullen 7. Wang Anshi and the Zhouli………………………………………229 Peter K. Bol 8. Tension and Balance: Changes of Constitutional Schemes in Southern Song Commentaries on the Rituals of Zhou…………. 252 Jaeyoon Song

vi

CONTENTS

III. Early Modern East Asia 9. Tokugawa Approaches to the Rituals of Zhou: The Late Mito School and “Feudalism”…………………………………………279 Kate Wildman Nakai 10. Yun Hyu and the Search for Dominance: A Seventeenth-Century Korean Reading of the Offices of Zhou and the Rituals of Zhou……………………..309 JaHyun Kim Haboush 11. The Story of a Chapter: Changing Views of the “Artificer’s Record” (“Kaogong ji” !"#) and the Zhouli……………………………………………………330 Benjamin A. Elman IV. Modern China 12. The Zhouli as the Late Qing Path to the Future…………………359 Rudolf G. Wagner 13. Denouement: Some Conclusions about the Zhouli.……………..388 Rudolf G. Wagner Bibliography…………………………………………………………397 Index…………………………………………………………………423

INTRODUCTION

Benjamin A. Elman and Martin Kern The Rituals of Zhou (Zhouli 周禮, hereafter Rituals) became one of the nine (later thirteen) Confucian Classics during the Tang dynasty (618–907). A late Warring States (480–221 BCE) or very possibly imperial Qin dynasty (221–206 BCE) text, it had reached its final form by the time of Zheng Xuan’s 鄭玄 (127–200) commentary. Written in early classical Chinese, the text is divided into six sections, the “offices” (guan 官) of Heaven, Earth, Spring, Summer, Autumn, and Winter, corresponding to the government Ministries of State, Education, Ritual, War, Justice, and Works, respectively. Its listings of 366 offices in these sections are arranged hierarchically, specifying the duties and staff members of each position and altogether presenting the structure of a six-part government with clearly divided responsibilities. While the first five sections of the received text are likely of pre-Han origin, the sixth, the “Artificer’s Record” (“Kaogong ji” 考工記), is a Western Han (202 BCE–9 CE) substitution for a purportedly lost earlier chapter (see chapter 11 in the present volume). The Rituals is an idealized blueprint for government organization that appears to include very ancient—indeed, Western Zhou (ca. 1046–771 BCE)—knowledge side by side with Warring States/early imperial political thought and government organization. In Han times (202 BCE–220 CE), the text was known as Zhouguan 周官, a title possibly better understood not as Offices of Zhou but as Comprehensive Offices, with the word zhou 周 used in its meaning of “comprehensive” rather than referring to the earlier dynasty (see chapter 1 in the present volume). The Rituals was special among the Classics in that the text at no point alludes to any particular historical frame of reference. The line that opens each of the six divisions of the text, wei wang jian guo 惟王建國, is most likely to be taken as a generic and timeless formulation that “it is the king who establishes the state” rather than “it was the (Zhou) king who established the state.” Moreover, it was apparently only with the appearance of

2

INTRODUCTION

Zheng Xuan’s commentary that scholars attributed the text to the sagely regent of the early Western Zhou, the Duke of Zhou (Zhou gong 周公; r. as regent 1042–1036 BCE). Like no other text among the Chinese Classics, the Rituals enjoyed a dual reception, first in imperial China and then across East Asia: on the one hand, it was cherished as a work of canonical learning; on the other hand, it was on several occasions applied to the actual administration of the state. The text is said to have been first employed for political ends by the Han minister Liu Xin 劉歆 (46 BCE–23 CE) in an effort to legitimize the short-lived Xin 新 dynasty (9–23) that Wang Mang 王莽 (45 BCE–23 CE) had established in a coup. While the full extent to which Wang Mang used the Rituals is difficult to pinpoint (see chapter 4), the text did become a blueprint for the structure of government when the Northern Wei dynasty 北魏 (386–534) designed their administration following the listings of offices in the Rituals and, in 398, constructed their new capital at Pingcheng 平城 (modern Datong 大同, Shanxi Province) according to the precepts of the “Artificer’s Record,” the sixth and final section of the Rituals. Between 491 and 495, in a series of somewhat symbolic moves, the royal ancestral temple and at least some of the palace halls, along with the official measures of length and capacity, were brought into accordance with the Rituals.1 Beginning in 546, the subsequent Western Wei 西魏 dynasty (534– 557) created its own government layout on the basis of the Rituals, with Yuwen Tai 宇文泰 (506–556), a military general who had become the highest official at court, being the driving force. In 556—a year before the official end of the dynasty—the central administration was reorganized according to the six ministries outlined in the classical text. In 557, following an act of usurpation, Yuwen Tai’s son Yuwen Jue 宇文覺 (542–557) established his new dynasty—now programmatically named the Northern Zhou 北周 (557–581)—and inherited the administrative system initiated by his father. In the process, he declared himself the new “Duke of Zhou” and elevated his father with the title of “King Wen” (Wen wang 文王) in imitation of the revered progenitor of the ancient Zhou, King Wen (who likewise had passed away shortly before his son founded the Zhou dynasty). 1 Pearce 2001. In addition to Pearce’s valuable article, our account here and below on the Northern Dynasties is informed by comments that Lu Yang (University of Kansas) presented during our final conference and later in writing.

INTRODUCTION

3

It must be noted that the Northern Wei 北魏 (386–533), Western Wei, and Northern Zhou were all non-Chinese dynasties, with their royal clans coming from the northern Xianbei 鮮卑 people (the Northern Wei was founded by the Tuoba 拓跋 clan; the Western Wei and Northern Zhou by the Yuwen 宇文). In contrast to their appropriation of the Rituals, outlined below, the Northern Zhou rulers highlighted Indian motifs in the Buddhist sculptures they sponsored in northwest China to augment their religious identities as a non-Han people.2 Yet their leaders also invoked the Rituals, and with it the founding heroes and purported institutional framework of the Zhou dynasty, to provide their non-Chinese dynasties, now ruling over key parts of northern China that in high antiquity had been the heartland of the Zhou dynasty, with classical precedent and Confucian legitimacy. Their attempts to put the idealized bureaucracy of the Rituals into actual political practice were embedded in a larger program of classical learning—complete with a royal academy (taixue 太學) and officially appointed “erudites” (boshi 博士)—which was focused on the ritual canons and strived to exploit the perennial ideal of “antiquity” (gu 古). It was perhaps because of the use of the Rituals first by the “usurper” Wang Mang and then by non-Chinese clans that neither the Sui (581–618) nor the Tang dynasties followed suit. Although classical scholarship on the Rituals continued (see chapter 6), the actual political use of the text declined. It was only during the Northern Song (960–1127) period that literati scholars such as Cheng Yi 程頤 (1033– 1107) argued once again that fathoming the Classics empowered rulers and officials to legislate according to the institutional model provided in the Offices of Zhou. Thereafter, Wang Anshi 王安石 (1021– 1086), as chief minister under Emperor Shenzong 神宗 (r. 1068– 1085), used the Rituals to authorize the ill-fated New Policies of 1069—undoubtedly the most infamous attempt throughout Chinese imperial history to put the text directly into governmental practice. Many later literati accused Wang Anshi of having politically exploited the Rituals in a manner that reminded them of its alleged earlier manipulations by Liu Xin and Wang Mang. As a result, Wang Anshi’s unprecedented statecraft repertoires were thought to have harmed the empire and impaired the clear fathoming of the Classics. For the state examinations of the Ming (1368–1644) and Qing (1644–1912) dynasties, official examiners frequently asked the candidates for the 2

See Dien 1962, 2–3, and Lu Haimo 2008.

4

INTRODUCTION

civil service to manifest the Rituals ideal in tangible ways or to critique its earlier misuses by Wang Mang, Wang Anshi, and others. In addition, the transmission of Chinese classical learning and statecraft to Japan, Korea, and Vietnam meant that the Rituals was also part of the repertoires of world-ordering techniques that East Asian kingdoms gleaned from the larger Chinese empire in their midst. The appropriations of the Rituals by Japanese, Koreans, and Vietnamese actually disputed the centrality of “China” and its heritage of “Confucianism” when non-Chinese such as the Mongols and Manchus militarily conquered and ruled China. During times of “barbarian” rule (by Chinese standards), the dragon throne in Beijing and the classical learning of Chinese officials were often gainsaid outside the empire, when first Korea and then Japan and Vietnam could legitimately claim they were the “second Rome” in East Asia. The present volume is the result of a series of three workshops, capped by a final conference in December 2006. When we began with our first workshop at UCLA in the fall of 2003, it was remarkable how little we knew about the Rituals as a classical text or its long-term historical role in the discourses of ideal government in premodern China, Japan, Korea, and Vietnam. Except for a few noteworthy studies that had focused on the Wang Mang era and Wang Anshi’s reform program, we knew little about much of its longer history of uses as a political paradigm for classical governance in China. We knew next to nothing about its significance in Japan, Korea, or Vietnam.3 During our meetings, we sought to present a more nuanced and balanced understanding of the Japanese, Korean, and Vietnamese appropriations of the Chinese statecraft repertoire enunciated in the Rituals. In reality, many disputed the centrality of “China” and its heritage of “Confucianism.” To clarify these conflicting views, we— to paraphrase John Duncan’s comments at the first workshop— explored the various polar tensions that inform social science and cultural history when approaching textual and historical sources: between the reductive theory that depicts texts as mere representations of social or economic reality and the equally reductive approach that sees all social or economic relations as culturally transparent, and between the urge to find unity of intellectual meaning in the Rituals and the drive to uncover only social, political, and economic differences in practice. 3

On China, see Bielenstein 1953–1959; Liu and Golas 1969. See also E. Biot 1851; Boltz 1993; Broman 1961.

INTRODUCTION

5

Our goal was to show how an apparently common Confucian statecraft discourse in Asia was appropriated differently in various regions and among various social groups as the consequence of divergent and historically contingent social, economic, and political circumstances. This volume, then, is a description of the diverse modes of appropriation over the centuries and across the East Asian region, which loosely define the Rituals of Zhou in light of the uses made of it. We do not attempt to ask, or answer, the essentializing question: What is the Rituals? Partly as a reflection of these multiple perspectives, we as editors do not think the volume requires a single, unified perspective in this “Introduction”; nor does it require a definitive “Conclusion” for such a deliberately wide-ranging conference volume. Instead, we want the richness of the Rituals to appear without conflating the different forms of its reception in China, Japan, Korea, and Vietnam. We realized at the final conference that we are not twelve actors in search of a single author who could encompass all our points of view. Taking into account the comments of our workshop and conference commentators, especially those of our consummate insider Willard Peterson, we as editors are, however, assuming a more authorial voice when in the following we are summarizing the key arguments from the collected articles. In this volume, every author attempts to identify how the Rituals entered into each era’s debates about premodern East Asian statecraft, which we define broadly as the concern for establishing sociopolitical, economic, and cultural order in the world. We also explore what those debates were about and how the Rituals related rhetorically to competing arguments from other texts concerning how such order could be best achieved. For all the authors, what unifies their distinct accounts of the various receptions of the Rituals across history and beyond China is how its codified vision of government offices was embraced and implemented, always with imperfect results, in a setting of political conflict requiring rhetorical debate. Thus, we also try to address scholars outside Asian studies who, like two of our outside commentators, Christopher Minkowski (Oxford) and Michael Cook (Princeton), are interested in ritual in India and Islam, in the history of law (especially the diffusion of Roman law but also the adaptations of Mosaic law), and in institutional history (Rome again but also Islam and India). The parallels we draw below to Rome are limited to the early period because of Rome’s precipitous fall when compared with the longevity of the Chinese empire, but the comparisons still have heuris-

6

INTRODUCTION

tic value. The medieval period from the Tang is better compared with the contemporary Abbasid Caliphate centered in Bagdad until the Mongol conquest of the Islamic world. Imperial China, like imperial Rome, represented the climax of ancient empire building. How did these new states of an unprecedented scale emerge at the end of the first millennium BCE? Their synchronicity as non-contiguous empires remains a historical puzzle, but they can be nonetheless compared in light of each other as empires of scale because each sought to unify huge landmasses and extraordinarily diverse populations. These new states were the first empires in the sense that we usually think of superstates of enormous size that became the dominant powers in their own world. Using our template based on the Rituals, we will note how imperial ideologies in China, although different from Rome’s, for example, enhanced their ability to extract resources via tribute from subject populations. Even if dissimilar in origins, the expansionist aims of these states and their agrarian economies, political cultures, and social organization were remarkably similar. In East Asia, contending states and armies constructed their political ideals and goals according to the models of past empires from an allegedly golden age. “What is an empire?” the Chinese asked themselves during the Warring States era, a time of turmoil when political unity seemed a far-off dream. Following the Qin unification, new rulers and subjects perceived Zhou ancient imperial culture as an ideal to be emulated. The Rituals represented that ideal for two millennia, a process that we trace in this volume. Looking to tradition, the Chinese tried to combine the earlier ideals of the Three Dynasties of Antiquity—Xia (not dated), Shang (ca. 1600–1046 BCE), and Western Zhou—with the kind of state power left to them by the discredited First Emperor of Qin (ruled as emperor 221–210 BCE). If empires were the ideal, then the bigger the better in both China and Rome. Size and scale were not just ideological ideals in the Rituals. Unlike rulers of more localized states in the Aegean who made a virtue out of decentralization, the Han Chinese and the Romans simply assumed that large empires brought more wealth, more power, more prestige, more of almost everything under their aegis, including colossal headaches when things went wrong. If any one characteristic set China and Rome as world empires apart from their predecessors, it was the scale of their operations and control mechanisms, although the scope of Alexander the Great’s empire provided a unique precedent in

INTRODUCTION

7

the Hellenistic world. Rome and the Han each became the single great power that dominated many different peoples and places. At roughly the same point in time, they were able to integrate more people and more productive, and destructive, force than any other human community before. A fascination with the past made earlier imperial models, particularly the early Zhou dynasty, compelling to the Chinese, a concern transmitted to their neighbors as well. Since the Han dynasty, they not only emulated the great Zhou that preceded them but also literally built their palaces right next to the previous or ancient capitals (not so much directly upon them as in Rome). They believed that the conquests of Han Wudi 漢武帝 (r. 141–87 BCE) would mark the climax of the ancient world. The Romans, on the other hand, although traditional, were less fascinated with earlier great powers in the Mediterranean. One obvious and important difference was that Rome’s predecessors were not Roman, while in China, they were Chinese. Thus, the Romans had to build their own new empire while the Chinese could claim to continue an earlier, if temporarily lost, cultural and political realm. Both empires created clear and indelible concepts of what it was to be an empire and to behave imperially. For China, that blueprint would often but not always be the Rituals, and occasionally also the “Grand Plan” (“Hong fan” 洪範) model of nested domains presented in the Classic of Documents (Shangshu 尚書). When the Han fell in the early third century and the Western Roman Empire disintegrated in the fifth, they both survived as model empires. While Rome replaced the Hellenistic model of Alexander in the Mediterranean, the Han overcame the legacy of the Qin in the name of restoring the ancient Zhou as the empire par excellence in East Asia, a claim that was arguably more important to the Eastern than to the Western Han. Successor states in the Mediterranean sought to become the “second Rome,” and after the Han dynasty fell, the Chinese continuously called themselves as a people and their language “the Han.” But the ideological underpinnings of empire in China and its institutional bulk were often traced back to Western Zhou times and were seen as having been articulated by the author(s) of the Rituals during the late Warring States and, possibly, Qin and Han times. David Schaberg and others contend in this volume that the Rituals originated in the Qin context after the immense power of oratory, scheming, and strategic thought had become widely recognized and, in some quarters, feared during the preceding Warring States era. Such

8

INTRODUCTION

oratorical claims can be somewhat compared to the discussions in fifth century BCE Athens, viz. Socrates and Plato versus the sophists. By then, the Rituals’ careful guidelines for channeling political speech for a unified realm were clearly preferred over the alternative—an uncontrolled flow of speakers and schemes among competing polities, which they could not have known prevailed in ancient Greece, for example. This was only one aspect of the rationale for the existence of the Rituals as a statecraft text, but the dangers of disruptive pluralism provided the historical context against which the idealization of a centralized state—as espoused in the Rituals—was written and cited during the early empire and thereafter. The political reading of the Rituals was not a Chinese monopoly, however, and the historical association of the Rituals with the Duke of Zhou, and hence with the early Zhou dynasty, during the early and middle empires of Han and Tang did not preclude its emulation beyond China’s borders. JaHyun Haboush (chapter 10) and Kate Nakai (chapter 9) contend that even though the Rituals had been officially part of the Confucian canon since Tang times, we cannot assume that it was drawn upon in identical ways throughout East Asia. Thus, Haboush and Nakai problematize the relationship between the distinctive nature of the Confucian canon and the cultural histories of the East Asian region, especially for those outside China who appropriated the Rituals for different purposes. What was distinct about the Confucian canon was that the Classics were simultaneously political and social, unlike the core text(s) in post-Roman Europe or the Middle East, which were primarily religious or sacral. Because the text was fundamentally open to appropriations beyond the Chinese political realm proper, one way to study the Rituals in its full historical complexity is to reimagine how the different states in East Asia, especially Korea and Japan, applied the canon to their uniquely different circumstances. Studying the various ways in which the Rituals was appropriated and reappropriated allows us to unpack the distinctive ways in which different East Asian societies imagined themselves as states at different moments in history. The later chapters in this volume take this perspective to heart and explore how the geopolitical spatial imaginaries of different societies and of later dynasties in China played out. When did a society conceive of itself as a state or as an empire? Since our goal is to think about the Rituals in the larger East Asian context, these issues force us to fully acknowledge the different ways in which the text was used in different East Asian coun-

INTRODUCTION

9

tries that did not share the Chinese conception of its own geopolitical spatial imaginary—and neither did the later imperial dynasties within China reproduce exactly the same statecraft policies. Why should we focus on the Rituals? Rather than another study of one of the Confucian Classics, the most persuasive answer we have given, and we hope our readers will share, is that this project is a step toward a better understanding of the statecraft terms in which premodern East Asian political thought, and with it the theory and practice of legitimacy, operated. Each of us seeks to clarify in our periods why the Rituals was a central, constitutionalizing text, especially but not only in the political-governmental sphere; and each of us has made it his or her task to thoroughly historicize the specific uses of the text. Based on the questions that came up during our meetings about relating actual historical rulers, courtiers, officials, and scholars with the Rituals, we have learned that we cannot just assume the text’s association with politics—for example, with Wang Mang’s troubled reign—without knowing the history of the text itself and its widely divergent uses. Hence we draw a distinction between the text itself and the uses made of it in different contexts within China proper by both Han Chinese such as Wang Mang and non-Chinese such as the northern Xianbei rulers of the Northern Zhou. As a group, the authors all recognized the need to render benign the tendency to polarize pragmatic versus rhetorical uses of the Rituals. The text, after all, was not merely an ideological smokescreen wielded by those who opportunistically appealed to ancient models and precedents. At the same time, the Rituals was endowed with great classical prestige since the Han dynasty (when it was probably completed), and it was rhetorically exploited for this reason. Since Tang times, it was classified as one of the Nine Confucian Classics because it described the ideal government established by the ancient sagekings. Soon thereafter, the Rituals was famously appropriated by Wang Anshi in his activist reforms called the New Policies (1070s– 1120s). For these, Wang interpreted the text as the ideal of a centralized bureaucratic system of government that prioritized interventionist strategies for social equity and welfare. Evidently, the Rituals did invite such vicarious co-option, since the time of Wang Mang, but such adaptations outside of China in Japan, Korea, or Vietnam were not always totalistic or based on the entire textual corpus as we have it now. While none of us has been so positivistic as to believe the Rituals was an actual blueprint for casting bronzes or creating political

10

INTRODUCTION

institutions, the legitimacy that the champions of the text tried to invoke encompassed political, social, and technical aspects of ideal statecraft. We have also discovered that, in addition to enunciating the statecraft uses of the Rituals, East Asian commentators over time found in it the ways in which the Zhou dynasty had bequeathed ancient civilizing processes through prescribed rituals and, furthermore, large-scale dynastic craft production that employed hundreds of artisans in courtsponsored mines and factories. Given these multiple perspectives and uses of the text, it has been our goal to be explicit in our diverse analyses of the statecraft, civilizational, or artisanal uses of the Rituals. In terms of statecraft, for example, we discussed at least three somewhat overlapping but also mutually exclusive views of the fundamental political differences enunciated in the text: Michael Nylan (chapter 3) stresses that, when seen as a Han dynasty composite text, the Rituals presents us with a statecraft discourse that sees states in light of “direct versus indirect” rule, not “central” versus “feudal.” Jaeyoon Song (chapter 8), on the other hand, maintains that the key division in Song times was that of a “centralized” versus “decentralized” administration. The usual concepts of “absolutist” versus “authoritarian” or “autocratic,” which have been used to understand the Rituals in the past, are problematized in this volume. We do not expect our authors or our readers to agree on which terminology might be best, but we think all of us are now explicit about why we use the terms we do. Overall, we think it worthwhile to try to show that the role of the “Classics” (jing, lit., “warp”—hence the image of guidance, coherence, and continuity) in premodern Chinese, Japanese, Korean, and Vietnamese politics was analogous to that of charters and constitutions in Western political culture. Premodern East Asian governments legitimated the constitutionality of their royal states (Korea and Vietnam) and empires (China and Japan) through classical political discourse. Political reform and classical debate went hand in hand. Imposing themselves on the interpretation of the Classics such as the Rituals allowed rulers and ministers to think they could limit the discourse regarding their own justifications for holding state power—even though the Classics were never actually controlled by those rulers and ministers. We must admit that, in the end, our volume has turned out to be more sinocentric than initially envisioned. Unfortunately, several participants in an earlier project entitled Rethinking Confucianism: Past

INTRODUCTION

11

and Present in China, Japan, Korea, and Vietnam (Los Angeles: UCLA Asian Pacific Monograph Series, 2002), including John Duncan, were unable to join us at the final conference. Martina Deuchler attended the conference and offered us some introductory notes about statecraft in premodern Korea, which inform our comments in this introduction. Meanwhile, we are all the more delighted that the two chapters on Japanese and Korean uses of Rituals statecraft help us to continue the larger intellectual project of rethinking the Chinese, Japanese, Korean, and Vietnamese uses of the Confucian Classics in East Asia. We hope that readers will sympathize with our efforts to be more ecumenical about East Asia and not just assume that the editors’ sinological predilections took easy precedence in this volume. The fact that classical scholars in Japan, Korea, and Vietnam could see themselves as representatives of “second Romes” when compared to China should caution us about a merely sinocentric view of the Rituals. When Korean yangban 兩班 (the two strata of literary and military elites) found in it the justification for the enslavement of criminals, and hence the raison d’être for slavery altogether, their way of appropriating the ancient text was as relevant as that of any Chinese official. Of course, the notion of a “second Rome” only takes us so far; after all, unlike Rome in 410 CE, the Chinese empire never really fell— continuing even after the onslaught of the Mongols in the thirteenth century and Manchu bannermen in the seventeenth. Moreover, the role of religion in the formation of imperial power and legitimation played out quite differently in the two contexts. After 410 CE, the Abbasid Arabian empire replaced the eastern and southern parts of the Byzantine (Eastern Roman) Empire; and by the eighth century, it represented a medieval Islamic polity as powerful as the Tang in China. However, the latter withstood the universalizing claims of religion via Buddhism and Islam that swept across the region after the rise of Arab power in Arabia and Central Asia. While certainly open to the flourishing of Buddhism and the influx of new religions, the Tang dynasty remained committed to its past political traditions and classical rituals in its quest for political legitimacy. In East Asia, it was particularly with the overthrow of the Ming dynasty by the Manchu Qing state in 1644 that Japanese, Korean, and Vietnamese elites felt increasingly encouraged to appropriate the Rituals for themselves. The Japanese in the eighteenth century claimed they were the true successors of the ancient Chinese sages.

12

INTRODUCTION

The Tokugawa regime, a sort of “centralized feudalism” from the seventeenth century on, refused to recognize the Qing as the “Central State” in East Asia. Annan (from a Chinese perspective literally the “Pacified South”) and Korea (similarly the “Pacified East”) “barbarized” the Qing and valorized themselves as the remaining land of the sage-kings; to scholars of these regions, the periphery had become the center. Hence, another perspective we pursue with the present volume is one of comparative classicism with regard to a text that was meaningful in different ways to different societies at different times. In imperial China, Japan, and Korea, classical and historical studies provided frameworks for the habits, interests, and values inherited by Confucian scholar-officials. Each classical text accumulated a history of its effects and interpretations, which then became a constituent part of the state’s raison d’être. As state ideology, the Classics—and particularly the ritual texts among them—represented the institutionalization of transhistorical truth. Although for the longest time the imperial state routinely tolerated, and indeed sponsored, competing interpretations of the Classics, steering clear of a narrowly defined orthodoxy, it remained deeply invested in their learning on the whole. Thus, at no point were the universalist truth claims advanced by Buddhism and Daoism able to replace the traditional Confucian canon. It was only in the aftermath of Song Dao Learning (daoxue 道學, i.e., what others today call “Neo-Confucianism”) that scholars and state authorities marked out clear and distinct formations of classical orthodoxy, leading the Yuan and Ming governments to successfully institutionalize selected commentaries on the Classics and Dynastic Histories that they hoped would help to consolidate state authority. Seeing themselves as late successors of the sage-kings and Confucius, who had derived from the Classics the guidelines for both political authority and dissent, scholars of the Ming and Qing dynasties comfortably stepped into their own roles as interpreters and transmitters of these precepts. The centrality of classical studies for the political discourse in premodern East Asia cannot be overemphasized. After rulers formally sanctioned classical learning during the early Han dynasty, political arguments were commonly expressed through the language of Classics such as the Rituals or through references to the Dynastic Histories. Scholar-statesmen, political opportunists, and even autocrats articulated their political views through the controlling medium of state ritual, classical sanction, and historical precedent. The ongoing connec-

INTRODUCTION

13

tion between the sages’ Classics of antiquity and later imperial Chinese political discourse (reactionary, moderate, or radical) suggests the power that some of these texts had over political behavior and expression in imperial China. Yet even in this general context of classical learning, we find the Rituals unique: like charters and constitutions in modern Western political culture, its model of statecraft legitimized the constitutionality of the imperial and royal states across East Asia. Among the Confucian Classics, the Rituals was exceptional due to its remarkable focus on statecraft to the exclusion of most other topics. Even when the larger set of the Thirteen Classics was replaced in relative importance by the more readable Four Books after the Song dynasties (960–1279), the Rituals retained its place as the supreme repository of statecraft models. Its repeated uses and abuses—the most notorious cases of the latter being those of Wang Mang in the early empire and of Wang Anshi in the Song—did not disqualify the text from remaining an integral part of the Classics. As such, and despite the late imperial emphasis on the Four Books, it belonged to the core curriculum for all those who would participate actively in the political arena. Throughout East Asia, century after century, the Classics—a set of abstruse and archaic texts that were translated into Japanese and Korean readings and glosses—preserved the orthodox teachings and political institutions of the sage-kings. In late imperial China, their mastery remained a key to advancement, fame, and power in the political arena. Classical erudition provided officials, scholars, and students in East Asia with a set of assumptions about good and evil in government and society; it also allowed them to manipulate the political machinery through historical references to the origins of those ideals on which the imperial state was founded. The Classics filled a simple need: if the ideals of the sage-kings were to be realized, the past—its ideas and institutions—had to be studied and cherished. While the Classics in general were called upon to provide the most acceptable justification for a dynasty’s rule, the Rituals at times “captured politics” in the name of restoring ancient governance and statecraft. The first three parts of this volume cover the remarkable historical ebb and flow of the Rituals’ place in classical discourse and imperial statecraft. With the fourth part, we turn to more modern issues in the mid–nineteenth century, when classical statecraft was increasingly challenged. Criticism accelerated after 1865, and a tug-of-war ensued among classical scholars over the proper evaluation of the Classics in

14

INTRODUCTION

general and the Rituals in particular. The source for the legitimation of political power remained located in the Classics; yet what it meant to be a classicist was called into question. The Classics were still inviolate, and they had always been reinterpreted according to the changing times. Now they were read and interpreted with new eyes and new strategies drawn from China’s increasing exposure to Western studies of state, society, and government. However, as the final chapters by Rudolf Wagner in this volume show, the Rituals remained a classical repository in which even modern reformers could locate native precedents for new styles of Western political institutions. Early China The chapters on the Rituals in early China are in different ways concerned with the nature of the text, its possible dating and authorship, its inner structure, its intellectual context, and its earliest interpretations. As with our volume altogether, we have chosen to allow each chapter to stand on its own and reach its own conclusions; to us as editors and contributors, the polyvocality of scholarly voices is not a problematic set of contradictions to be homogenized but a rich diversity of perspectives to be cherished and respected. In his opening chapter, David Schaberg argues the case for calling the Rituals a constitutional text in the sense of the Greek politeia, defined by Aristotle (384–322 BCE) as “the arrangement of magistracies in a state, especially of the highest of all,” and hence pointing to a monarchy with “certain internally justified arrangements for the distribution of power and responsibility” (Schaberg). Reviewing the evidence and considering the arguments advanced by traditional and contemporary scholars regarding the possible composition date of the Rituals, Schaberg sides with the modern scholar Jin Chunfeng 金春峰, who places the work in the Qin dynasty. In accepting Jin’s argument, Schaberg connects the Rituals to a string of late Warring States and early imperial texts that present similar constitutional schemes, including lists of offices, though never as comprehensively or coherently as the Rituals. From here, Schaberg considers how two central aspects of statecraft—the circulation of both goods and opinion—are represented in the Rituals, which thus can be read as “a comprehensive system for knowing and governing.” Most importantly, the textual organization

INTRODUCTION

15

of offices and duties presents a central government that “draws to itself material and informational resources and then reproduces its power as it redistributes these resources across the realm in the form of commands and public works.” As a system, the Rituals integrates cosmological thought with administrative practices, and its strong emphasis on unifying material resources and information connects well with pre-imperial and imperial Qin ideals of universal governance. In this bureaucratic order designed to control the all-encompassing flow of goods and knowledge to and from the emperor, the ruler—as in both the text of the Rituals and early imperial Huang-Lao theories of rulership—remains the invisible center detached from the management of daily affairs. From this analysis, Schaberg derives the interpretation that it was only in Han times, when the text was linked to the Duke of Zhou and the early Zhou dynasty, that the original list of “comprehensive offices” (Zhouguan) was gradually reframed as the “rituals of Zhou” (Zhouli), giving the text a new title and meaning that allowed its transformation from a Qin work of technical writing into, finally, one of the Classics that invoked the cultural and historical model of the ancient Zhou dynasty. Discussing the Rituals as a work of technical and formulaic writing in chapter 2, Martin Kern describes it as a textual artifact that in its account of governmental scribes “embodied the very nature and pervasive function of bureaucratic writing that it consistently asserted for its universe of governmental offices.” Nearly one thousand menial scribes (shi 史)—unranked commoners outside and below the Zhou aristocratic ranks—are listed as staff in virtually all offices of the central administration, and this concurs with information on scribes found in recently excavated manuscripts from the third and second centuries BCE as well as in several received Han dynasty texts. Yet in distinction to these low-level functionaries of writing, Kern traces a second group of scribes, a group of Zhou court officials of aristocratic rank who were also labeled shi, through the text of the Rituals. While forty-two offices include ranked scribal positions in charge of writing, reading, and storing texts on fiscal, legal, administrative, military, ritual, and astrological matters, the Ministry of Ritual contains five of the six central offices that are actually headed by scribes: the grand scribe (taishi 大史), minor scribe (xiaoshi 小史), scribe of interior affairs (neishi 內史), scribe of exterior affairs (waishi 外史), and scribe in royal attendance (yushi 御史). In the Rituals, these offices are among the highest positions of the royal court and

16

INTRODUCTION

concerned especially with the presentation of texts (as opposed to the menial production of documents)—just as they are in Western Zhou bronze inscriptions as well as in the early (possibly Western Zhou) layers of the Classic of Poetry (Shijing 詩經) and the Classic of Documents. On the other hand, Kern argues, transmitted Eastern Zhou texts such as the Zuo Tradition (Zuo zhuan 左傳), the Discourses of the States (Guoyu 國語), the Records of Ritual (Liji 禮記), and the Ceremonial Rituals (Yili 儀禮) show not only different scribal titles but also different functions associated with them that match neither the Western Zhou evidence nor the account of the Rituals. It therefore appears that in their representation of writing, the authors of the Rituals distinguished clearly between ancient and contemporaneous knowledge. As a result, “although the Rituals of Zhou was a late Warring States or early imperial artifact of writing, the memory of scribes enshrined in it was that of high antiquity; its authors, it appears, imagined themselves as the true successors of the ancient scribes they knew so well.” The text’s unified bureaucratic order was mapped onto the ideal of the Western Zhou only beginning in Han times, but its authors had already found themselves, or their forebears, in this past both remembered and imagined. In chapter 3 Michael Nylan explores the early legends surrounding the Duke of Zhou. As noted by Schaberg and Kern, it was the attribution of the Rituals to the duke during the Han that helped historicize and canonize the Rituals as purportedly depicting the early Western Zhou government. But who was the duke in historical memory and imagination? Examining his role in both the Classic of Documents and the Remnant Zhou Documents (Yi Zhoushu 逸周書), Nylan shows a meandering history of conflicting images from late Western Zhou through early Han times. The five “proclamations” (gao 誥) and related chapters from the Classic of Documents that have been traditionally attributed to the duke show him as a sage advisor to the young King Cheng 成 (r. 1042/35–1006 BCE), guiding and exhorting his ruler and, in fact, governing on his behalf. It was, however, the duke’s unique position of power that also made him appear, to no few Warring States thinkers, like a usurper of royal prerogatives. This lasting tension in the duke’s image, never fully suppressed in the Classic of Documents, prompted a forceful rebuttal in the Xunzi 荀子 and still left Sima Qian’s account in the Archivists’ Records (Shiji 史記), completed around 100 BCE, with a decidedly ambiguous tenor. Only in the wake of the new classicism

INTRODUCTION

17

emerging in late Western Han times, especially with the influential voices of Yang Xiong 揚雄 (53 BCE–18 CE) and Liu Xin, did thinkers wish “to cleanse the duke’s record of all the aspersions cast against him, so that he, who had once wielded all the power behind an imperial throne, might provide a more perfect model for contemporary aspirants to political power”—chief among them, of course, Wang Mang. Yet as Nylan shows in her account of the Remnant Zhou Documents, in Warring States times a third element was attached to the image of the duke, namely, that of a strategist versed in political expedients that were defined in terms both cosmological and pragmatic. His emphasis on harsh punishments, attention to administrative reforms, advocacy of war and agriculture, and overall efforts toward the centralization of power all show the duke not as a man driven by moral principle (as in the Classic of Documents) but as an adherent of the precepts of the Springs and Autumns Annals of Mr. Lü (Lüshi chunqiu 呂氏春秋) and a representative of Legalist thinking in the terms of the Qin reformer Shang Yang 商鞅 (ca. 385–338 BCE). Ironically, considering Schaberg’s conclusions on the Qin origins of the text, this Duke of Zhou would have been the perfect author of the Rituals—as opposed to the Duke of Zhou imagined by Zheng Xuan and other Han scholars who idealized him solely on the terms of the Classic of Documents. In chapter 4, Michael Puett discusses the Rituals’ unique usefulness to the “usurper” Wang Mang, who took the text as a blueprint for the ritual system, taxation policies, and administrative organization in his newly founded yet short-lived Xin (New) dynasty. Reviewing late Warring States and early imperial texts that advance theories of political organization and unified rule, Puett finds the Rituals’ closest parallels not in texts like the Records of Ritual that promoted the decentralized moral and ritual order of old but in Legalist works such as Shang Yang’s Writings of Lord Shang (Shangjun shu 商君書) and the Han Feizi 韓非子 as well as in the systematizing state cosmology of the Springs and Autumns Annals of Mr. Lü, all of which advocated a centralized state (and all of which were, in different ways, connected to the state of Qin). However, what fundamentally distinguished the Rituals from these texts were, in Puett’s analysis, two elements: the absence of any specific political or cosmological claims and the lack of a view of the ruler as a founding sage who replaces a previous order with his own, new creation of a state. While the ruler in the Rituals,

18

INTRODUCTION

like the sages in other writings, “establishes the state,” he does so in an act not of creation but of organizing and arranging what is already there: a vast bureaucracy of existing offices and officers who, once put into order, are left to themselves as a self-contained, smoothly operating system. Unlike the other early statecraft texts, the Rituals did not claim to replace the order of old with a sage’s new creation. Texts like the Writings of Lord Shang, Han Feizi, and the Springs and Autumns Annals of Mr. Lü, however, were profoundly antitraditional. They shared the Rituals’ centralizing ideology, but they also represented the strong-handed Qin (under the First Emperor 秦始皇帝; r. as emperor 221–210 BCE) and early Han (especially under Emperor Wu 漢武帝; r. 141–87 BCE) interventionist policies of new creation that by late Western Han times had become largely discredited and rejected under the influence of the new ritual and political classicism that claimed to return to the model of the Zhou dynasty. It was the Rituals’ lack of any explicit political argument, cosmological or historical, that made the text acceptable in the classicist environment of Wang Mang’s time, even though its actual administrative outlook was that of a centralized state headed by a powerful ruler. Ideologically, the Rituals—now firmly conceived as the Rituals of Zhou—seemed to advocate a return to antiquity; pragmatically, it helped a ruler with strong ambitions to create a new institutional framework. It was for these reasons, Puett concludes, that Wang Mang and his allies retrieved the Rituals from obscurity. While the first four chapters concern the nature, ideology, and early political use of the Rituals, the concluding chapter of our section on early China moves forward to Han times and to the defining moment in its scholarly interpretation and canonization. As Andrew Plaks points out in his report (chapter 5) on reading through the entire Zheng Xuan 鄭玄 (127–200) commentary on the Rituals, it was “Zheng’s fixing of the primary phonetic glosses and his basic elucidation of many of the central terms of ancient ritual lore that define his unparalleled contribution to the canonical tradition” and his status “as a synthesizer of the full range of Han textual learning and as the key transmitter of a common pool of received knowledge” to all subsequent scholarship on the Classics. In fact, we simply do not know what might have become of the Rituals without Zheng Xuan’s commentary, which not only built upon the glosses of earlier Eastern Han scholars but also brought his unrivaled erudition to the explanation—and presumably also editorial fixation—of the text. Most important, we think,

INTRODUCTION

19

is the fact that Zheng integrated the Rituals with some of the previously already established Classics, especially the Classic of Poetry and the additional ritual texts of the early canon, raising the hitherto still relatively obscure Rituals up to their status. In general, scholars acknowledge the central importance of Zheng Xuan’s commentary for the interpretation, canonization, and transmission of the Rituals, providing the basis for a monumental tradition of later imperial exegesis, but Plaks documents that his stature was not always unquestioned. While Tang and later commentators praised Zheng’s broad erudition and ability to draw on a very large range of sources for his glosses, others, especially in the Song, criticized him sharply for what they considered his promiscuous inclusion of dubious materials. On the other hand, as Plaks observes, it is not always clear which parts of the commentary are Zheng Xuan’s, which are earlier glosses he quotes from others, and why subsequent subcommentaries quote him with lines that are not found in his commentary proper. Zheng’s commentary on the Rituals—in distinction to his work on some of the other Classics—consists primarily of brief lexical or phonological glosses of individual words and the realia they refer to, ranging from all matters of ritual, especially sacrificial ritual, to those of law, economic control, market regulation, taxation, and disaster relief, and then further on to issues of clothing, hygiene, music, medicine, marriage, and other, more personal issues (that, however, are all within the scope of the overall social order). While Zheng assumes a critical historical attitude to other texts, especially the “Royal Institutions” (“Wangzhi” 王制) chapter in the Records of Ritual (which Plaks surveyed for comparison), he maintained “a supposedly unquestioning faith in the sagely origins of the ritual order enshrined in the Zhouli” (Plaks). This attitude by the foremost classicist of his time, perhaps even more than the specifics of his commentarial method, granted the text an authority that secured its status for later ages. Altogether, the five chapters on the Rituals in early China—that is, to the end of the Han—show a remarkable consistency in their historical evaluation of the text. While working from diverse perspectives and advancing discrete arguments, Schaberg, Kern, Nylan, and Puett all agree on the proximity of the text to a circumscribed political discourse of late Warring States and early imperial—especially Qin— times that combines cosmological thought with an ideal of a highly centralized and bureaucratically organized state. In their accumulation, these analyses detach the Rituals from its traditional context of the

20

INTRODUCTION

Confucian Classics and instead recontextualize the work within the discourse on statecraft surrounding the formation of the Chinese imperial state. Furthermore, both the strenuous attribution of the text to the Duke of Zhou and its clever appropriation by Wang Mang can be fully explained within the framework of Qin-Han historical and ideological developments. The same is true for Zheng Xuan’s commentary toward the very end of this period, which proved formative for both the imperial state and the canonization of the Classics. For different reasons and in different ways that all converge on the same historical moment, the inclusion of the Rituals into the canonical heritage depended on all three: the Duke of Zhou, Wang Mang, and Zheng Xuan. Medieval China Generalizing about the influence of the Rituals during the period between the fall of the Han and the rise of Sui and Tang (220–580) is difficult. Although the study of the ritual Classics played a significant role in shaping the institutional identity of the successive non-Han regimes in the north, described above, the Rituals in particular did not generate much enthusiasm among Han Chinese classicists. Often the ritual specialists who were masters of the canon of the Three Ritual Classics (sanli 三禮) actually were just experts of the Records of Ritual and the Ceremonial Rituals. They usually focused on the parts of these works dealing with mourning rituals (sangfu 喪服). For instance, Lu Zhui 陸倕, a personnel official of the Liang 梁 dynasty (501–557), was concerned about the lack of Rituals specialists among southern literati. In a letter to his chief minister, Xu Mian 徐勉, he lamented how the Rituals, which in his view was the key to the teaching of the sages, no longer received the attention it deserved. To remedy the situation, he recommended Shen Jun 沈峻, an expert on the Rituals from the Yangzi delta, to Xu Mian for the position of Erudite of the Five Classics.4 One important reason behind the decline of scholarly interest in the Rituals during the earlier half of the Southern Dynasties period was the growing influence of the more abstruse and metaphysical “Mystical Learning” (xuanxue 玄學) in the south, which shifted exegetical interest from liturgical texts like the Rituals to the Classic of Changes (Yijing 易經) and the Analects of Confucius (Lunyu 論語). In this 4

Nanshi 71.1741.

INTRODUCTION

21

connection, the popularity of Wang Bi’s 王弼 (226–249) commentaries in the south also undermined the Han commentarial tradition centered on Zheng Xuan. Many of the experts on the Rituals came from gentry families in the Yangzi delta, where the state of Wu’s 吳 traditions of Han classical learning had survived relatively intact. In addition to Shen Jun, mentioned above, exegetes of the Rituals, such as Gan Bao 干寶 of the Western Jin (265–316), Qi Gun 戚袞 of the Liang, and Shen Chong 沈重 of the Liang, were all natives of the Wu region, and their works clearly bore the imprint of Zheng Xuan’s exegetical oeuvre.5 The influence of the Rituals during this period survived somewhat in the political and institutional realms, however. The ritual and bureaucratic models presented in the text, as we have seen, continued to inspire northern “barbarian” regimes in the early medieval period to formulate their own ritual and bureaucratic systems. In the early days of the northern conquest regimes, exegesis of the Rituals was not much better than in the south. In fact, Fu Jian 苻堅, the ruler of the Former Qin 前秦 (351–395) state, arguably revived the study of the Rituals in the north by appointing the elderly Lady Song 宋氏, daughter of a renowned Rituals specialist, as an Erudite to pass on her family’s knowledge of the Rituals.6 It did not take long, however, before the Rituals was discovered by rulers of the Northern Dynasties as a valuable source for state building and cultural legitimization. The silence of the Rituals on the issue of “Chinese” (hua 華) versus “barbarians” (yi 夷) and its emphasis instead on an all-encompassing polity within the “Nine Provinces” (jiu zhou 九州) of ancient China also accommodated the needs of the politically ambitious but ethnically sensitive non-Han rulers of the north. On the other hand, even though the Northern Zhou adopted the Rituals, we cannot conclude that the use of the name of “Northern Zhou” was itself an effort to pay homage to the former “Zhou dynasty” of antiquity. In the opening essay of this second part of the volume, chapter 6, David McMullen explores the historical evolution of civil and military institutions during the mid–Tang period. He also explicates how the “Zhouli model” influenced the court and its officials in elaborating the theory and practice of the imperial state, although less so than during the Northern Wei and Northern Zhou, when outsiders used the Rituals 5 6

See Jian Boxian 1975, 76–79, 176–177. Jinshu 96.2521–2522.

22

INTRODUCTION

to disguise their alleged barbarization of the northern Chinese world. The Tang ritual code was the result of three forces: (1) the authority of the Confucian canon and its three ritual texts, the Ceremonial Rituals, the Records of Ritual, and the Rituals; (2) historical practices since the Confucian canon had been formed during the Han dynasty; and (3) the requirements of the Tang dynasty itself and its emperors, as they developed their own ambitions or responded to contemporary political situations. The Rituals maintained its status as a canonical text during the Tang, but the scholarly community took a cautious attitude to the picture it painted of the ideal state in antiquity. Although Tang statecraft and political thought overall did not favor efforts to revive the Rituals’ program literally, scholars continued to hold the text in high regard and frequently cited it in reviews of institutional history. Although the Rituals was never the exclusive authority for Tang rituals, it complemented other Confucian canonical texts to sanction state rituals that played an indispensable part in defining the dynastic house, the official hierarchy, and the society at large. The charter document for the Tang from the late Kaiyuan period (712–741) on, however, was not the Rituals but the Compendium of the Sixfold Administration of the Great Tang (Da Tang liudian 大唐六 典), a governmental compendium somewhat related to the Rituals. The integration of directives from the Rituals with those of other canonical sources, and with historical practices in state ritual, made it an integral, and at times the most influential, part of Tang statecraft. When the concerns of the Tang political elite changed significantly, the Rituals continued to play some role, but confidence in the state system was eroded by the An Lushan 安祿山 rebellion after 756. Citations of the Rituals in key sections of the Tang Criminal Code (Tanglü 唐律) suggest that its judicial principles also remained authoritative in legal matters. In the skeptical scholarly climate of the late eighth century, however, the Rituals lost authority. Some scholars such as Du You 杜佑 (735–812), and perhaps Liu Zhi 劉秩 (d. after 758) before him, questioned the simplistic correlation of the Kaiyuan Compendium with the six bureaucratic divisions of the Rituals. Picking up from McMullen’s chapter, Peter Bol contends in chapter 7 that there was a sea change in classical interpretations of the Rituals between the Tang and Song dynasties. Bol agrees that during the Han-Tang era, scholars focused more on the reading of the text as a historical document and less on the contemporary significance of the institutions described in the text. During the Northern Song dynasty,

INTRODUCTION

23

however, Wang Anshi offered a new reading of the Rituals, which illustrated and justified his view that there were universal principles for the organization of human affairs. Thereafter, succeeding literati considered Wang’s activist reading of the text to have furthered his own political agenda—which in their view came at the cost of a nuanced understanding of the classical vision—a prime example of the misuse of a Classic. However, Bol cautions us not to read this “failure” narrative backward into Wang’s own political agenda. Often the villain in later accounts of the fall of the Northern Song dynasty in 1127, Wang Anshi was a high-minded minister who cherished the Classics and believed he had discovered how to unleash their full political and economic potential for the benefit of the Song dynasty. In chapter 8, Jaeyoon Song explains for the first time why, in the aftermath of Wang Anshi’s reform failures, the Southern Song still produced a host of new commentaries on the Rituals. Song asks why the Southern Song scholars from Yongjia !" (in Wenzhou, modern Zhejiang Province) in particular paid attention to the text. He concludes that these scholars interpreted the Rituals as suggesting a decentralized and noninterventionist model of government based on conciliatory foreign policies, domestic stability, regional and local autonomy, low taxation, economic growth, and community building— and hence the opposite of Wang Anshi’s “New Commentary” on the text, which served as the ideological basis for his New Policies agenda. Those who opposed any recurrence of Wang Anshi’s reforms in the Southern Song attacked his interpretations. First, they reasserted canonical authority by refuting the official interpretation of the Rituals produced during the New Policies. Second, they redefined statesociety relations by producing new commentaries. Song also shows how these commentaries influenced statecraft ideas in Ming and Qing times, despite the great difference between the Ming autocratic state and the prime minister–centered Song state. Early Modern East Asia In chapter 9 Kate Nakai describes how Japanese classicists viewed Japan through the prisms of the Chinese Classics after 1600. Mito # ! scholars defined their political concerns during the Tokugawa period (1603–1868) in light of the Rituals’ wide-ranging program for state and society. On one level, they favored decentralized supporting

24

INTRODUCTION

structures, and they argued that the text offered a model for integrating military and civil administrative offices within a decentralized framework, which could be adapted to contemporary Japanese society. On the other hand, they criticized what they regarded as the disorganized character of Tokugawa social and political life. In the Rituals they found both motivation and validation for reshaping their political order. The reception of the text in Japan seems to have disguised or even encouraged a slippage between ceremonial commitments to a decentralized feudal system and an actual preference for more centripetal mechanisms of rule. Mito scholars stressed above all reliable methods for organizing and utilizing human resources, and the methods they singled out could be applied at either a macrocosmic or microcosmic level. Rather than just an authoritarian model for Japan, the Rituals gave Japanese thinkers enough intellectual space to see themselves in light of their own struggles between centralized and regional interests. Nakai also shows that Dazai Shundai 太宰春臺 (1680–1747), a scholar of ancient Confucian learning writing a century after those of the Mito school, opined that government in China had evolved from decentralization to centralization, while Japan’s state institutions had moved from centralized to decentralized. In his view, China’s centralized bureaucracy, formed as a structure of commanderies during the Qin dynasty, had betrayed the concept of decentralized statecraft enunciated in the Rituals, which represented the ideals of the feudal Three Dynasties of Antiquity (Xia 夏, Shang 商, and Zhou). He therefore saw the Japan of his day as organizationally closer to the ancient feudal ideal than imperial China, although he did not simply affirm the patterns of Tokugawa society as such. To the contrary, he found in the Rituals a model for a more systematic form of feudalism that, if instituted, would correct the inadequacies of Tokugawa political and social structures. We should add, based on Alexander Woodside’s presentation entitled “Vietnam and the Zhouli: The Political Theory of Phan Huy Chu (1762–1840)” at one of the workshops, that eighteenth-century Vietnamese classical scholars also contended that the rural ideals of a more locally oriented village elite remained in place in their country. They concluded that the Rituals emphasized primordial (pre-imperial) ideals of decentralization and that, therefore, Vietnam remained truer to the ancient classical ideal of the Chinese sage-kings than had imperial China. At the same time, however, as in China and also Japan,

INTRODUCTION

25

scholars in Vietnam used the text as a basis for critiquing the contemporary state of their own society and drew inspiration from it in developing programs of reform. Following up on Nakai’s account, JaHyun Haboush, in chapter 10, looks at the uses of the Rituals during the Chosôn 朝鮮 (1392–1910) period. In addition, John Duncan contributed a short piece on Korea at one of the earlier workshops, and Martina Deuchler presented a summary entitled “Some Notes on the Zhouli in Korea,” in which she maintained that for Koreans “the Zhouli was not an inspirational book, but its minute descriptions of a vast, hierarchically structured bureaucratic apparatus made it a practical handbook for dynastic founders. It is conceivable that early reformers in late Koryô 高麗 (918–1392), who strove to realize their vision of a new government order on a Confucian basis, were alerted to the work’s usefulness for administrative, ritual, and educational purposes by the Regulations of the Zhizheng [1341–1368] Era (Zhizheng tiaoge 至正條格), a slim Chinese (Yuan 元 dynasty, 1271–1368) compendium of government deliberations much consulted in late Koryô. In its preface, the late Yuan scholar Ouyang Xuan 歐陽玄 (1238–1357) referred to the vital importance of statutory laws and [positively] referred to the Zhouli as just such a compendium.” Haboush’s chapter goes into more depth to demonstrate how the Rituals continued to inspire reform-minded yangban 兩班 intellectuals after the Koryô period. Focusing on what meanings yangban drew from the text in Korea, her account addresses how well-known seventeenth- and eighteenth-century scholars (and scholar-officials) such as Yun Hyu 尹鑴 (1617–1680) and Chông Yagyong 丁若鏞 (1762–1836) used it to formulate their reform plans. In Haboush’s analysis, they acknowledged the Rituals as a central document of the East Asian textual community but also recognized its function as the most political of the classical texts. Yun’s and Chông’s blueprints for reform shared striking resemblances on crucial points. In their proposals for restructuring the government and political economy, for instance, they “both proposed the restoration of the State Council as the central decisionmaking organ, the abolition of the Censorate, and the discontinuation or drastic reconfiguration of the civil service examination.” Earlier, Chông Tojôn 丁若鏞 (1342–1398), one of the most powerful advisors to the Chosôn founder, Yi Sônggye 李成桂 (r. 1392– 1398), had prepared his Statutes for the Governance of Chosôn (Chosôn kyôngguk chôn 朝鮮經國典) as a radical departure from the

26

INTRODUCTION

practices of the Koryô dynasty. He favored a return to a prime minister–centered government and voiced strong commitment to a meritocracy. This emphasis on merit over birth divided symbolic and administrative power, with the throne possessing royal prestige and the prime minister holding the reins of administrative authority. Invoking the Rituals selectively to suit their purposes, Yun Hyu and Chông Yagyong returned to such views to empower new proposals for dealing with a hostile world. They were convinced that reconfirming Korea’s cultural identity was its only path toward survival. Ritual controversies were also informed by the Chosôn elite’s anxiety over their cultural and political identity in a Manchu-dominated China. Benjamin Elman opens chapter 11 by describing how the first Ming emperor, unlearned in the Classics, sought legitimation for his rule by relying initially on the activist institutions and precepts presented in the Rituals. As Song’s paper also shows, during both the Southern Song dynasty and Yuan-Ming transition (1250–1390), the text had remained respectable despite its association with the failed reforms of Wang Anshi’s earlier program. Ming dynasty rural control systems for village tax collection and household registration, for instance, had their locus classicus in the Rituals. Ming Taizu 明太祖 (r. 1368–1398) invoked these activist models for his statecraft policies. The emperor was persuaded by his scholar-official advisors that the classical canon would offer the best model of governance for his dynasty. Elman further shows that literati interest in the “Artificer’s Record” chapter of the Rituals increased during the Ming and Qing precisely at a time when Chinese scholars could appeal to it as a record of ancient technology that demonstrated Chinese counterparts to—and possibly the origins for—early modern European machinery. With the “Artificer’s Record” as an acknowledged Han dynasty addition to the Rituals, later evidential research scholars who championed Han classicism over Song learning could now valorize the chapter as an authentic record of ancient technology and use it to point to the “Chinese origins of Western learning.” Furthermore, at a time when the Rituals was increasingly considered a Wang Mang period forgery, classical scholars could now extract and isolate the “Artificer’s Record” from the rest of the text in order to point to the strengths of ancient Zhou technology without being struck broadside by the philological debates pointing to the questionable provenance of the Classic as a whole.

INTRODUCTION

27

Modern China In chapter 12 and in the “Denouement” for the volume, Rudolf Wagner shows that, while the Qing dynasty’s Manchu court continued to use the Rituals as a master text for dealing with unusual situations, such as visits by British and French emissaries, it had also become an essential reference point for late Qing reform projects. It was considered the ideal normative framework for revivifying the Manchu state by reapplying the precedents laid down by the Duke of Zhou but long since lost. The Rituals now belonged to a reimagined past before the purportedly corrupted “Imperial Confucianism” of Qin and Han had subverted ancient classical ideals. Thus, the text was now used to redress what its champions considered the main causes for China’s present problems, which had derived from the Qin conquest and the ensuing despotic governments that had lasted two millennia. Wagner adds that there was no antiquarian intent in this late Qing return to the Rituals. Rather, a revisionist classical interpretation emerged that claimed that the text had correctly encoded sagely principles of good governance. Rather than reviving the particular institutions and rules of the Zhou dynasty, however, late Qing reformers extracted these principles and held them up as a mirror to Western institutions to verify their compatibility with modern times. Wagner demonstrates that the late Qing uses of the Rituals as a reformist text lasted from the mid-nineteenth-century Taipings 太平 through the late-nineteenth-century self-strengthening discourse of “learning from the West” (yangwu 洋務) and to the time of Kang Youwei 康有為 (1858–1927). Sun Yirang’s 孫詒讓 (1848–1908) monumental commentary Corrected Essentials of the Rituals of Zhou (Zhouli zhengyi 周禮正義) was written in 1903 to refocus the uses of the text via reading strategies aimed at dealing with contemporary problems. *** At the end of our conference, one of our commentators, Christopher Minkowski of Oxford, presciently asked: “In what ongoing conversation does the project intervene?” In the end, what we have presented here is the life of a text that has been designated a Classic since the Tang dynasty and that was always at the cutting edge of imperial and royal statecraft in East Asia. However, the Rituals is arguably the sin-

28

INTRODUCTION

gle most obscure of the Classics, familiar to only a few sinologists, if the lack of devotion of Western studies to it is any measure. We have tried to articulate for a larger audience its constitutive meanings, who its readers were, and how expert they were. We have also addressed for which political, social, and intellectual constituencies the Rituals mattered most. The early chapters detail how the text was created, contextualized, and commented upon. The chapters dealing with the medieval and early modern periods show how the imperial state could intervene in the life of a Classic straightforwardly and with seeming impunity. Though the ongoing life of the Rituals was largely determined by literati in China, yangban in Korea, and commoners in Japan, who all acted collectively as classical scholars, our conference detailed—and this volume is filled with—historical examples of court officials promoting new or old commentaries or revising or excising canonical lists. We still have a distance to travel before we can show how and why rulers and scholars in a variety of East Asian states and societies ultimately decided what to make of the Rituals in practice. We present many important clues, however, in this volume. Nevertheless, the variety of uses of the text and its commentaries over time, from the second century to 1900, and across civilizations, from China to Japan, Korea, and Vietnam, belongs to a different story from that of the internal development of the meaning of the text, which, as some at the conference felt, had an internal logic of its own. But hardly anyone argued that such a logic was independent of the purposes and uses of the text. In fact, we would rather argue here that, from its inception, the textual logic of the Rituals was the result of its purposes and uses. We have tried to encompass both here, but our efforts to link the internal and external aspects of the life of the Rituals of Zhou in premodern East Asian statecraft remain a work in progress. The same is true for our understanding of the actual textual formation of the text, although the early chapters in this volume concur remarkably on how this formation should be historically contextualized. What we may have unintentionally accomplished with this volume is to unsettle the ways in which we should understand how some texts went about becoming—and unbecoming—“Confucian Classics.” The vicissitudes of the Rituals are instructive; for instance, by early modern times one of its minor and derivative chapters on technology could “wag the dog,” enabling classical scholars to unceremoniously scrap the rest of the Classic or place it in a lesser corner of the canon. Here

INTRODUCTION

29

we see a healthy tradition of polemical refutation, dismissal, and ridicule, which is so common in the Indian Sanskrit tradition but which we have usually left out of the story in East Asia. Others have often argued, since Edward Said, that “Oriental” classics such as the Rituals have been exhumed as idle curiosities prepared by “Orientalist” scholars. Although such views have created straw men—who have been dead for a long while, if they ever had a life to begin with—we have tried very hard to steer clear of such untimely accusations. We hope not only that this volume will contribute to further progress in understanding the Rituals of Zhou but that its publication will also encourage future volumes delineating the complex life of each of the Classics in East Asia. We are grateful for the support we received from the American Council of Learned Societies in cooperation with the Chiang Ching-kuo Foundation for Scholarly Exchange that enabled us to hold a four-day conference at Princeton to discuss the theme “New Perspectives on Chinese Culture and Society.” We are also thankful for the early support we received from the UCLA Center for Chinese Studies and the Princeton University Program in East Asian Studies to power up this project. At Princeton, the Humanities Council and the Davis Center for Historical Studies also provided timely support for the last workshops and final conference. And finally, all the authors in the volume are indebted to Pamela Bruton, our truly excellent copy editor; to Albert Hoffstädt and Patricia Radder at Brill who shepherded the volume through the publication process; and to Eno Compton and Michael Hunter, Ph.D. candidates in East Asian Studies at Princeton, who performed superb editorial work as we moved together from many conference papers to a coherent book.

PART ONE: EARLY CHINA

CHAPTER ONE

THE ZHOULI AS CONSTITUTIONAL TEXT David Schaberg, University of California at Los Angeles Since the Han dynasty (202 BCE–220 CE), the Rituals of Zhou (Zhouli 周禮) has had a place in the Chinese canon that can be understood fully only with reference to its “constitutional” promise. The text, initially known as the Offices of Zhou (Zhouguan 周官) (or, as I will argue below, Comprehensive Offices), incorporates both an office list and prescriptions for regular duties, and therefore bears comparison with similar blueprints of government from other places and times.1 The most influential legend of its origins, the attribution to the Duke of Zhou, tellingly echoes stories of constitution making from early European history. The work shows clear generic affinities with passages in other early Chinese texts, offering an administrative system that is internally well ordered, clearly founded on traditional and philosophical principles, and potentially workable; yet the Zhouli dwarfs all its Chinese analogues in scale and, if recent hypotheses on its provenance are correct, is unique among them for its close association with an actual government, that of the Qin dynasty (221–206 BCE). Finally, in a way that is characteristic of several early approaches to the problem of governing a unified empire, the text places two of the key activities of the central government—management of commerce and of communications—within a single system of circulation, making this system the foundation of the government’s legitimacy. In all these respects, even though it has never been adopted in toto as a template for government, the Zhouli has had a constitutional character. It is important to specify the sense in which the Zhouli might usefully be called “constitutional.” The Zhouli is clearly not “constitutional” in the same sense that a modern document like the United States Constitution is; it does not guarantee the rights of the people or specify the duties of government. Instead, I mean the word “constitu1

For an account of the changing titles of the work, see Wang Xueping 2007.

34

CHAPTER ONE

tion” as a translation of the term politeia as used by Greek writers debating the virtues of different “forms of government.” Aristotle (384– 322 BCE) defined the term thus: “A constitution is the arrangement of magistracies in a state, especially of the highest of all” (Aris. Pol. 3.6.1278b). This is the sense of politeia that had been adopted by Thucydides (ca. 460–ca. 395 BCE) in recounting Pericles’ (ca. 495– 429 BCE) remarks on the virtues of Athens’ democracy (Thuc. 2.37) and by Plato (429–347 BCE) throughout the work he called Politeia, now known to us as the Republic (e.g., Rep. 562a). Herodotus (fifth century BCE) had recounted a Persian deliberation, very likely legendary, over the relative merits of democracy, oligarchy, and monarchy (Herod. 3.80–83), and this juxtaposition of governments on the basis of their broad or narrow distribution of power continued to inform later Greek philosophical approaches to the question of politeia. For our purposes, then, the term “constitution” (as translation of politeia) is useful in that it identifies the Zhouli system as a monarchy, but one with certain internally justified arrangements for the distribution of power and responsibility. By reading the Zhouli “constitutionally,” we can begin to imagine how its framers would answer the kinds of questions a Plato or an Aristotle might ask of them. A “constitution” in the sense intended here may, but does not necessarily, take the form of a written document setting forth the laws, statutes, and customs that give a government its identity. Sometimes by committing its laws to writing a community commits itself to the rule of law. When Solon had framed a constitution for his fellow Athenians, he traveled abroad for a decade, binding them with a vow not to repeal the laws during his absence (Herod. 1.29).2 The Spartans’ lawgiver, Lycurgus (ninth century BCE), went further in his efforts to make amendments impossible, starving himself to death at Delphi once the oracle had confirmed that the laws he had imposed were good (Plutarch Lyc. 29; see also Herod. 1.65–66). Writing’s detachment from the living occasion of speech—the quality that made Plato regard the written text as a kind of orphan (Phaedrus 275e)—is indispensable in myths of this sort, which date to the first era of widespread Greek literacy, in the middle of the first millennium BCE. Even Lycurgus, who did not put his Spartan laws in writing, resorted to the medium when he sent news home of the oracle’s positive response. It 2

Herodotus does not call Solon’s laws a politeia; but the reforms are identified as a politeia in the Constitution of the Athenians attributed to Aristotle (Ath. Pol. 9).

THE ZHOULI AS CONSTITUTIONAL TEXT

35

is as if, one way or another, there must be a tangible relic of the speech acts that brought the laws into being: either a written legal code or, at the very least, a record of its divine ratification. The absence of legendary lawmakers is in this way figured as a virtue. They may be dead and gone or even fictional, but in their wisdom they still stand against frivolous amendments. The codes they have left behind are supposed to be sufficient on their own. The Zhouli seems to have been composed in an era and a political milieu that favored encyclopedic projects of writing. Jin Chunfeng 金 春峯, whose reasons for dating the Zhouli to the Qin dynasty are reviewed below, rightly compares the text to the Springs and Autumns Annals of Mr. Lü (Lüshi chunqiu 呂氏春秋), another third-century Qin effort notable for its orderly structure and its air of comprehensiveness.3 The formal perfection of the Zhouli was perhaps from the beginning understood to compensate in part for the lack of any internal claim of authorship: here was the system, perhaps even named “comprehensive” (zhou 周) in its title, and no personal attribution could add to its authority.4 It may have been predictable that by the late Western Han (202 BCE–9 CE) the Zhouli would come to be attributed to a Solon-like figure, the Duke of Zhou (Zhou gong Dan 周公旦), one of the eleventh-century BCE founders of the Zhou dynasty (ca. 1045–256 BCE). The attribution was in keeping with larger trends in Han classicism, which sought to link each of the Five Classics to historical moments of composition or collection, typically in the early Western Zhou or in the era of Confucius. But the association with the Duke of Zhou also eased the acceptance of the work as one of the ritual classics, and roughly corresponded with the change of the work’s title to Zhouli. For a text that in many places shows a greater affinity with the Writings of Lord Shang (Shangjun shu 商君書) or the Guanzi 管子 than with the Ceremonial Rituals (Yili 儀禮) or the Records of Ritual (Liji 禮記)—and that has been characterized as an amalgam of Legalist, Ru, and other ideas on statecraft5—this attribution would allow a reframing of its contents and a downplaying of its more technical and legalistic features. Once the identification with the Duke of Zhou was widely accepted, a work and a system that had once fit very aptly in their Qin 3

Jin Chunfeng 1993, “Zixu” 自序, 5. Jin Chunfeng 1993, 1–2. 5 Hou Jiaju 1987, 47–108. 4

36

CHAPTER ONE

context could be praised and cited by governments that otherwise disavowed their debts to Qin. To all appearances, the Duke of Zhou was the missing author and legislator; but the real absent Solon here was Qin itself and the anonymous Qin scholars who framed the Zhouli. To establish the relevance of a constitutional reading of the Zhouli, I first review arguments on the dating of the Zhouli and then examine passages from pre-Han and Western Han literature that are generically similar to the Zhouli. These office lists, all of them far shorter than the Zhouli itself, provide the immediate intellectual and compositional context for the Zhouli. They show that the project of administrative design was one undertaken by numerous thinkers in the final three centuries BCE, and as alternative designs they illuminate by contrast the principles adopted in the Zhouli. In the second part, then, I focus on one distinctive feature of the Zhouli as constitution, its provisions for the management of communication and commerce within the realm. The Zhouli designers appear to have believed that the state’s claims to legitimacy and all its prospects for survival depended both on centralized guidance of many economic activities and on the orderly transmission of information upward and downward within the political hierarchy. This governing philosophy jibes poorly with early stigmatizing views of the Qin as a regime reliant on harsh laws, but it may nonetheless reflect considered Qin ideals of governance. Finally, in a concluding reflection on the Zhouli’s idealization of comprehensive knowledge and the perfection of the administrative system, I argue that the text may well have been understood in the early Han as an example of Huang-Lao statecraft. This association may have helped justify the text’s initial transmission, its transformation from Comprehensive Offices (Zhouguan) to Rituals of Zhou (Zhouli), and its ultimate acceptance as a description of the Duke of Zhou’s system of royal governance. The Zhouli in its Generic Context Few Chinese texts have been dated as variously as the Zhouli. Zheng Xuan 鄭玄 (127–200) is the first scholar known to have attributed it to the Duke of Zhou; references to the text before his time do not seem to concern themselves with the composition of the text, focusing instead on its appearance in Han courts and the eventual appointment of Liu

THE ZHOULI AS CONSTITUTIONAL TEXT

37

Xin 劉歆 (46 BCE–23 CE) as its “academician” (boshi 博士).6 As Michael Nylan has argued, Zheng Xuan’s dating flew in the face of linguistic and other evidence but facilitated the bundling of the Zhouli with the other two rites classics, Yili and Liji, as a seemingly complete ritual system.7 The attribution to the Duke of Zhou has now been abandoned by careful scholars, though some have held that the Zhouli preserved some elements of a system first promulgated by the duke.8 Mainstream opinion on the dating of the work is now divided into two major camps, with the second clearly prevailing. The first group of scholars holds that the Zhouli is a concoction, possibly even a wholecloth fabrication, designed by Liu Xin and Wang Mang 王莽 (r. 9–23) to legitimize the latter’s dynastic pretensions. This view traces back at least to the Northern Song, when it figured in the objections of Sima Guang 司馬光 (1019–1086) and others to Wang Anshi’s 王安石 (1021–1086) reformist invocations of the Zhouli. Kang Youwei 康有 為 (1858–1927) famously attacked both the Zuo Tradition (Zuo zhuan 左傳) and the Zhouli in his Study of Forged Classics in Xin Dynasty Scholarship (Xinxue weijing kao 新學偽經考), sparking a debate that lasted through much of the twentieth century. The most potent renewal of the forgery charges in recent years has been that of Xu Fuguan 徐復觀 (1903–1982), who argued that Wang Mang himself had composed the Zhouli during five years out of power, and that the text had still lacked its final section, the “Winter Offices” (Dongguan 冬官), when Wang and Liu presented it at court.9 A larger group of scholars holds that the Zhouli is a product of the Warring States period. As a denial of the attribution to the Duke of Zhou, this view was at first expressed in pejorative terms: according to Jia Gongyan 賈公彥 (fl. 650), author of the Subcommentary on the Zhouli (Zhouli shu 周禮疏), He Xiu 何休 (129–182) considered it “a work of the devious strategists of the Warring States period” (liuguo yinmou zhi shu 六國陰謀之書), while Lin Xiaocun 林孝存 (second century CE; also known as Lin Shuo 林碩), a contemporary and opponent of Zheng Xuan’s, regarded it as “a presumptuous, unsubstantiated text from a recent and decadent era” (moshi douluan zhi shu 末世 6

Zhang Xincheng 1979, 282–283. Nylan 2001, 174–175. 8 See the view of Zhu Xi 朱熹 (1130–1200), cited in Zhang Xincheng 1979, 292. 9 Xu Fuguan 1980, 44–45, cited in Yu Yingshi 1990, 8–9. 7

38

CHAPTER ONE

瀆亂無驗之書).10 Su Che 蘇轍 (1039–1112) denied that the Duke of Zhou could have had anything to do with the work, arguing instead that it is contaminated by the opinions of Qin and Han Ru scholars perversely devoted to harsh laws.11 Milder evaluations of “Legalist” thought and Qin administration have brought more nuanced expressions of this view; some of these leave a minor editorial role to Liu Xin and Wang Mang. Zhang Xincheng 張心澂 concluded that, though it was based in part on Western Zhou and Spring and Autumn period institutions, the Zhouli was composed in the first part of the Warring States period by a Ru scholar with some expertise in Legalist thought and economic planning, then later tampered with and promulgated by Liu Xin, who adapted it to the interests of Wang Mang.12 On the basis of its handling of sacrifices, punishments, land tenure, and other institutions, none of which appears to have been implemented systematically during the early Han, Qian Mu 錢穆 (1895–1990) argued that the Zhouli was a late Warring States product.13 Hu Shi 胡適 (1891–1962) dated it to the Han, understanding it as Wang Mang’s amplification of a work that Sima Qian 司馬遷 (145?–86? BCE) had seen in an earlier form.14 Shi Jingcheng 史景成, citing the history of the Jupiter cycle, dated the text to the period between the completion of Lüshi chunqiu (241 BCE at the earliest) and the Qin unification (221 BCE).15 Hirase Takao 平 勢隆郎 critiques Qian Mu’s thesis, drawing on calendrical evidence to push the composition of the Zhouli back to the mid- to late Warring States period and argues for an association with the state of Yan 燕.16 The work of Jin Chunfeng has brought a new precision to research on the provenance of the Zhouli. In a monograph written in direct refutation of Xu Fuguan’s arguments, Jin contended that the Zhouli was the fruit of a characteristic Qin interest in continuing and reforming the practices of earlier governments, especially the Zhou, that it was composed in the final years of the Warring States period by scholars from the various conquered states, and that the zhou 周 of the title reflects the “comprehensiveness” of the presentation rather than 10

Cited in Zhang Xincheng 1979, 284. Compare the views of Zhang Zai 張載 and Cheng Yi 程頤, cited on the same page. 11 Cited in Zhang Xincheng 1979, 284–286. 12 Zhang Xincheng 1979, 313–327. 13 Qian Mu 1932. 14 Hu Shi 1935. 15 Shi Jingcheng 1966. 16 Hirase 2001.

THE ZHOULI AS CONSTITUTIONAL TEXT

39

any reference to the Duke of Zhou.17 Through a systematic comparison with Qin legal materials (especially those from Shuihudi 睡虎地) and received Qin texts (Shangjun shu and Lüshi chunqiu), he established that in numerous areas—taxation and corvée labor, punishments, education, sacrifices, funerals, slaveholding, and Five Phases theory, to name a few—the Zhouli most closely reflects Qin practices and beliefs. The study is cautious and comprehensive and leaves little doubt that, however Wang Mang and Liu Xin may have touched up the text, what we have is essentially a Qin document. As a project undertaken by Qin scholars in a spirit of preservation and adaptation, the Zhouli leaves behind the misleading implications that come with anachronistic attributions. It may in some ways reflect a long-standing Qin elite interest in the early Zhou kings whose territory they came to occupy, but it need not be read as an account of any political and ritual system imposed by the Duke of Zhou. It may have been useful in the hands of Wang Mang, but it need not be read as a forgery of his era. Instead, the Zhouli represents the familiar Qin drive to unification and rationalization carried out in the medium of political science. A feature of the text that several writers have noted, its fusion of “Confucian” and “Legalist” ideas, is in fact part of a much larger synthesis of all relevant political techniques, from the mild methods of education and religious display to the harsher rules of punishment, taxation, and conscription. The Confucian virtues are put in their place as one tool among many. It is within this synthetic statecraft that the Zhouli’s characteristic means of ordering commerce and communication come to the fore. These methods for controlling the circulation of goods and opinion were the main lesson the Qin scholars had learned from the past. To see how the Zhouli is distinct, it is necessary to place it within the context of other texts of its kind. In a very general way, then, the Zhouli might be compared with other ancient efforts to capture in writing the essential information necessary for the successful maintenance of a political or social entity. Such efforts would include lists of laws like the Code of Hammurabi or the Roman Twelve Tablets; they would include prescriptions for ritual practice like those found in the Pentateuch; and they would include titles lists of the kind known from

17

Jin Chunfeng 1993, 5–6, 288.

40

CHAPTER ONE

ancient Mesopotamia.18 In China, the broad context for the Zhouli would comprise scattered passages outlining ritual procedure and personnel, along with legal texts like the “Code of Lü” (“Lü xing” 呂刑) and other lost codes (perhaps unwritten, perhaps legendary) attributed to sage-kings.19 None of these compares directly with the Zhouli project, but all do reflect the strong motivation, early in the history of literacy, to fix in writing a perfected system of political organization. The office list as specific form fits within this larger category of legal and administrative writing. Some of the organizational and rhetorical habits that would ultimately issue in the Zhouli were in place long before that text and the other more modest constitutional schemes were produced. The earliest known precedents for direct prescription of the duties connected with a particular official title are the investiture commands (ming 命) that the Zhou kings issued to their subordinates. These commands depended for their effectiveness on the accoutrements of the speech act, including the occasion of a royal audience, the bestowal of gifts, and the pronouncement of the command itself.20 But often, if not always, they were also recorded in writing: texts of many commands were apparently kept in archives at court, while bronze inscriptions made by the addressee of the command also preserved information on the terms of the commands.21 Even as it continued, in the writings of the Western Zhou and later, to designate an official royal act of recognition and appointment, the word ming also acquired a more general nominal meaning: it came to stand for the whole lesson of political wisdom and virtue learned and taught by the early Zhou kings, a sort of total command from them to their lesser successors.22 The Zhouli may be read as a compendium of all possible and necessary commands that an ideal founding ruler might invite later imitators to obey. There are early precedents too for the recording, in a simple list, of offices and officeholders. In three passages relating to the state of Song 宋, the Zuo zhuan lists appointees to the positions of commander of the army of the right (youshi 右師), commander of the army of the 18 Nissen, Damerow, and Englund 1993, 110–115. I am grateful to Wang Haicheng for bringing the Mesopotamian parallel to my attention. 19 Chunqiu Zuozhuan zhu 1275 (Zhao 6.3). 20 Kern 2007, 140–151. 21 Kern 2007, 150. 22 I have discussed the complementary workings of writing and ceremonial speech in Schaberg 2004.

THE ZHOULI AS CONSTITUTIONAL TEXT

41

left (zuoshi 左師), minister of war (sima 司馬), minister of the multitude (situ 司徒), inspector of walls (sicheng 司城), and minister of justice (sikou 司寇).23 Another passage, also on Song, divides the duties of the minister of justice between two officials, distinguished as greater (da sikou 大司寇) and lesser (shao sikou 少司寇), and adds a premier (taizai 大宰) and vice premier (shaozai 少宰).24 These passages are remarkable not only for their form, which anticipates the more substantial lists of the constitutional texts, but also for their official titles, which appear also in the Zhouli and other texts.25 The closest parallel to these Song lists in the Zuo zhuan is a shorter list naming the Duke of Zhou as premier (taizai 太宰), Kang Shu 康叔 as minister of justice (sikou 司寇), and their younger brother Dan Ji 聃季 as minister of public works (sikong 司空).26 Several other passages give formally similar lists of top Jin 晉 appointments, mainly to military positions, but except in the case of the chief minister (qing 卿) and minister of war (sima), these do not overlap with the system apparently used in Song and remembered in later constitutional schemes.27 We come now to the Zhouli’s closest parallels in the early literature. The early Chinese passages that advance constitutional schemes seem to derive, conceptually and perhaps historically, from a combination of the rhetorical functions of commanding and listing. They give a sense, impossible to divine from the Zhouli alone, of the range of choices the constitutional imagination could draw upon. As Mark Edward Lewis observed in his brief discussion of several of them, they show the emerging separation of the king from the system of bureaucratic duties and ranks, the echoing of prescriptions associated with Shang Yang’s 商鞅 (ca. 385–338 BCE) “Legalism,” and the gradual accommodation of this system to the ordering principles of correlative cosmology.28 They also help to illustrate the rise of writing 23

Chunqiu Zuozhuan zhu 620–621 (Wen 16.5). Chunqiu Zuozhuan zhu 874 (Cheng 15.4). 25 Broman (1961) carefully notes which states used titles that also appeared in the Zhouli. 26 Chunqiu Zuozhuan zhu 1541 (Ding 4.1). 27 See Chunqiu Zuozhuan zhu 445–447 (Xi 27.4), 589–590 (Wen 12.6), 721 (Xuan 12.2), 865 (Cheng 13.3), 880 (Cheng 16.5), 999 (Xiang 13.3). One late passage gives similar lists for Zhou royal and Qi military commanders: see Chunqiu Zuozhuan zhu 1661–1662 (Ai 11.3). 28 Lewis 1999b, 42–48. The texts Lewis mentions are the “Li zheng” 立政 chapter of the Shangshu 尚書, the now-fragmentary “Wangdu ji” 王度記 and “Wangzhi” 王 制 chapters of Xunzi 荀子 and Liji 禮記, a long Zuozhuan 左傳 speech on the officers 24

42

CHAPTER ONE

as a tool of intellectual conquest. Lewis argues that “the culmination of the authority of writing in the Warring States was texts that sought to record the entire cosmos. The Zhouguan is an early example of such a text created as a model of the world.”29 Although it is not literally true that these texts seek “to record the entire cosmos” or to offer models of the world, they do offer comprehensive systems of higherlevel administrative organization by which rulers may hope to govern a part of the human and natural world. Rereading the several early Chinese constitutional passages, including some that Lewis did not discuss, it is possible to show in greater detail how they resemble the Zhouli and what they bring out in the Zhouli by serving as a foil for it. Each has its own internal principles of order, which bespeak basic assumptions about the key to political control. Neither the Zhouli nor any of its shorter counterparts truly seeks “to record the entire cosmos,” even if they do make provisions for a full complement of sacrifice. Instead, they seek to constitute a government that will respond most effectively to all the human, natural, and spiritual stimuli that it confronts; they seek to create a new organism, political in nature, that can thrive in the environment as given. To repeat, the Zhouli’s unique solution to this problem involves a coordinated approach to commerce and communications. But this becomes clear only after reading the whole set of texts. 1. Zuo zhuan, Zhao 17.3. When the ruler of the small state of Tan 郯 visits Lu 魯, a Lu nobleman asks him why his ancestor, the legendary emperor Shao Hao 少皞, named his official positions after birds. The visitor explains that several early rulers built their systems of official titles around omens that attended their rise to power: for the Yellow Emperor, for instance, clouds were the omen and officials were named after types of clouds, while others took their auspices from fire, water, dragons, or birds. He then lists his ancestor’s officials, translating bird names into more familiar equivalents: regulator of the calendar (lizheng 歷正), supervisor of equinoxes (sifen 司分), supervisor of solstices (sizhi 司至), supervisor of beginnings of the seasons (siqi 司 啟), supervisor of ends of seasons (sibi 司閉), minister of the multitude (situ), minister of war (sima), minister of public works (sikong), of King Shao Hao 少暤, and the “Wu xing” 五行 chapter of the Guanzi 管子. Most of these are also discussed below. 29 Lewis 1999b, 48.

THE ZHOULI AS CONSTITUTIONAL TEXT

43

minister of justice (sikou), supervisor of affairs (sishi 司事), gatherer of the people (jiumin 鳩民), regulator of artisans (gongzheng 工正), and regulator of farmers (nongzheng 農正). In each case the bird that gives the office its title corresponds somehow to the functions associated with that office. The swallow, for instance, arrives on the spring equinox and departs on the fall equinox and therefore corresponds to the supervisor of equinoxes.30 The bird scheme outlined here and the others mentioned in passing imply a faith that buried in sets of natural phenomena—in the set of all bird species, for example—are paradigms for the discrete set of offices that can form a successful government. Birds have no government themselves; Shao Hao’s paradigm of thirteen offices is initiated by the interpretation of an omen, that is, by the recognition of bird species’ distinct meanings for humans. Government in this paradigm is mainly a matter of scheduling agricultural activities and organizing collective labor. In this mythologizing vision of the government of distant antiquity, the focus is all on direct contact between humans and the usable resources of the natural world. Minimal attention is accorded the problem of social interactions among humans.31 2. Classic of Documents (Shangshu 尚書), “Canons of Yao” (“Yao dian” 堯典). In consultation with his senior advisors, the sageemperor Shun 舜 identifies qualified men and makes appointments to nine positions: minister of public works (sikong), ruler of the millet (houji 后稷), minister of the multitude (situ), justiciar (shi 士), master of artisans (gonggong 共工), royal forester (zhenyu 朕虞), regulator of the ancestral cult and rites (zhizong 秩宗), supervisor of music (dianyue 典樂), and informant (nayan 納言). 32 Duties for each are clearly prescribed in commands from Shun.33 Compared with the bird offices of Shao Hao, Yao’s system is slightly more removed from the confrontation with nature and slightly more attentive to the internal management of the human community. Agriculture and hunting and gathering still have dedicated appointees (houji and zhenyu), as does the shaping of the environment (sikong, a 30

For the explanations of Du Yu 杜預 (222–284) and other commentators, see Chunqiu Zuozhuan zhu 1386–1389. 31 See also Chunqiu Zuozhuan zhu 1500–1504 (Zhao 29.4), discussed below, which offers a lengthy mythical account of offices correlated with the Five Phases. 32 For official titles, I have adopted translations from the draft version by Serruys, Nylan, and Schaberg (n.d.), in some cases modifying them to match Broman’s translations for the Zhouli. 33 See Gu Jiegang and Liu Qiyu 2005, 1.191–334.

44

CHAPTER ONE

position filled by Yu 禹 himself), but the management of human affairs such as ritual (zhizong), public moral education (situ), and the education of young noblemen in harmony (dianyue) now dominates. Most remarkably, there is now an officer for the transmission of language to and from the ruler (nayan), whom Shun appoints in these terms: “Long, I abhor slanderous speech and destructive behavior that agitates and alarms my populace. I charge you to act as informant. Day and night send out and bring in our royal commands. Be you only truthful!” (龍，朕堲讒說殄行震驚朕師，命汝作納言，夙夜出納 朕命，惟允). 34 The supervisor of music, too, manages speech in teaching correct song lyrics (shi 詩). Although Yao’s appointments in the opening of the “Yao dian” show some rudimentary correlation with the four directions, Shun’s appointments are entirely lacking in correlative patterning, and the archaic or archaizing language of the text discourages most verbal parallelism. The effect is to focus attention on the administrative foresight and authority of the ruler, who is quoted throughout. This is the sage-king as prime mover. He is a heroic inventor of institutions, but the system will run of itself once he is finished. One is reminded of the empty space assigned king or emperor in the Zhouli system, which is designed to operate without him.35 Further, where Yao appointed officials in the four directions, the Shun of the “Yao dian” embarks on grand tours of his realm, to the east, to the south, to the west, and to the north, figuratively bringing governed space under control by moving through it; these travels prefigure the First Emperor of Qin’s (Qin Shihuang 秦始皇, r. 221–210 BCE) later imperial progresses. It is worth mentioning here that numerous scholars have regarded the “Yao dian” as a Qin court edition transmitted into the Han.36 Perhaps the role assigned the ruler and the travels attributed to him signify a historical link with the Zhouli constitution. Three other texts are relevant to the Shangshu’s depiction of early government order. In what may be the earliest extensive list of official titles, the “Establishing Government” (“Li zheng” 立政) chapter includes certain familiars—situ, sima, sikong—among a larger number of titles that have little resonance in later texts; it does not discuss duties. In a single sentence and without elaboration, the “Great Plan” 34

Serruys, Nylan, and Schaberg n.d. See Jin Chunfeng 1993, 6–8. 36 Chen Mengjia 1985, 135–146. See also Kern 2000b, 111, 192–193. 35

THE ZHOULI AS CONSTITUTIONAL TEXT

45

(“Hong fan” 洪範) chapter lists the “eight governances” (ba zheng 八 政): food (shi 食), commodities (huo 貨), sacrifices (si 祀), supervising of public works (sikong), supervising of the multitude (situ), supervising of justice (sikou), guests (bin 賓), and master or army (shi 師). Unsurprisingly, the “old text” (guwen 古文) chapter “Offices of Zhou” (“Zhouguan” 周官) presents the same system of officials as the Zhouli. This chapter is a late reconstruction of a missing chapter of this title, offered to the Jin (265–420) court in 317 by Mei Ze 梅賾; it is interesting largely for describing the duties of each officer in regular trisyllables, a prosodic feature that is seen in some of the other constitutional texts, including parts of the “Yao dian.” The presence of these passages in the Shangshu, even without elaboration of the kind found in the “Yao dian,” reinforces the impression that for Qin and Han elites, the Shu served as a kind of mythical thesaurus of government practices. 3. Xunzi, “Royal System” (“Wangzhi” 王制). This essay, closely linked to others in the Xunzi by its concern with resource allocation, hierarchy, and the justification of ritual, contains a brief section separately titled “Ordering the Officers” (“Xu guan” 序官). Its officers are the following: the steward chief (zaijue 宰爵), minister of the multitude (situ), minister of war (sima), grand master (taishi 大師), minister of public works (sikong), controller of fields (zhitian 治田), forestry master (yushi 虞師), district master (xiangshi 鄉師), master of artisans (gongshi 工師), hunchbacked mediums and lame shamans (yuwu boxi 傴巫跛擊), controller of markets (zhi shi 治市), minister of justice (sikou), premier (zhongzai 冢宰), ruling lord (bigong 辟公), and Heaven-appointed king (tianwang 天王).37 Although the ordering principles of this list are by no means obvious, overall it appears to move inward from peripheral areas, building toward the urban and palatial settings where language, wealth, and power are most concentrated. The first three officers “know the numbers” (zhi…shu 知…數) of various things in their charge; they act as managers of the rituals, human resources, and techniques (also shu 數) that will support the other officers’ work. For each of the remaining twelve, we get a short list of the duties that are “the business of” the officer in question (…zhi shi 之事). The grand master (taishi) is in 37 Xunzi jijie 1.166–171. Tianwang (the king who is “Heaven-appointed,” “Heaven-favored,” or otherwise identified with Heaven) is the regular way of referring, in the Zuozhuan and Xunzi and a number of other texts, to the Zhou king.

46

CHAPTER ONE

charge of the forms of language—commands and poetry—that put the resources in motion. The next three officers manage the paths and substance of actual economic circulation, including waterways (sikong, successors to Yu), agricultural lands (zhitian), and uncultivated lands (yushi). Next, moving inward toward the urban center, the district master organizes dwellings, small-scale agriculture, and moral education in the areas around the city. The master of artisans, the mediums and shamans, and the controller of markets oversee activities that would have taken place near the center of the city, in the marketplace. The last officers control instruction of all kinds—by punishment, by reward, by moral example—presumably from the palace itself. A recurrent word in this passage is “following” (shun 順), usually in accordance with the proper timing, sometimes with periodic “extension” or “adjustment” (xiu 修). Everything in this hierarchically arranged space is to follow smoothly through top-down teaching and the proper management of resources. 4. Liji, “Royal System” (“Wangzhi” 王制). This chapter resembles many others in the Liji in that it gives the impression of having been compiled from disparate passages, some of them quite brief. Recent studies have dated its composition as early as the fourth century BCE and as late as the middle second century BCE.38 The account of the main offices, though at points detailed, is not entirely coherent or well ordered and is perhaps made up of categorized fragments of information. In order, the offices listed are premier (zhongzai), minister of public works (sikong), minister of the multitude (situ), regulator of musicians (yuezheng 樂正), minister of war (sima), minister of justice (sikou), chief recorder (taishi 大史), and accountant (sikuai 司會); in some cases subordinate personnel are mentioned.39 A later part of the chapter specifies salaries for various ranks of official.40 Although the Liji office list lacks the internal organization of the Xunzi “ordering of the offices,” it recalls that passage in its attention to the maintenance and cultivation of human resources. The zhongzai oversees all income and budgeting, attending especially to expenditures on sacrifices and funerals. The sikong has charge of standards, geographical space, and the way that space is occupied; when all is in 38 For a recent effort to date the oldest sections of the “Wangzhi” to the late fourth century, see Wang E 2006. Wang Yunfei (2007) argues that the text must have undergone revision in the era of Emperor Wu (r. 141–87 BCE). 39 Liji jijie 1.337–377. 40 Liji jijie 3.395–396.

THE ZHOULI AS CONSTITUTIONAL TEXT

47

order, he “promotes study” (xingxue 興學). The situ then teaches rites and moral standards, enforces the latter, and receives nominations of talented locals for office or advanced study. The yuezheng instructs young noblemen in the songs (shi 詩), documents (shu 書), rites, and music. The sima controls appointments to office, not only within the military hierarchy but elsewhere as well. Finally, the sikou enforces educational gains at the center, hearing cases and punishing legal hairsplitting, ritual violations, sophistical excesses, and religious fakery. The pattern is less pronounced than in the Xunzi, but here too there is a movement upward and toward the center of power, with an accompanying cultivation of the best human talent for service there. 5. Guanzi 管子, “Five Phases” (“Wu xing” 五行).41 The offices in this relatively brief list are the following: minister in charge of the seasons (dangshi 當時), minister in charge of granaries (linzhe 廩者), master of lands (tushi 土師), minister of the multitude (situ), minister of war (sima), and minister of justice (li 李).42 The list is historicized as an account of the Yellow Emperor’s (huangdi 黃帝) appointments of named individuals, each distinguished by his “discrimination” (bian 辨/辯) of affairs in a particular geographical quadrant, to the offices appropriate to those quadrants. In an echo of the Zhouli, the system of officers is correlated with the cardinal directions and then with the seasons. Minimal attention is given to the duties of the officers here, though there is much elsewhere in the Guanzi that has near parallels in the Zhouli.43 6. The Elder Dai’s Records of Ritual (Da Dai liji 大戴禮記), “One Thousand Chariots” (“Qian sheng” 千乘).44 An unidentified master (zi 子), speaking to an unidentified lord (gong 公), outlines a system of four offices—minister of the multitude (situ), minister of war (sima), minister of justice (sikou), and minister of public works (sikong)— coordinated with the four seasons, spring through winter.45 There is some attention to speech: to the prayers of the situ, to the teachings of the sima, and to the sikou’s judicial duties and controls on speech. 41 Rickett (1998, 119–120) cites views placing the “Wu xing” chapter as early as before the fourth century BCE and as late as the early Han. Rickett himself dates the text to the first half of the third century. 42 Guanzi jiaoshi 364. I have adopted some of the translations of the titles, and altered others, from Rickett 1998, 123–124. 43 Gu Jiegang (1979), among others, drew attention to connections between the Zhouli and the Guanzi. 44 For an overview of the provenance of the Da Dai liji, see Huang Huaixin 2005. 45 Da Dai liji jiegu 157–161.

48

CHAPTER ONE

Another Da Dai liji passage, from the “Flourishing Virtue” (“Sheng de” 盛德) chapter, offers a list of chief officers identical to the Zhouli’s but coordinated with key Confucian terms: the zhongzai with the Way, situ with virtue, minister of cult and rites (zongbo 宗伯) with humaneness, sima with the sage, sikou with rightness, and sikong with ritual.46 7. Luxuriant Dew of the Springs and Autumns Annals (Chunqiu fanlu 春 秋 繁 露 ), “Five Phases Mutually Generating” (“Wuxing xiangsheng” 五行相生).47 Expounded in an essay without narrative frame, this scheme is more highly correlative than any other, not only matching offices and duties with cardinal directions, virtues, historical individuals, and the Five Phases but also linking them in a cycle of mutual production: supervisor of agriculture (sinong 司農), supervisor of horse (sima), supervisor of organization (siying 司營), situ, and sikou.48 The succeeding chapter, which presents the same offices in the mutual-conquest order of their corresponding phases, is extraordinary for its focus on the consequences of official malfeasance.49 Attention to the management of speech and revenues is concentrated under the duties of the siying. The several constitutional schemes appear to draw upon a core of shared knowledge and adopt similar techniques. The heart of each system, however it might be elaborated, is the ministers or supervisors (si 司), one or more of whom appear in every text. Building upon its version of the list of officers, each text replicates in rhetoric what it hopes the official system itself will accomplish with regard to the world being governed: it distributes duties among offices according to a common ordering principle, which is implied in part by the verbal parallelism that characterizes any list. Each text proposes a slightly different way of accommodating the world’s divisions and needs and controlling territory and time. The intensely correlative accounts of the later texts are in some ways the logical culmination of the necessity of distribution itself: even before the sikou, for example, is identified with

46

Da Dai liji jiegu 146–147. There is a close parallel to this passage in the “Zhi pei” 執轡 chapter of Kongzi jiayu 孔子家語 (Kongzi jiayu zhuzi suoyin 47). 47 Arbuckle (1991, 490–491) argues that this chapter derives from the end of the Western Han. 48 Chunqiu fanlu yizheng 362–366. 49 Chunqiu fanlu yizheng 366–371.

THE ZHOULI AS CONSTITUTIONAL TEXT

49

the west, autumn, and rightness, his punitive duties have had to be separated in a coherent way from the duties of his colleagues. Two opposing impulses stand out in the progression represented by the seven texts. The most correlative and also possibly the latest among them favor parallelism and abstract order at the expense of practical engagement with real natural and human resources. These take the logic of the list to its extreme. The others, by contrast, seem to derive from the tradition of individual commands or appointments (ming) and favor specific forms of service and accountability at the expense of overall systemic coherence. Either extreme is a failure. Without a purchase on the practical details of resource management, the government will starve, but without some clearly represented principles of its own order, the government will be unable to preserve and reproduce itself. From this perspective, the success of the Zhouli as model constitution lies in its combination of particular duties and abstract principles of order. Regulated Exchange: Speech and Commerce in the Zhouli The Zhouli is easily summarized. As presently arranged, the text, like the royal government it is supposed to represent, is divided into six sections: the offices of Heaven (tianguan 天官), led by the premier (taizai 大宰) and responsible for court and palace affairs; the offices of Earth (diguan 帝官), led by the minister of the multitude (da situ 大 司徒) and responsible for the affairs of the populace in the countryside and cities; the offices of spring (chunguan 春官), led by the minister of cult and rites (da zongbo 大宗伯) and responsible for sacrificial affairs; the offices of summer (xiaguan 夏官), led by the minister of war (da sima 大司馬) and responsible for warfare and relations with the feudal nobility; the offices of autumn (qiuguan 秋官), led by the minister of justice (da sikou 大司寇) and responsible for penal law; and the offices of winter (dongguan 冬官), which was perhaps originally designed to be led by a minister of public works (sikong) but is now devoted to specifications for the work of artisans. The numbers of offices and their apportionment among the six departments imply an effort to create numerological correspondences that would facilitate central control of the realm.50 50

Lewis (1999b, 47) discusses other aspects of numerical correlation in the construction of the text. See also Hou Jiaju 1987, 19–28, on numbers as they relate to the

50

CHAPTER ONE

The Zhouli reconciles the tensions of particularity and principle that emerged in the other constitutional schemes. There are enough offices and enough prescribed duties to imply a clear-eyed understanding of a central government’s full range of responsibilities. On the side of principle, the six-part division, corresponding to the canonical set of six directions (above, below, east, south, west, north), suggests an overarching, rational architecture for the system. Given the seasonal associations provided for in correlative cosmology (e.g., of punishment with autumn), the relevant image for the Zhouli is perhaps that of the four seasons revolving like a wheel on an axis formed by Heaven and Earth. This image would certainly complement the use of the term zhou in the title of the work and in the compound zhou zhi 周知, discussed below. Still, despite the conspicuousness of the six-part arrangement, its correlative aspects barely penetrate to the level of the particular offices and their duties. The Zhouli’s true achievements in the rationalization of offices lie elsewhere. Other constitutional schemes had paired the management of the realm’s resources with a commitment to top-down moral education, “transformation by teaching” (jiaohua 教 化). To borrow a term used in the Xunzi chapter “Wangzhi,” a cultivation of compliance (shun 順) throughout the realm would make for uninterrupted economic activity and in turn support the central institutions that protected the people’s livelihood. The Zhouli adopts the same combination, but on a much vaster scale. And it amplifies one element that is just barely introduced in one of the earlier schemes, the “Yao dian”: this is the channeling of information and opinion from below in reciprocation of the teachings flowing outward and downward from the central court. Drawing on older notions of circulation and exchange within the realm, the Zhouli scholars devised a system in which resources and information moved in the same networks, with the capital and court in each case serving as a place for collection and redistribution. Beneath the six-part surface, then, the principle that truly informs the Zhouli offices is the assimilation of communication and material exchange within a single system of circulation. This system’s durability depends largely on its success in cultivating a wealth of resources and information and in converting each into the other.

question of the authenticity of the present Offices of Winter, the final section of the text.

THE ZHOULI AS CONSTITUTIONAL TEXT

51

In its economic prescriptions, the Zhouli reflects a late Warring States awareness of the workings of supply and demand and a fascination with the central government’s unique power as an economic player. As a matter of course, the government is to charge fees on every sort of profit and exchange in the territory it governs. It will not seek monopolies or deliberately manipulate prices, but it will buy up goods that are not selling and have become a burden on the market.51 By collecting resources at numerous local checkpoints and markets, pooling them, then redistributing and reinvesting them in its own projects of maintenance and expansion, the government manages natural economic circumstances to its own advantage. The notion is one that would emerge in clearest and most extreme form in the economic chapters of the Guanzi, which hold that the state should accumulate a massive concentration of wealth in various forms and then control prices and supplies in the public markets by adjusting (and sometimes cutting off) the flow of its own resources into these markets.52 The same basic principle found its way into Han administrative practice in the form of the “balanced standard” (pingzhun 平準) and the salt and iron monopolies.53 The Zhouli’s management of communication assumes similar principles. In doing so, it draws upon ideas already well developed in Warring States thought. It approaches language itself not only as a thing in the world but also as a model of the world and a tool for bringing order to it. Many offices are named for the object or resource they manage—the salt man (yanren 鹽人), for example, or the dog man (quanren 犬人)—and their human energies matter to the court only as a means of realizing the order of words and thereby securing for the court the resources it needed. Language use is treated as a matter of matching names and substances (ming shi 名實), and one official after another is enjoined to “make correct distinctions among the names and objects” (bian qi mingwu 辨其名物) that his duties involve.54 In the background lie the ambitions both of the “rectification 51 See, however, Zhouli zhengyi 4.1095. Such purchases put the massive economic power of the state to work in stimulating the market; see Li Puguo 1987, 73. 52 According to their able translator W. Allyn Rickett, the economic chapters of the Guanzi probably date to the early or middle Han, although the evidence is murky enough to allow some scholars to date them to the Warring States period. See Rickett 1998, 357. 53 On the “balanced standard,” see Shiji 1417–1443; Swann 1950, 65. On the salt and iron monopolies under Qin and Han, see Swann 1950, 61–63; D. Wagner 2001. 54 E.g., Zhouli zhengyi 1.257.

52

CHAPTER ONE

in accordance with names” (zhengming 正 名 ), expressed in the Analects, and of the matching of “forms and names” (xingming 形名) espoused by Han Fei 韓非 (ca. 280–ca. 233 BCE) and others. In zhengming, naming fixes and specifies the usable resources of the world.55 In xingming the system of names is put into motion, as an official’s duties and performance are tested rigorously against the letter of his job description.56 The Zhouli’s tightly organized office list and clear enumerations of duties establish a basic verbal framework for the control of the world’s resources. In its comprehension of geographical and social space, the Zhouli likewise draws older ideas into a new synthesis, in which information and other resources tend to move along the same or similar conduits. In keeping with the Zhouli’s inclusiveness—its readiness to join ideas that had previously been expressed in opposition to one another— local society is organized both in nine-part well-field units of the sort Mengzi 孟子 (fl. ca. 330 BCE) favored and in the groups of five and ten households advocated in the Guanzi and the Shangjun shu.57 The well-field units are perhaps mentioned with a nod to traditional or legendary ways of agricultural organization. One of the officials who is responsible for keeping them in order, the vice minister of the multitude (xiao situ 小司徒), relies much more heavily on the grouped households, which serve as targets of taxation, conscription, and moral education, and whose members are bound to one another in mutual legal responsibility. Many of the functions of the associations, especially the legal and educational, involve the transmission of language to and from the central government.58 In its attention to the management of the unified realm, the Zhouli reflects Warring States thinkers’ admiration of detailed geographical and ethnographical knowledge and their awareness of the way that material exchanges, often of a frankly commercial kind, strengthened ritual and cultural networks. A listing of the nine regions, their ethnic characteristics, and their local products and tribute goods recalls the “Tribute of Yu” (“Yu gong” 禹貢) chapter of the Shangshu 尚書. In 55

See Lunyu jishi 3.855–857, 885–896 (Analects 12.11, 13.1). Han Feizi jishi 2.111–112 (“Er bing” 二柄). 57 For the well-field system, see Zhouli zhengyi 3.786. For grouped households, see Zhouli zhengyi 11.2785 and Guanzi jiaoshi 30 (“Li zheng” 立政). The Zhouli, like the Shangjun shu, also organizes groups of five for taxation and legal responsibility; see Zhouli zhengyi 3.776 and Shangjun shu 114–116. Lewis (2006, 247–252) discusses the historical and intellectual context of the well-field concept. 58 Zhouli zhengyi 1.222, 3.776, 3.830, 3.881, 11.2785. 56

THE ZHOULI AS CONSTITUTIONAL TEXT

53

both texts, ethnographic knowledge is a prerequisite for economic and political control, with cultural diffusion from the center helping to hold the network together.59 Both texts also incorporate the schematic view of the realm plotted in concentric squares around the royal domain, assigning each zone its proper name in the hierarchy.60 As a comprehensive text and as a blueprint for government, the Zhouli applies the lessons of economic management to the problem of information management. It does not prescribe monopoly or total control, opting instead for countless smaller prescriptions and interventions, which all together make the Zhouli a web of sanctioned types and occasions of speech, a guide to all of the utterances, verbal or written, that should and should not take place in the well-ordered realm. In a figurative sense, the Zhouli collects resources from a whole preceding history of prestigious speech, incorporating the forms and occasions of speech that had been most specialized and most effective in the administration of the Zhou dynasty. But then, in a practical and literal sense, it arranges these forms and occasions into a circulatory system for the realm’s internal communications. Predictably, many of the traditional speech forms that the Zhouli incorporates into its system are used for diffusing information from the top and the center and in economic terms are equivalent to public works projects. The chief criminal judges (shishi 士師), for example, are responsible for the “five admonitions” (wu jie 五戒), by which they “aid in punishments, preventing incriminations from besetting the people: first, the harangue, used with the army; second, the proclamation, used at meetings and gatherings; third, the prohibition, used in labor conscriptions; fourth, the correction, used in the walled city; and fifth, the announcement, used in lesser towns and rural areas” (先後刑 罰，毋使罪麗于民：一曰誓，用之于軍旅；二曰誥，用之于會同 ；三曰禁，用諸田役；四曰糾，用諸國中；五曰憲，用諸都 鄙。).61 The chief prayer masters (taizhu 大祝) “create the six disqui59

The Zhouli’s ethnography is to be found in the prescriptions for the master of the four quarters (zhifangshi 職方氏); Zhouli zhengyi 10.2640–2671. In addition to the “Yu gong,” see the remarks on the minister of the multitude’s duties and on geography in the “Wangzhi” chapter of the Liji (Liji jijie 1.390, 359–360). Huang Huaixin, Zhang Maorong, and Tian Xudong (1995) point up close parallels between the Zhouli’s zhifang ethnography and the “Zhifang jie” 職方解 chapter of the Yi Zhoushu 逸周書; see Yi Zhoushu huijiao jizhu 2.1038–1063. On material and cultural networks, see Schaberg 1999. 60 On systems of concentrically arranged zones, see Lewis 2006, 271–273. 61 Zhouli zhengyi 11.2783.

54

CHAPTER ONE

sitions in order to connect superior and inferior, intimate and strange, far and near: first, the address; second, the command; third, the proclamation; fourth, the collective pronouncement; fifth, the supplication; sixth, the eulogy” (作六辭，以通上下親疏遠近，一曰祠，二曰命 ，三曰誥，四曰會，五曰禱，六曰誄。).62 The prayer masters for covenants (zuzhu 詛祝) “are responsible for the invocations cried aloud in covenants, imprecations, prayers to the high lord, prayers to ancestors, prayers of blame, prayers of explanation, prayers in time of disaster, and prayers in time of plague. They create the written addresses for covenants and imprecations in such a way as to regulate the good faith of the state and to substantiate the agreements of the local polities” (掌盟、詛、類、造、攻、說、禬、禜之祝號。作盟 詛之載辭，以敘國之信用，以質邦國之劑信。).63 Some of the genres mentioned in connection with these offices are otherwise known to us from collections like the Shangshu or from accounts of high-level diplomacy during the Spring and Autumn period. The command-like nature of many of these forms of speech is clear. They are delivered from above and invoke the power of the government to enforce its policies. Prayer, directed upward, serves likewise to reinforce secular power by displaying the prestige of the sacrificants and their access to the divine world. But the distinction of the Zhouli system lies in the way it organizes and stimulates certain forms of speech directed from the locales toward the center, much in the way it forces the rendering of other sorts of resources to the royal court. Exploiting the resources of earlier speech genres, the central government is to establish and control the conduits by which local, popularlevel knowledge is collected and used. Perhaps the most telling of the Zhouli’s provisions for collecting information is the procedure known as “consultation” (xun 詢, OC *sәwjin).64 In “consultation,” an officer solicits the views of a group of people who are not his direct subordinates and not necessarily themselves part of the official structure. The consultation is thus used to investigate opinion and foster consensus: It is the duty of the vice minister of justice to be responsible for the administration of the outer court in such a way as to muster the myriad people and consult with them: first, to consult about dangers facing the state; second, to consult about relocating the state; and third, to consult 62

Zhouli zhengyi 8.1992. Zhouli zhengyi 8.2060–2061. 64 For this and other Old Chinese (OC) reconstructions, see Schuessler 1987. 63

THE ZHOULI AS CONSTITUTIONAL TEXT

55

about installing a ruler. Positions for consultation: the king faces south; the three lords, the regional chiefs, and the hundred clans face north; the assembled ministers face west; and the assembled functionaries face east. The vice minister of justice receives the guests, leading them in in order and inquiring with them. He uses the populace to support proper aims and to determine plans. 小司寇之職，掌外朝之政，以致萬民而詢焉。一曰詢國危。二曰 詢國遷。三曰詢立君。其位：王南鄉，三公及州長百姓北面，群 臣西面，群吏東面。小司寇擯以敘進而問焉，以眾輔志而弊謀。65

Here, as Sun Yirang observes, the king faces the multitude (the hundred clans) directly, signaling his esteem for their opinion as the members of his administration look on. The “consultation” is likewise to be used in the cultivation of human resources, specifically in the identification of new personnel from the locales. Employing the conventions of the “district archery ritual” (xiangshe 鄉射), the xiang district administrator (xiang dafu 鄉大夫) “consults with the populace” (xun zhongshu 詢眾庶) and invites them to promote the worthies (xing xian 興賢) among them. “When the state holds a great consultation with its populace, each district administrator leads many or few from his district and delivers them to court” (國大詢于眾庶，則各帥其鄉之眾寡，而致於朝).66 Texts securely dated to the Zhou attest to the antiquity of this ideal of consultation.67 The term xun 詢 is linked in various ways with four other Zhouli terms relating to communication. Its phonetic xun 旬 (OC *swjiin), “everywhere, all around,” suggests that xun 詢 is cognate with jun 鈞 (OC *kwjin), “potter’s wheel,” an implement that would have provided a vivid image of comprehensiveness through its constant rotation.68 A member of the same word family is xun 徇 (OC *swjins), “circuiting,” a practice that in the Zhouli and other texts typically takes the consultant out of court and puts him in motion among his public, sometimes with a wooden-tongued bell to summon them to attention.69 The focus here is less on the exchange of speech than on 65

Zhouli zhengyi 11.2766. Zhouli zhengyi 3.850–851. 67 See Shijing, “Huanghuangzhe hua” 皇皇者華 (Mao 163) and “Ban” 板 (Mao 254); the interpretation of Mao 163 at Chunqiu Zuozhuan zhu 933 (Xiang 4.3), and Guoyu jijie 178–181 (“Lu yu” 2); and the interesting echoes of the poem at Guoyu jijie 361–362 (“Jin yu” 4). 68 Schuessler 2007, 328, 548–549. 69 Zhouli zhengyi 1.186, 3.815, 3.834, 11.2782. Schuessler (1987) does not include this graph. I adapt the reconstruction of Schuessler (2007, 328). 66

56

CHAPTER ONE

physical circulation, which is undertaken for the sake of displaying a criminal justly punished, for reminding military officers of the deadly consequences of disobeying a harangue (shi 誓), or for annual reminders (the language of which is quoted) to the populace that violations of the law will be punished.70 In the Zuo zhuan, the words uttered in the course of circulation have a marked generic character, always brief and usually consisting of promises, injunctions, or forceful definitions.71 In one passage on strife in the state of Song, xun 徇 involves a good deal of dialogue, negotiation, and consultation with the governed populace.72 Another Zhouli term that may be etymologically related to xun 詢 is xun 訊 (OC *sjins), “conferring,” in which the vice minister of justice inquires with court subjects, functionaries, and commoners as to the appropriateness of criminal penalties he has imposed.73 The emphasis in xun 訊 is on questioning, that is, on gathering language rather than prescribing it; the text states explicitly that the vice minister is to hearken to the people in cases of capital punishment.74 But like xun 詢, xun 訊 puts the representative of the government in a position to exploit the opinions of the populace in an overt and organized fashion. A third term that can, like xun 徇, involve travel and the transmission of opinion is xun 循 (OC *sljuәn), which ordinarily means “to follow” or, in Schuessler’s more elaborate definition, “to follow something that is pre-determined, inspection tour.”75 Here—also considering the difference in the main vowel—there is no direct etymological connection with xun 詢 in its sense of comprehensive circulation. There may, however, be an indirect connection by way of xun 訊, with 70 For the criminal displayed, see Zhouli zhengyi 4.1092; for the quoted reinforcement of the harangue, see 9.2333; and for the quoted reminders, see 1.186, 3.815. 71 See Chunqiu Zuozhuan zhu 305, 1253, 1675 (Xi 5.2, Zhao 4.4, Ai 13.1). 72 Chunqiu Zuozhuan zhu 1731 (Ai 26.2). 73 Despite the associations of xun 詢 with xun 旬 and the image of the potter’s wheel, it is also possible that xun 詢 and xun 訊, “to ask,” belong to the same word family, both words being derived ultimately from yun 允 (OC *ljuәnʔ). Schuessler (2007, 539) equates xun 訊 and xin 信 (OC *sjin), joining these in the same word family as xun 恂／洵 (OC *sәwjin) and stating that these words are “probably derived from yun 允 (OC *win), ‘trust, be true, sincere.’” Although xun 詢 is not included in Schuessler 2007, Schuessler offers that xun 詢 may be the same word as xun 恂／洵 in Schuessler 1987, 705; for early examples of interchangeable use, see Gao 1989, 80. 74 Zhouli zhengyi 11.2766–2776; cf. 2841. 75 Schuessler 1987, 705. For xun 循 in the circulation of speech, see Guoyu jijie 188 (“Lu yu” 2).

THE ZHOULI AS CONSTITUTIONAL TEXT

57

an emphasis on getting at the truth of a matter, as for instance by means of direct inspection.76 The grandest of the terms for traveling and communicating is xun 巡 (OC *swjin), which in its most elevated usage denotes an inspection tour undertaken by an official or by the sovereign himself.77 Some Zuo zhuan attestations of xun 巡 suggest a definition very close to that of xun 徇, with the significant difference that, when xun 巡 involves the delivery of speech, it is always the ruler himself who goes or sends a personal proxy to the commoners to speak to them.78 In the Zhouli, too, xun 巡 can entail the delivery of the sovereign’s speech to his subjects: “The herald is responsible for reciting the aims of the king and speaking of the administrative affairs of the capital city. He makes a tour of all the dependent states and discourses on these things” (撢人 掌誦王志，道國之政事，以巡天下之邦國而語之). 79 And again, after the description of the idealized concentric geography of the realm: “When the king is about to make a tour of the holdings, then [the keeper of maps, zhifangshi 職方氏] makes admonition in the four quarters, saying: ‘Each of you, cultivate and make tranquil your holdings, perfect your appointed undertakings, and dare not fail to be respectful and admonished in them, since then for the state there will be a great penalty’” (王將巡守，則戒于四方，曰：「各脩平乃守，考 乃職事，無敢不敬戒，國有大刑。」).80 This xun 巡 can also be an occasion of royal largesse.81 It may be that, for the architects of the Zhouli, a king’s or official’s inspection tour, xun 巡, was conceptually linked with the consultation implied in xun 詢, in xun 訊, in xun 徇, and in xun 循 and was thought of as the most exalted way of confronting and incorporating the views of the governed populace. The turning potter’s wheel makes a fine 76

Schuessler (1987, 801) relates yun 允 to xun 循; see n. 73 above for the possible derivation of xun 訊 from yun 允. Xun 循 is attested as a loan for xun 恂, “be sincere; certainly,” one of the graphs that, according to Schuessler (1987, 705), may represent the same word as xun 詢; see Gao Heng 1989, 132. 77 Xun 巡 is attested as a loan for xun 徇 (Gao Heng 1989, 80); and xun 循 is attested as a loan for xun 巡 (Gao Heng 1989, 80). 78 See Chunqiu Zuozhuan zhu 682, 1428, 1616 (Xuan 4.3, Zhao 21.6, Ai 2.3). 79 Zhouli zhengyi 10.2704. 80 Zhouli zhengyi 10.2692. 81 Zhouli zhengyi 4.1023. See Shiji 30.1441; Guanzi jiaoshi, 639 (“Qingzhong ji” 輕重己). I understand the phrase xun yu baixing 循於百姓 to refer to a tour of the sort normally called xun 巡. See also Zhouli zhengyi 3.834, which brings together the elements of the inspection tour, the giving of largesse, and the use of the woodentongued bell.

58

CHAPTER ONE

figure for the circulation around a center that is required to achieve a “comprehensive” (zhou) view of all that surrounds that center. In the physical act of going around to the people he governs, the king (or his proxy) descends across social and geographical space and marks out the extent of his sway. On his return to the center, his new knowledge of local circumstances becomes useful not only in its many particulars but also in the aggregate, as the court considers and classifies circumstances throughout the realm. The fundamental model for the court as governing institution is the treasury as a destination for resources or, even more pertinently, the market as a central place for assessing value and trading. By such communicative actions as consultation and travel, the king and his officers reinforce the networks necessary for effective taxation and redistribution. Conclusion: “Universal Knowing” and the Intellectual Context of the Zhouli Taken together, the mechanisms for aggregating and using resources of all kinds are the basis of the Zhouli’s proposed constitution; their role in the text is such that it may require a reinterpretation of the title itself. What makes the government a government, and the center a center, is the whole set of transactions that make of the capital a hub. The figure of the complete circle, seen in the word family of xun 詢, reappears in the term zhou 周, which within the Zhouli is for the most part used in the compound zhou zhi 周知, “to know” some phenomenon “as it occurs on all sides”: “universal knowing,” one might call it. This use of zhou accounts for nearly all appearances of the word in the text. Significantly, both in the Zhouli and in the few other early texts that use the compound zhou zhi, the object of knowledge is in all but a few cases something of relevance to economic resources and taxation.82 Clerks in the accountant’s department (sishu 司書), for example, are versed in all tax regulations and maps, and by these they “know all around them the income and expenditure of the hundred different types of tribute good” (周知入出百物); they are responsible for tax receipts, which they categorize, record, and turn over to other

82 For zhou zhi outside the Zhouli, see Yi Zhoushu huijiao jizhu 2.1042, 1061 (“Zhifang jie”; cf. Zhouli zhengyi 10.2636, 10.2684, for close parallels); Shuoyuan jiaozheng 511 (“Fan zhi” 反質); Kongzi jiayu zhuzi suoyin 25 (“Zhipei” 執轡).

THE ZHOULI AS CONSTITUTIONAL TEXT

59

officers.83 In two places, this universal knowing denotes something grander: it is the knowledge that the keeper of maps has of the realm’s idealized geography; and it is the knowledge that the king obtains by reading five types of written reports (shu 書) on conditions in the various states, submitted by the chief assistant to the director of the Bureau of Foreign Affairs (xiao xingren 小 行 人 ). 84 The “comprehensiveness” that Jin Chunfang read into the zhou of the title Zhouguan is not the knowing of a single thing from all sides but—by the internal criteria of the text—the knowing of a thing as it comes to the center from all sides, that is, as it exists in and comes from all parts of the realm. In this interpretation, the “offices” of the Zhouguan or the “rites” of the Zhouli are not an object for general contemplation throughout the realm but tools for central administrative knowledge of the realm. Any association with the dynasty named Zhou is accidental. This investigation has provided an intellectual historical context for reading the Zhouli as a comprehensive system for knowing and governing. As an office list, the Zhouli resembled other, shorter lists that appeared in the late Warring States and early Han. These lists, which formed a minor genre of composition, bespoke a drive to create a complete overview of the functions necessary to a successful government. As constitutions, they determined the distribution of political power in the state and in this respect resembled the kind of constitution, or politeia, that Greek philosophers studied. Perhaps because of the long history and prestige of the command (ming), or perhaps because somewhere early in the history of writing in China there was a tradition of office lists, these Chinese constitutional texts took a form not found in early European literature. A list always presumes some principles of inclusion and distribution. In the case of office lists, these principles reflect some fundamental philosophical conception of the truest and most effective ways to govern the world. Among the shorter constitutional texts, the principles of organization vary widely. Some, like the Xunzi’s, are quite practically organized around the concentration of resources and power in the royal center. Others, like the Western Han examples, are dominated by the relatively abstract categories of correlative cosmology. 83 See Zhouli zhengyi 2.478. See also Zhouli zhengyi 2.479, 3.689, 3.779, 4.1144, 4.1236, 10.2636. At Zhouli zhengyi 9.2408, zhou zhi denotes a militarily useful knowledge of surrounding terrain; at 10.2455, a quantitative knowledge of the numbers of locales and the noblemen in them. 84 Zhouli zhengyi 10.2684, 12.3007.

60

CHAPTER ONE

These texts throw into relief the Zhouli’s elegant combination of a workable, detailed list of offices and duties with a higher-order classification system. They also point up the Zhouli’s unique elaboration of ideas of controlled circulation: the central government draws to itself material and informational resources and then reproduces its power as it redistributes these resources across the realm in the form of commands and public works. Other constitutional texts do prefigure the Zhouli’s implicit claim to be a complete listing of the officers required for a functioning royal government, but none achieves the intellectual coherence that comes with the Zhouli’s vision of centralization, universal knowing, and management of information and resources. If the Zhouli is indeed the work of predynastic or dynastic Qin scholars who drew earlier ideas on administration and statecraft into a new systematic vision, then its idealization of universal knowing follows logically. The Qin’s many projects of unification—of script, of measures, and of the legal system itself—belonged to an overall effort to make the realm universally knowable and its resources exploitable. The image of the ruler who can know all that occurs under Heaven without so much as venturing out his door is one that was first expressed in the Laozi 老子, then assimilated in the Han Feizi to a vision of government by inflexible rule of law.85 It appears also in the Lüshi chunqiu, a work closely associated with the ruling elite of predynastic and dynastic Qin, and in this work it is matched to an ideal of government that is far less harsh and far less dependent upon punishment than that found in the Shangjun shu or the Han Feizi.86 In all of these texts, the system of officials and duties, functioning according to xingming ideals of prescription and performance, promises the ruler security, quiescence, and insulation from the day-to-day anxieties of governing. This model of the ruler’s role recalls both the Zhouli constitution, which largely frees the king from day-to-day decision making, and the final years of Qin Shihuang’s life, when the emperor is said to have withdrawn into tranquil seclusion in an effort to become a “true man” (zhenren 真人).87 The ruler who can know all without leaving his palace may nevertheless choose to travel across his realm and show himself to his subjects. The tours of inspection undertaken by the First Emperor of Qin 85

Laozi jiaoshi 189 (Laozi 47); Han Feizi jishi 409–410 (“Yu Lao” 喻老). Lüshi chunqiu jiaoshi 1.145 (“Xian ji” 先己), 2.1049 (“Jun shou” 君守). 87 Shiji 6.257. 86

THE ZHOULI AS CONSTITUTIONAL TEXT

61

had the functions of circulation and territory marking attributed to xun 巡 in the Zhouli.88 The stelae erected in the course of these tours, along with the possible performance of the texts inscribed upon them, were not only an address to spiritual powers but a display to the governed populace that their sovereign had acquired the power to circulate everywhere among them and to speak to them and for them.89 It is not surprising to find that in two of the stela inscriptions, zhou functions precisely as it does in the compound zhou zhi, describing the stabilizing (zhou ding 周定) and surveying (zhou lan 周覽) that the sovereign performs all around him as he moves throughout the realm.90 These depictions of the ruler’s movement recall the enduring topos of the powerful man’s roaming flight (zhou liu 周流), exemplified both in the Lyrics of Chu (Chuci 楚辭) and in the Han fu 賦.91 If we are to assess how the Zhouli, with its apparent associations with Qin institutions, would have been received during the earliest stages of its post-Qin transmission, then we must understand the work in the context of the ideas on statecraft, self-cultivation, and rulership that were prevalent in the early Han. These ideas, often collectively associated with the figures of the Yellow Emperor and Laozi and referred to as “Huang-Lao” 黃老 thought, were represented in a surprisingly large number of Western Han texts and seem to have attracted several powerful second-century figures, including emperors and empresses.92 To judge from extant second-century BCE texts that have been identified with Huang-Lao, this configuration of ideas would have offered a singularly receptive matrix for the Zhouli. Adopting as a sample of Huang-Lao thought the four texts on statecraft found at Mawangdui in 1973 and sometimes identified as the Four Political Treatises of the Yellow Emperor (Huangdi sijing 黃帝四經), it is easy to see the appeal that the Zhouli would have held for anyone interested in these theories. The Mawangdui texts idealize the ruler who is both 88

Shihuang’s tours are referred to as xun 巡 at Shiji 6.241, 243, 252, 261, 267. The Zhouli’s prescriptions for xun 巡 may thus corroborate the interpretation of the tours offered by Kern (2000b, 114–118). 90 Shiji 6.249, 261. 91 For zhou liu, see Chuci buzhu 1.32, 42, 43 (“Li sao” 離騷), 3.110 (“Tian wen” 天問), passim; Shiji 117.3026. For zhou lan, see Shiji 117.3025. 92 Emmerich (1995) gives a careful accounting of the Han texts likely to have been associated with Huang-Lao thought and provides biographies of the major figures said to have practiced Huang-Lao ideas. With regard to the possible existence of a selfidentified Huang-Lao “school” practicing a distinctive and internally coherent set of ideas, I share the skepticism expressed by Csikszentmihalyi and Nylan (2003, 80–87). 89

62

CHAPTER ONE

quiescent and concerned for his people; in this respect they differ from the Shangjun shu and Han Feizi and resemble the Zhouli.93 The ruler of the Mawangdui texts, like the ruler of the Zhouli, uses the tools of his government, particularly the laws and administrative regulations, to know all that occurs in the realm; for any adherent of xingming theory in the abstract, the Zhouli’s office list would have represented a practical guide to the “names and forms” of a complete government.94 Finally, in their attention to timeliness, the Mawangdui texts raise the possibility that the distinctive arrangement of the Zhouli office list according to departments (Heaven, Earth, and the four seasons) would have resonated with Huang-Lao ideas.95 To be sure, the Zhouli government differs in certain key respects from the system adumbrated in the Mawangdui texts; its economic policies, for instance, are far more interventionist. But it is true too that within the diverse field of texts that were known in the Western Han and that have been linked to Huang-Lao, more robust proposals for government economic guidance were not unknown.96 Whether it was regarded as a Huang-Lao text, as a product of Qin expertise, or as a model from an even earlier time, the Zhouli came into the early Han as a technical guide to the xing and ming of government, and specifically to a government built upon “universal knowing” and the controlled exchange of information and resources. This reception set the stage for the gradual reframing of the office list (the Zhouguan, or “comprehensive offices”) as historicized ritual text (the Zhouli, or “rituals of Zhou”). As read from the perspective of the 93

See Leo Chang and Yu Feng 1998, 52–56, 108–113. Leo Chang and Yu Feng (1998, 48–52) note the central importance of xingming in the model of governance advanced in the opening sections of the “Jing fa” 經法, 100. The work of John Major on cosmological and topographical knowledge in the Huainanzi suggests that that text itself is in some ways a Huang-Lao-affiliated descendant of the Zhouli, sharing some of its themes (Major 1993, 165) and its overall conviction that “knowledge of the natural world translates into political power” (13). 95 Leo Chang and Yu Feng 1998, 65–67, 120–122. Although there is in the Lüshi chunqiu a Qin precedent for calendrical ordering of court duties, it is nevertheless conceivable that the six-part arrangement of the Zhouli text reflects, not an original Qin arrangement, but the work of Han transmitters interested in calendrical correlations. 96 For the Mawangdui texts’ views on economic policy, see Leo Chang and Yu Feng 1998, 111–112; Peerenboom 1993, 99–100. Kanaya Osamu and others have identified the Guanzi 管子 with Huang-Lao thought, though it remains uncertain whether that text’s most striking chapters on economic intervention would have fit with the laissez-faire ideals found in other texts; see Rickett 1998, 8; Peerenboom 1993, 234–236. 94

THE ZHOULI AS CONSTITUTIONAL TEXT

63

Eastern Han and afterward, the Zhouguan or the Zhouli was the set of “offices” or “rituals” that could be traced to the beginning of the Western Zhou, and specifically to one founding figure, the Duke of Zhou. Especially as the duke came to be associated with Zhou culture and with certain of the Classics, this reassignment kept the text usable, lending to its prestige and making it a respectable inspiration for later reformist-minded thinkers. The scant tincture of ritual guidance found in the text provided a sort of alibi for its xingming formulas. The hallowed name of the Zhou dynasty and the title of its early regent came to overshadow the sense of “comprehensiveness” that had been signaled both in the original title of the work and in its prescriptions for rule. Nonetheless, the qualities that many later thinkers found worthy of imitation in the text—its pairing of certain of the virtues lauded by Confucius and his followers with hardheaded technical prescriptions for government—derive from the sense of zhou as exemplified in the text’s use of zhou zhi. The legitimacy of a monarchy is normally understood to rest upon tradition, often bolstered by an overt claim to divine sanction. In China, the rulers of the early Western Zhou dynasty had long been represented as defenders of the governed people’s interests, both in overthrowing an abusive regime and more generally in consulting (xun 詢) even the humblest groups. The monarchy envisioned by the Zhouli presumes these two old forms of legitimation but adds to them a newer form, a rationalized, technical vision of the total management of the resources needed to sustain the realm. While this technical vision owes much to late Warring States philosophy and to the practical administrative techniques of the Qin and other states, it does not exclude older, more traditional techniques. Instead, it incorporates them as one kind of resource, extracting from them the value of their traditional prestige. The procedure seems to have been typical of Qin intellectual endeavors. To the great extent that later Chinese administrative structures are inspired by the model of the Zhouli, they owe an unacknowledged debt to the Qin state and its principles. Like the early Greek constitutions with which it might be compared, the Zhouli has its absent legislator, the Duke of Zhou; but behind him, not quite obscured by the legends, there is another far more influential Solon, the Qin.

CHAPTER TWO

OFFICES OF WRITING AND READING IN THE RITUALS OF ZHOU Martin Kern, Princeton University The Rituals of Zhou in the Context of Formulaic Technical Writing Of all known early Chinese texts, the Rituals of Zhou (Zhouli 周禮)— in Han times initially called the Offices of Zhou (Zhouguan 周官)— comes closest to a survey of bureaucratic order. Other texts offer meticulous accounts of specific formalized activities (e.g., divination records) or stage-scripts for ritual performances (e.g., the Ceremonial Rituals [Yili 儀禮]). But none rivals the encyclopedic scope of the Rituals of Zhou as it lays out a vast and systematic hierarchy of governmental functions, no matter how historically accurate or idealized we may assume them to be. As such, the Rituals of Zhou is the perfect textual reflection of the bureaucratic idea itself, an idea that in early China resounded with strong religious overtones and manifested itself in administrative documents as well as in political rhetoric and religious communication.1 The master tool of bureaucratic rule is formulaic writing; when in Han times the Rituals of Zhou rose from obscurity, such writing had continuously existed for at least a millennium, from late Shang (ca. 1200–ca. 1046 BCE) oracle records through Western (ca. 1046–771 BCE) and Eastern Zhou (770–256 BCE) bronze inscriptions and texts

In addition to the participants at the workshops and conference, I wish to thank Lothar von Falkenhausen for his, as always, excellent and helpful comments on an earlier version of this essay. 1 The literature on the nexus of Chinese religion and bureaucracy, ancient and modern, is extensive; see, e.g., Keightley 1978; Xu Zhongshu 1936; Vandermeersch 1977–1980; Falkenhausen 1993, 152–167; C. Cook 1995; Venture 2002; Kern 2007; Lewis 1999b, 13–51; Schipper 1974; Seidel 1985; Weld 1990; Harper 1994, 2004; Li Ling 2000; Ahern 1981. Lewis (1999b, 42–48) has argued the point explicitly for the Rituals of Zhou.

OFFICES OF WRITING AND READING

65

on stone, bamboo, and wood.2 Furthermore, for the fourth and third centuries BCE in particular, numerous excavated manuscripts display formulaic writing for a variety of bureaucratic purposes, and regularly in close association with religious ones: administrative records, tax records, calendars, archival records, judicial records and legal statutes (including contracts for the netherworld), economic records, divination records, sacrificial records, etc. Among writings found in tombs, by far the most common type is the inventory of funerary items, that is, a simple list that occasionally was accompanied by a letter to the netherworld bureaucracy.3 While the hallowed texts of the literary and philosophical tradition existed in both writing and—primarily—in the learned elite’s oral memory, the texts truly dependent on the technology of writing were pragmatic ones that recorded, stored, or circulated information or provided the blueprint for specific bureaucratic or religious tasks.4 Certain types of administrative writing were based on fixed forms where one only had to fill in the specific information (e.g., dates and names), and where the actual demands on the scribe were quite limited.5 Altogether, formulaic writing served the required uniformity of administrative, legal, and economic records that could be produced and reproduced in large numbers. Given its occasion-specific function, most of this writing was not preserved by the tradition. Written on wood and bamboo stationery, it has perished from view, yet circumstantial evidence still proves its existence already for late Shang and early Western Zhou times.6 Writing clearly reached a fundamentally new status with the early empire, but this status pertained not so much to bureaucratic records (which just multiplied with the needs of the imperial state) as to the developing commitment to enshrine the hallowed tradition— especially the Five Classics (wujing 五經) and the various genres of commentaries and essays arranged around it—in a fixed written form.7 To write and read these texts properly required a profound education 2 In addition to the works listed in n. 1, see Bagley 2004; Xu Fuchang 1993; Barbieri-Low 2007, Xi Hanjing 1983; Chen Hanping 1986. Ledderose (2000) discusses “modular mass production” across a broad range of artifacts, including texts, through the Chinese tradition. 3 For a survey of such writings, see Giele 2001. 4 See the discussion in Assmann 2000, 131–138. 5 Xing Yitian 1998; Loewe 1967, vol. 1, 16. 6 Bagley 2004; Falkenhausen 1993, 163–164; Li Feng 2006. 7 Kern 2001.

66

CHAPTER TWO

not only in the Classics themselves but also in the forms of political, ritual, philosophical, and historical discourse that were seen as direct extensions of the canon. Especially in the archaic Classics of Poetry (Shijing 詩經), Documents (Shangshu 尚書), and Changes (Yijing 易 經), the frequent use of “loan characters” (jiajie zi 假借字) to write homophonous but different words necessitated more than the knowledge of the script: in order to read and understand the text, and to identify the correct words behind the ever-changing ways of writing them, one had to know it already. By contrast, pragmatic texts could be readily written and comprehended in this script for the simple fact that blueprinted administrative records were largely unambiguous in a way archaic poetry and speech were not; their straightforward diction did not require any hermeneutic approach or particular cultural competence. A subordinate scribe merely trained in the writing system was well equipped to produce and read such records but would have felt disoriented in the archaic language of the classical tradition. The Rituals of Zhou does not fit neatly into the category of either pragmatic writing or the Classics. In Han times, it became associated with the Duke of Zhou 周公 (r. 1042–1036 BCE) and gained much in stature through Zheng Xuan’s 鄭玄 (127–200) commentary, which connected the text with the established Classics. While not yet included in the inscription of the Classics on stone stelae that were erected outside the Imperial Academy in 175 CE,8 it was a prominent text by late Six Dynasties times, received another major commentary by Jia Gongyan 賈公彥 (fl. 627–656), and became part of the nine (later twelve and finally thirteen) Classics during the Tang. Yet, its later career as a Classic notwithstanding, the text seems far closer to the recently excavated bureaucratic texts than to the high tradition of literary, historical, and philosophical works that in late Warring States and early imperial times formed the traditional canon. By virtue of its own structure as a vast and highly systematic (and, one may say, rather dull) survey, its existence outside the neat columns of a bundle of knotted bamboo slips is hard to imagine. As a material artifact, the Rituals of Zhou embodied the very nature and pervasive function of bureaucratic writing that it consistently asserted for its universe of governmental offices. Unlike the transmitted poetic, historical, and philosophical literature of Warring States and Han times, the Rituals of Zhou nowhere 8

Wilkinson 1998, 464; Boltz 1993, 30.

OFFICES OF WRITING AND READING

67

advances an explicit argument. It poses no hermeneutical challenge, contains no direct historical references or textual citations—even though certain parts of the text may have been composed with information gained from archaic texts such as the Classic of Poetry (Shijing 詩經) or the Classic of Documents (Shangshu 尚書)—and makes no claim as to its authorship or circumstances of composition. It also is not invoked by other texts before late Western Han times. As a result, scholars have variously assigned the text to the Duke of Zhou, to anonymous compilers of late Warring States or Qin-Han times, or, finally, to Liu Xin 劉歆 (46 BCE–23 CE), allegedly to serve the political purposes of the “usurper” Wang Mang 王莽 (r. 9–23).9 I do not agree that the Rituals of Zhou depicts “a governing system which, in all essentials, prevailed in middle and late feudal Chou in the various states and has its roots in the system pertaining to late Yin and early Chou,” from which time it “continued to exist in a more or less modified form up to 221 B.C.”10 To my mind, the Rituals of Zhou fits well with late Warring States or imperial Qin textual culture (the latter being argued in Jin Chunfeng’s 金春峰 superb study of 1993) on three grounds: linguistically, philosophically, and as an exemplar of formulaic writing.11 Yet its idealized image of perfect administrative order meets the intellectual needs of the dawning (Warring States) or newly established (Qin–early Han) empire just as well as those of an imperial rule in collapse (late Western Han). Within this range of some three hundred years, it seems difficult—pending the next archaeological sensation—to put a precise date on its composition or editorship. With the present essay, I will, however, suggest that the Rituals of Zhou is a text of at least two very distinct layers: composed in Warring States/early imperial language, it conforms to early imperial ideals of universal order12 and matches accounts of scribal culture of that time. Meanwhile, it also reveals a profound knowledge of far older— indeed, Western Zhou—administrative units and their titles that by 9

For a survey of these opinions, see Boltz 1993, 24–29; see also the discussions in Karlgren 1931; Jin Chunfeng 1993; Zhou Shifu and Zhou Wenxiang 1981; Qian Xuan 1996, 21–33; among others. 10 Broman 1961, 73–74. 11 See also David Schaberg’s superior discussion of the possible date of composition in chapter 1 of this volume. 12 Lewis (1999b, 48) understands the Rituals of Zhou as a comprehensive vision of “the state as a replica or image of the cosmos.” While any numerologically driven order, one might argue, is inherently cosmological, I am reluctant to see the text in these terms. It is not how it was discussed during its early reception.

68

CHAPTER TWO

Warring States times had long been discontinued and replaced.13 As such, the text systematizes and seemingly unifies two disparate and diachronic sets of administrative data. The Low-Ranking Scribes in the Rituals of Zhou Among the offices listed in the Rituals of Zhou, one would expect many to be devoted to the composition, circulation, and reception of formulaic written documents as key to bureaucratic rule. The actual number of office descriptions that explicitly mention the reading, writing, and handling of written documents—42 out of 366 offices— might therefore appear low. However, it was not only these 42 offices that produced and kept writings. The vast majority did. This fact is not expressed in the actual office descriptions but in the general “outlines of offices” (xu guan 敘官) that form the introductions to the six major sections of the text. In these outlines, all the positions within each office are listed in hierarchical order, beginning with the head of the office and ending with those in charge of the lowest duties. A comprehensive example is the outline for the prime minister (taizai 大宰), head of the Ministry of State (tianguan 天官). This highest office of the state, placed at the outset of the Rituals of Zhou, contains the following positions and ranks: Prime minister (taizai), one man in the rank of one of the six ministers (qing); vice prime ministers (xiaozai), two men in the rank of ordinary grand master (zhongdafu); assistant ministers of state (zaifu), four men in the rank of junior grand master (xiadafu); senior servicemen (shangshi), eight men; ordinary servicemen (zhongshi), sixteen men; numerous junior servicemen (lüxiashi), thirty-six men; storehouse keepers (fu), six men; scribes (shi), twelve men; aides (xu), twelve men; runners (tu), one hundred and twenty men. 大宰卿一人。小宰中大夫二人。宰夫下大夫四人。上士八人。中 士十有六人。旅下士三十有二人。府六人。史十有二人。胥十有 二人。徒百有二十人。

13

As Broman (1961, 66–74) has argued, the offices in the Rituals of Zhou coincide to a very substantial extent with those known from other early texts. Yet more to the point, recent Chinese scholarship on Western Zhou administration has pointed out numerous correspondences between the Rituals of Zhou and Western Zhou bronze inscriptions; see Xi Hanjing 1983; Lai Changyang and Liu Xiang 1985; Zhang Yachu and Liu Yu 1986.

OFFICES OF WRITING AND READING

69

The second office listed under the Ministry of State, that of the commandants of the palace (gongzheng 宮正), contains the following positions: two senior servicemen, four ordinary servicemen, eight junior servicemen, two storehouse keepers, four scribes, four aides, and forty runners. The third office, master of the palace guard (gongbo 宮 伯), is staffed with two ordinary servicemen, four junior servicemen, one storehouse keeper, two scribes, two aides, and twenty runners. The fourth office, food supervisor (shanfu 膳夫), has two senior servicemen, four ordinary servicemen, eight junior servicemen, two storehouse keepers, four scribes, twelve aides, and one hundred and twenty runners. These examples from the beginning of the text show the overall, largely unified pattern representative of much of the Rituals of Zhou: while the office of the prime minister contained at its top the three high-ranking positions of one minister (qing, Rank 1) as prime minister, two ordinary grand masters (zhongdafu, Rank 3) as vice prime ministers, and four junior grand masters (xiadafu, Rank 4) as assistant ministers, the lower offices were merely staffed with personnel of the rank of senior servicemen (shangshi) and below. The sequence of positions in these offices is consistent and reflects their rank: the three levels of servicemen (shi) were still part of the nobility, but the four ranks—in this order—of storehouse keepers, scribes, aides, and runners were not; they were recruited from commoners. Serving as messengers, the runners were the lowest level of the personnel, and their members were usually the most numerous. According to Zheng Xuan’s commentary, the aides above them are characterized as men possessing “talent and knowledge” (you cai zhi 有才智).14 In a few cases, other groups are listed between the scribes and the aides, such as artisans (gong 工) and merchants (gu 賈), as in the offices of the storehouse keeper of jades (yufu 玉府) and supervisor of palace women’s work (dianfugong 典婦功).15 In this outline, the storehouse keepers are primarily in charge of storing official documents and contracts;16 the scribes, usually double the number of storehouse keepers, were their subordinates who created these writings. Zheng Xuan notes that both groups were appointed by the respective ministers, suggesting their relatively low status. In his explanation of the phrase “com14

Zhouli zhengyi 1.21. Zhouli zhengyi 1.40, 54. 16 See the Shuowen jiezi 說文解字 definition of fu as wenshu zang 文書藏 (storage of writings) as quoted by Sun Yirang 孫詒讓 (1848–1908) in Zhouli zhengyi 1.20. 15

70

CHAPTER TWO

moners in office” (shuren zai guan 庶人在官) in the “Royal Regulations” (“Wangzhi” 王制) chapter of the Records of Ritual (Liji 禮記), Zheng refers to the four Rituals of Zhou categories of storehouse keepers, scribes, aides, and runners, noting again that they were appointed by the ministers and adding that they did not receive orders (or writs of appointment?) directly from the Son of Heaven or the ruler of the state (bu ming yu tianzi, guojun 不命於天子國君).17 Altogether, the outlines to the five original chapters of the Rituals of Zhou—no scribes are mentioned in the “Artificer’s Record” (“Kaogong ji” 考工記)—a chapter that was included with the text only in Han times) list for the regular offices of the central government 442 storehouse keepers and 994 scribes, all of whom are practically invisible in the actual office descriptions. For both groups, additional numbers are given for administrative units beyond the central government (in the case of the scribes, an additional 101).18 While the outlines seem too schematic to reflect the actual administrative needs and practices of any particular period, the large number of officials charged with the production and storage of documents still points to an extensive amount and broad variety of pragmatic writing. This finding accords well with the archaeological evidence. Already for Warring States times, the reality of extensive pragmatic writing— implying large numbers of scribes—is well supported by excavated manuscripts. For earlier periods, the evidence is indirect yet not to be dismissed: the archives that underlay the bronze inscriptions required dedicated scribes who were most likely separate from the specialists who carved the inscription texts into the clay molds used for bronze casting. Considering just the social, technological, and economic implications of the production of thousands—likely tens of thousands— of bronze artifacts, one can easily imagine a substantial group of scribes in charge of records, even though the actual number may not have reached the dimensions noted in the outline sections of the Rituals of Zhou. This is not to suggest a particularly early date for the composition of the Rituals of Zhou or for its temporal frame of reference. Leaving aside such considerations of dating, direct archaeological evidence for the status of governmental scribes comes from the Qin statutes, writ17

See Zhouli zhengyi 1.20; Liji zhengyi 11.94c. The later commentarial tradition has repeatedly applied Zheng’s Records of Ritual commentary to the Rituals of Zhou. 18 All numbers follow Sun Yirang’s calculations at the end of each outline in the Zhouli zhengyi.

OFFICES OF WRITING AND READING

71

ten on bamboo, that were excavated in 1975 from Shuihudi 睡虎地 (Yunmeng 雲夢, Hubei) tomb 11, which was sealed in 217 BCE. The tomb occupant, a man named Xi 喜 who lived from 262 to 217 BCE, began his governmental career as a local scribe (shi 史) at the age of eighteen or nineteen (244 BCE) and was promoted to the position of a prefectural scribe (lingshi 令史) three years later. He was appointed to the same position in a different district in the following year. As a prefectural scribe—still a rather low and junior position19—he had more extended responsibilities, including the investigation of criminal cases.20 In 235 BCE, he was promoted to be in charge of criminal suits.21 From this career, it becomes clear that the position of the scribe, although hereditary in Qin, was an entry position and could be attained at a young age immediately after being trained in an office that the Qin statutes identify as “study room” (xueshi 學室).22 The principal prerequisite was the ability to read and recite a text of a certain length—a skill particularly important in legal matters. There is further evidence, now from the Han, concerning the competence of governmental scribes. According to Xu Shen’s 許慎 (ca. 55–ca. 149) postface to the Shuowen jiezi, his dictionary contained a total of 9,353 characters, plus 1,163 variant forms.23 He also informs us that the character glossary Cang Jie 蒼頡, attributed to the Qin chancellor Li Si 李斯 (d. 208 BCE) and further elaborated upon by Yang Xiong 揚雄 (53 BCE–18 CE), included 5,340 characters.24 According to Qiu Xigui, Shang oracle bone inscriptions include fewer than 5,000 different characters, and the Thirteen Classics contain a total of 6,544 individual characters.25 As Qiu points out, the Thirteen Classics developed over a long period of time and accordingly include characters that were in use at different periods. Thus, “if it were possible to calculate the statistics of a period limited to one or two hundred years of the Zhou dynasty, the number of characters in general use 19

Xu Fuchang 1993, 378–382. Hulsewé 1985, 39n4; see also the various entries on lingshi as keyed in Hulsewé’s index. 21 See Xu Fuchang 1993, 8–14; Hulsewé 1985, 1. 22 Xu Fuchang 1993, 358–360; Hulsewé 1985, 87. 23 Shuowen jiezi zhu 15B.1a. As noted by Qiu Xigui (2000, 48n10) and others before him, the actual total number of characters (including variants) is slightly higher, namely, more than 10,700. This is probably to be explained by later additions. 24 Shuowen jiezi zhu 15A.14b. Yang Xiong’s Cang Jie xunzuan 蒼頡訓纂 was apparently a glossary to complement the earlier dictionary; see Hanshu 30.1720–1721. 25 Qiu 2000, 49–50. 20

72

CHAPTER TWO

during that period would probably fall short of the total number used in the Thirteen Classics.”26 By way of comparison, Qiu further notes that the modern dictionary Xiandai Hanyu cidian 現代漢語辭典 includes about 4,900 different characters. This broader perspective is relevant to a passage in the Shuowen jiezi postface—with an earlier version already included in the History of the Han (Hanshu 漢書) “Monograph on Arts and Writings” (“Yiwen zhi” 藝文志)—that is often read as saying that a Han scribe had to memorize 9,000 different characters.27 The passage in the Shuowen jiezi reads as follows: “at the age of seventeen or above, the young students are first examined, and those who can recite and write nine thousand characters can serve as scribes” (學僮十七以上始試，諷籀 書九千字乃得為史). Here, the Qing commentator Duan Yucai 段玉 裁 (1735–1815) glosses zhou 籀 as “to read out loud a written text” (dushu 讀書). Earlier commentators have understood zhou as referring to an early Zhou dictionary, still listed in the Hanshu “Monograph on Arts and Writings”: the Scribe Zhou (Shi Zhou pian 史籀篇) in fifteen bundles.28 The parallel version of the Shuowen passage in the Hanshu does not include the character zhou: “the grand scribe examines the young students, and those who can recite and write more than nine thousand characters can serve as scribes” (太史試學僮，能諷書九千 字以上乃得為史).29 A third version of the passage, showing it in its original legal context, is now available in the “Statute on Scribes” (“Shilü” 史律) in the Statutes and Ordinances of the Second Year (Ernian lüling 二年律令, likely referring to the year 186 BCE), a bamboo manuscript from tomb 247 at Zhangjiashan 張家山 (Jingzhou 荊州, Hubei). This text reads: “[One examines] the young students for scribal office by using fifteen bundles of bamboo slips [of text], and those who can recite and write more than five thousand characters can serve as scribes” ([試] 史學僮以十五篇，能諷書五千字以上乃得為 史). Here, the modern manuscript editors explain shiwu pian 十五篇 (fifteen bundles of bamboo slips) as a reference to the Shi Zhou pian 史籀篇, which in the Hanshu happens to be listed as of just that size.30 26

Qiu 2000, 50. See, e.g., Winter 1998, 566; Wilkinson 1998, 48n29. 28 For the Shuowen passage and its commentaries, see Shuowen jiezi zhu 15A.11b– 12a; for the Hanshu mention of the Shi Zhou pian, see Hanshu 30.1719. 29 Hanshu 30.1721. 30 Zhangjiashan ersiqi hao Han mu zhujian zhengli xiaozu 2001, 203. I find this conclusion too tenuous to support. 27

OFFICES OF WRITING AND READING

73

Like the Shuowen, the Zhangjiashan manuscript includes a reference to the students’ age of seventeen sui 歲 before they are examined for the position of a scribe; thus, it places the examination about two years before the age at which the Shuihudi tomb occupant was appointed to his position, strongly suggesting that the two manuscripts from Zhangjiashan and Shuihudi both refer to roughly the same entrylevel function of a local governmental scribe. Considering (a) the total numbers of characters included in the Shuowen, in the Qin-Han dictionary Cang Jie, or in the entire Thirteen Classics and (b) the rather simple formulaic tasks of an entry-level governmental scribe, it seems extremely unlikely that he had to master 9,000 (or 5,000, as in the Zhangjiashan manuscript)31 individual characters. As Michael Loewe has noted with respect to the Han administrative wooden slips from Juyan (Edsen-gol; Inner Mongolia/Gansu), “the majority of the inscriptions were probably made by low-grade clerks or officials, whose education had been scanty, and whose practice at writing may have been somewhat limited.”32 Anthony Hulsewé, writing almost fifty years ago, doubted the number of 9,000 individual characters as well, noting that even all Western Han dictionaries combined would not have reached such a number. Hulsewé therefore proposed to read the Shuowen and Hanshu accounts as referring to the memorization of writings that were altogether 9,000 characters long.33 Depending on the text, this understanding would reduce the number of individual characters drastically. If it were a group of primers, perhaps including texts like the Classic of Filial Piety (Xiaojing 孝經, ca. 1,800 characters) or parts of the Analects (Lunyu 論語, ca. 12,000 characters),34 the number of different graphs would come down to fewer than 2,000. However, Hulsewé draws attention to a fragment of the Han System of Offices (Han guanyi 漢官儀, attributed to Ying Shao 應劭 [ca. 140–before 204]), preserved only in the Comprehensive Statutes (Tongdian 通典) of 801, that states that, in Han times, a lingshi 令史 31 The textual difference between 5,000 and 9,000 remains difficult to explain. The numbers “five” and “nine” were clearly distinct not only in sound (ruling out an oral misunderstanding) but also in character shape (precluding a simple copyist’s error). 32 Loewe 1967, vol. 1,16. 33 Hulsewé 1959, 246–247. This understanding seems also implied by Thern (1966, 13), who translates the passage as “When male students of seventeen sui and older are tested for the first time, they must recite and write 9000 words before they can be appointed officials.” It is unclear what Thern means by “words.” 34 Numbers from Wilkinson 1998, 465.

74

CHAPTER TWO

mastering both the Cang Jie pian and the Shi Zhou pian could be appointed to the Magnolia (or Thoroughwort) Terrace (lantai 蘭臺), the imperial palace depository for charts and writings.35 Hulsewé therefore concludes that the reference to 9,000 characters may have meant different things at different times: a text of a certain length, the entire contents of the two dictionaries, or certain other works.36 However, the Han System of Offices discussion may not pertain to governmental scribes in general, and certainly not to the entry-level position of an ordinary local shi (as opposed to a curator of imperial documents) in particular. According to the very limited information available, Magnolia/Thoroughwort Terrace was the office where historical and other important records of the textual tradition were stored.37 It was an imperial institution in charge of forming and preserving historical and cultural memory, not an agency that generated day-to-day administrative documents for the local bureaucracy. This impression is well supported by the fact that in the first century CE, Ban Gu 班固 (32–92), son of the eminent official Ban Biao 班彪 (3–54) and soon to be the most accomplished scholar and writer of his time, was appointed to the Magnolia/Thoroughwort Terrace for the express purpose of completing his history of the Western Han.38 I therefore agree with Kenneth Brashier’s interpretation of the Zhangjiashan passage as indicating that a text of a certain length had to be read, recited, and written.39 Whether or not this text was the Shi Zhou pian remains undecided. In the same Zhangjiashan “Statute on Scribes,” it is noted that diviners (bu 卜) had to master writings of 3,000 characters, but that some of them could recite more than 30,000; and an invocator (zhu 祝) was expected to recite 7,000. While the knowledge of 7,000 different characters would have been a highly unlikely demand for a low government position, the number of 30,000 characters cannot possibly be understood in this way, simply because there were not 30,000 different characters. This number can only refer to the length of a text, and given that the passage on the invocator is exactly parallel and directly related to those on the scribe and the diviner, its understanding should be extended to those as well.

35

See Hanshu buzhu 30.25a. Hulsewé 1959, 248. 37 See Hanshu 19A.725. 38 See Hou Hanshu 40A.1334. 39 Brashier 2003. 36

OFFICES OF WRITING AND READING

75

The words feng 諷 and zhou, both meaning “to read out loud,” point to a particular technical competence expected from a scribe. Why would a scribe be asked to recite a text? For a low-level government scribe, who was certainly not a member of the learned elite (as opposed to the scholars and high officials who memorized and recited the Classics), recitation was important because it demanded control of the phonetic correlation between graphs and the words they were writing. With the writing system not yet standardized and the use of loan characters for homophonous characters widespread, the scribe had to be able to resort to a phonetic use of the writing system as necessary, especially when encountering some less common words or in situations where a text was dictated to him. Such phonetic use of a limited number of loan characters for a much larger number of homophonous words was far more economical and easier to master than a large repertoire of different characters specific to particular words. As noted above, the low-level scribe in charge of administrative and other bureaucratic writing, while directly visible only in Warring States and later writings, was most likely already part of the Western Zhou royal court, even though probably in far lesser numbers. Indeed, the origins of the very word shi may be found in the kinds of administrative scribal activities mentioned in the Rituals of Zhou as well as in the Shuihudi and Zhangjiashan manuscripts. Morphologically, the word shi seems to have originally indicated someone who “marked” or “scratched” things and belongs to a group of early terms that according to Wolfgang Behr also includes shi 使 (to employ, send), li 吏 (clerk, official), li 理 (to divide, regulate, mark), shi 事 (to serve for), shi 仕 (to be a clerk, to serve as), and shi 士 (servant, retainer).40 As Behr points out, whether any of this was still on Xu Shen’s mind when glossing shi 史 as ji shi zhe 記事者 (one who notes down affairs) is doubtful;41 in typical Eastern Han fashion, Xu may have simply produced a pun. Be this as it may, it appears that the basic acts of mark40

See Behr 2005, 15–18. It should be noted that the morphological analysis of the word takes primacy over the paleographic interpretation of the graph, despite the latter’s prominence in scholarship since Wang Guowei’s seminal essay “Shi shi” 釋史 (Guantang jilin 6.1a–6b). Altogether, the literature on shi is vast; see the bibliographic information in Behr 2005, 15nn7–8; C. Cook 1995, 250–254; Gentz 2001, 9. A good range of graphic interpretations is included in Matsumaru and Takashima 1994, nos. 0004, 0024, 3371, 3425, 5881. For the fallacies involved in the interpretation of graphs to “decipher” the meaning of the words they are writing, see Boltz 1994; Takashima 2000. 41 Behr 2005, 18.

76

CHAPTER TWO

ing (including calculating) were at the original root of scribal identity.42 Furthermore, the clerical functions of day-to-day bureaucratic recording match well with the low status of the scribes in the Rituals of Zhou, where they are grouped with other unranked officials, among them artisans (gong). Of all transmitted texts from the Warring States and early imperial periods, it is the Rituals of Zhou—the very representation of bureaucratic order—that alone registers the rather pedestrian dimension of scribal activity for virtually all governmental offices, real or imagined. Records of simple administrative duties, abundant as they must have been, were discarded by the literary tradition. Apart from the Rituals of Zhou, only late Warring States and early imperial manuscripts like those from Shuihudi or Zhangjiashan speak explicitly of such lowlevel writing. That this kind of administrative writing derived from an originally religious framework, as has been repeatedly proposed,43 strikes me as problematic in two ways: First, it might neglect the obvious bias of the archaeological and received textual record as well as all indirect evidence for an early, concomitant use of writing in several distinct cultural spheres and on several distinct social levels. Second, it has the potential to exaggerate the status and power of writing in early Chinese culture by wrongly generalizing its most exalted instances44 and to overburden the low-level clerical work with vast assumptions of cultural competence and meaning. (Likewise, it would be anachronistic to project the particular characteristics of Warring States menial writing, and of low-level clerks, wholesale into much earlier periods.) Such confusion is apparent, for example, in the assumption that an ordinary Han scribe had to master 9,000 different characters. It creates the danger of overlooking differences in the functions and levels of literacy and of blinding us to the particularities of writing for diverse purposes. The 1,095 low-level clerks listed in the Rituals of Zhou were producers of mundane records, and probably so were their counterparts in the early Zhou archives. However, 42 Kominami (1999), while arguing from the problematic vantage point of graphic analysis, arrives at much the same conclusion, namely, that the origin of the scribe should be seen in duties of calculation. 43 See, e.g., Lewis 1999b, 28. Shirakawa (1974) has suggested that the early graph 史 seems to depict the offering of a basket of inscribed slips upward, namely, to the ancestral spirits. From this perspective, the function of the shiif not the function of writing altogetherhas often been seen as having originated in a religious context. I myself have changed my mind on the issue. 44 See Nylan 2000; Kern 2000a.

OFFICES OF WRITING AND READING

77

scribal officers in the Rituals of Zhou are not limited to the menial positions and purposes implied in the office outlines. The High-Ranking Scribes in the Rituals of Zhou Since at least Eastern Zhou times, the Chinese tradition has privileged the representation of the scribe from the perspective of high rank, religious duty, historical knowledge, moral judgment, and political influence. The Rituals of Zhou account corresponds partially with this image. The forty-two offices that mention the writing, reading, or storing of texts are largely devoted to fiscal, legal, administrative, military, ritual, and astrological matters. However, it is the Ministry of Ritual (chunguan 春官) that includes all but one of the actual offices of scribes: the grand scribe (taishi 大史), the minor scribe (xiaoshi 小史), the scribe of interior affairs (neishi 內史), the scribe of exterior affairs (waishi 外史), and the scribe in royal attendance (yushi 御史). In addition, only the Ministry of State (tianguan) includes the office of the scribe of female affairs (nüshi 女史) in the service of the queen. In other words, the offices of the shi belong largely to the ritual and religious administration, while the offices of writing in the Ministries of State, Education (diguan 地官), War (xiaguan 夏官), and Justice (qiuguan 秋官) bear different titles and are mostly concerned with administrative, legal, and economic purposes. As will be seen below, the ritual dimension of the shi office is also the one that overwhelmingly dominates its appearance in both bronze inscriptions and early transmitted texts. For the purposes of the present chapter, I relegate the offices of writing in the Ministries of State, Education, War, and Justice to the appendix while focusing on those in the Ministry of Ritual. In the Ministry of Ritual, the following offices of writing and of the proclamation of texts are listed:45 • The great invocator (taizhu 大祝) is responsible for the prayer words for the six kinds of invocations. He tells the officer in charge of the public fields (dianren 甸人) to read out loud the prayer text. Staff: two junior grand masters, four senior servicemen; at the same time, the office is directly superior to that of the minor invocator (xiaozhu 小祝), which has a staff of eight ordinary servicemen, sixteen junior 45

The following list contains direct translations from the Rituals of Zhou while also summarizing and/or further elaborating on the text for the individual offices.

78

CHAPTER TWO

servicemen, two storehouse keepers, four scribes, four aides, and forty runners. • The grand scribe (taishi 大史) is in charge of the six codices of the state. During the fasting periods before the major sacrifices, he, together with other ritual officers, reads out loud the ritual writings (lishu 禮書). On the day of the sacrifice, those in charge of the ritual writings take their positions according to the appropriate sequence. On grand assemblies and audiences, the grand scribe uses the ritual writings to harmonize the ritual affairs. On the day of granting silk, he takes the writings to report to the king. At a grand funeral, on the day of sending off the dead to his place in the ancestral temple, he reads out loud the dirge (lei 誄). Staff: two junior grand masters, four senior servicemen. The office is directly superior to that of the minor scribe. • The minor scribe (xiaoshi 小史) is in charge of the records of the state, of establishing genealogies, and of defining the alternating zhao 昭 and mu 穆 order in the ancestral line. At a grand sacrifice, when reading out loud the ritual regulations (lifa 禮法), the scribe uses his writings to put the sacrificial stands into the correct sequence. At a grand funeral, when the posthumous name is bestowed, he reads out lout the dirge. Staff: eight ordinary servicemen, sixteen junior servicemen, four storehouse keepers, eight scribes, four aides, forty runners. • The royal astrologer (baozhangshi 保章氏) is in charge of observing the astral bodies. He records the movements of the stars, constellations, sun, and moon in order to observe the changes under Heaven, so as to discriminate auspicious and inauspicious events. Staff: two ordinary servicemen, four junior servicemen, two storehouse keepers, four scribes, eight runners. • The scribe of the interior (neishi 內史) is in charge of the laws of the king’s “eight handles” of government and of proclaiming the king’s ordinances. He reads out loud the writings that record the affairs from the four quarters. He is further in charge of writing down the king’s commands and of storing them in the archive. Staff: one ordinary grand master, two junior grand masters, four senior servicemen, eight ordinary servicemen, sixteen junior servicemen, four storehouse keepers, eight scribes, four aides, forty runners. • The scribe of the exterior (waishi 外史) is in charge of writing down the commands issued to the realm beyond the capital. He also is in charge of the records of the four quarters. Staff: four senior serv-

OFFICES OF WRITING AND READING

79

icemen, eight ordinary servicemen, sixteen junior servicemen, two aides, twenty runners. • The scribe in royal attendance (yushi 御史) is in charge of royal edicts and commands issued to the state, the capital, and the general population. All those in governing positions receive their laws and ordinances from him. Staff: eight ordinary servicemen, sixteen junior servicemen, one hundred and twenty scribes, four storehouse keepers, four aides, forty runners. From this list and the staff members as listed in the “outlines of offices,” it is clear that the various scribes are of very different ranks and in charge of very different duties. The office of the grand scribe, for example, does not have its own menial scribes; these are included only in the lower-ranking office of the minor scribe with its far more extensive staff. Apparently, the office of the grand scribe is in charge not of producing documents but of presenting them on important occasions. The highest-ranking scribal office is that of the scribe of the interior, headed by an ordinary grand master; on the other hand, the office of the scribe in royal attendance is headed by eight ordinary servicemen—that is, members of the lowest stratum of the ranked nobility— yet commands no fewer than one hundred and twenty menial scribes. The symmetrical account of offices and their staff is a utopian ideal of bureaucratic order, and subsequent commentaries have further added to the idealization. But the Rituals of Zhou is not entirely fictional. It comprises the knowledge of a wide range of ritual and administrative practices that at least in part can be traced back to middle and late Western Zhou times; its fiction is to synthesize and recast this diachronic and fragmentary knowledge as a synchronic and comprehensive blueprint of government. The scribal offices of the Ministry of Ritual exemplify this development. Three other types of sources (two transmitted, one excavated) provide the bulk of references to these offices: Western Zhou bronze inscriptions, the ritual canons of the Ceremonial Rituals (ten passages) and Records of Ritual (twenty-six), and the anecdotal historiographies of the Zuo Tradition (Zuo zhuan 左 傳; sixty-six) and the Discourses of the States (Guoyu 國語; twentyseven).46 By comparison—considering just the remaining texts of the Thirteen Classics—the Documents mention the shi in four passages, 46

My count of “passages” is the result of an electronic search of the Academia Sinica database.

80

CHAPTER TWO

the Poetry in two, the Gongyang Tradition (Gongyang zhuan 公羊傳) in two, the Guliang Tradition (Guliang zhuan 穀亮) in three, the Analects in three, the Approaching Elegance (Erya 爾雅) in one, and the Mengzi in two. While these numbers do not constitute a complete account, they show an overall tendency of the received tradition to regard the high-level scribes, on the one hand, as ritual specialists and, on the other hand, as political advisors to, and speakers for, their rulers. In fact, these functions are related because the advisors (e.g., the remonstrators in the Zuo Tradition and the Discourses of the States) are men profoundly learned in the past as well as in the operations of the moral cosmos. Their knowledge comprises, among other branches of learning, the command of historical precedents and their applicability to the present, the interpretation of portents, the calculation of the calendar, and the correct order of ritual, especially sacrificial, activity. Much of their competence thus overlaps with that of the ritual specialists noted in the Ceremonial Rituals and the Records of Ritual. However, the Zuo Tradition, the Discourses of the States, the Ceremonial Rituals, and the Records of Ritual mention scribes not only as officers of general competence in ritual matters but in two distinctly religious functions, namely, as diviners and invocators. Only once does the Rituals of Zhou mention a scribe as a participant in the actual practice of divination (in the discussion of the diviners [zhanren 占人] in the Ministry of Ritual); by contrast, the Records of Ritual and the Discourses of the States contain three such passages each, and the Zuo Tradition contains seven.47 Likewise, the Rituals of Zhou contains not a single reference to the officials that in the other sources are named invocator scribe (zhushi 祝史), spirit medium scribe (wushi 巫 史), sacrificial scribe (jishi 祭史), divination scribe (bushi 卜史), or milfoil diviner scribe (shishi 筮史)—all of them religious specialists who seem to have been in direct contact with their rulers. While the Rituals of Zhou, like the Zhangjiashan Statutes and Ordinances of the Second Year, mentions the high-ranking scribes directly next to the invocators and other religious specialists, it never conflates the two. On the other hand, the Ceremonial Rituals, Records of Ritual, Discourses of the States, and Zuo Tradition clearly speak of these officials not as menial clerks but as men of considerable status; they are 47

Zhouli zhengyi 48.1960; Liji zhengyi 17.153a (“Yueling” 月 令 ), 29.247a (“Yuzao” 玉藻), 40.323a (“Zaji” 雜記); Chunqiu Zuozhuan zhengyi, Zhuang 22, Xi 15, Cheng 16, Xiang 25, Zhao 7, Zhao 31, Ai 9; Guoyu, (“Jin yu”) 1–2, 4; Yili zhushu 47.252a–c (“Shaolao kuishi” 少牢餽食).

OFFICES OF WRITING AND READING

81

decidedly not the low-ranking scribes that are, as a matter of course, found in the offices of the minor invocator and virtually all other religious officials of the Rituals of the Zhou. Remarkably, these terms also do not appear in Western Zhou bronze inscriptions, and they are equally absent from the early layers of the Five Classics that mention scribes and may, in one version or another, possibly date from late Western Zhou times.48 In the Poetry, scribes are mentioned only in two of the “minor elegantiae” (xiaoya 小雅): In Ode 193, “At the sun-moon conjunction in the tenth month” (Shi yue zhi jiao 十月之交), the scribe of the interior is mentioned among the highest dignitaries of the royal court. In Ode 220, “When the guests first take their places on their mats” (Bin zhi chu yan 賓之 初筵), a scribe—possibly of relatively low status—assists an inspector who records those who are drunk at a lavish court banquet.49 In the royal speeches that compose the early layers of the Documents, scribes are mentioned in a number of chapters. In the “Metal-Bound Coffer” (“Jin teng” 金縢), the scribe presents the Duke of Zhou’s 周 公 written prayer (ce zhu 冊祝); when the writing is later recovered, the collective group of scribes is consulted about its contents.50 In the “Announcement about Alcohol” (“Jiu gao” 酒誥), both the grand scribe and the scribe of the interior are mentioned among the high dignitaries; furthermore, the grand scribe is so mentioned in the “Establishment of Government” (“Li zheng” 立政) and the “Testamentary Charge” (“Gu ming” 顧命).51 In the “Announcement concerning Luo” (“Luo gao” 洛誥), the king orders a “maker of records” (zuoce 作冊) to announce a written prayer and, later, a written charge.52 In the “Testamentary Charge,” the late king’s testamentary charge is first recorded on bamboo slips; later, in an elaborate ceremony, the grand 48

See Xi Hanjing 1983; Lai Changyang and Liu Xiang 1985; Zhang Yachu and Liu Yu 1986. In my following account of the Poetry, Documents, and Western Zhou bronze inscriptions, I draw on Kern 2007, where I extensively discuss writing as performance and representation of writing during the Western Zhou; see also there for the arguments concerning a late Western Zhou date for the relevant Poetry and Documents passages. The Documents chapters discussed here all belong to the authentic “modern text” version of the text, and within that version, they come from the earlier chapters. 49 Legge 1895, vol. 4, 322, 399; Karlgren 1950a, 139, 174. 50 Legge 1895, vol. 3, 353, 359–360; Karlgren 1950a, 35–36. Here and in the following, references to Legge and Karlgren are given for convenience; see Kern 2007 for my disagreements with many of their translations. 51 Legge 1895, vol. 3, 410, 515, 557; Karlgren 1950a, 45, 68, 71. 52 Legge 1895, vol. 3, 451–452; Karlgren 1950a, 55.

82

CHAPTER TWO

scribe presents the recorded charge to the new king.53 Finally, another text possibly of Western Zhou origin,54 the Remnant Zhou Documents (Yi Zhoushu 逸周書) chapter “The Great Capture” (“Shi fu” 世俘) that relates the Zhou conquest of the Shang, notes that the victorious King Wu 武 asked his scribe to recite a document to announce the royal conquest to Heaven.55 In sum, all the appearances of scribes in the early Documents chapters match perfectly well with the account of their position and duties in the Rituals of Zhou. In both sources, the different types of scribes, as well as the makers of records, are high officials close to the king. Their duties are centered on ritual as well as on governmental matters, including the public proclamations regarding the ritual order. The same is true for Western Zhou bronze inscriptions; here as in the Documents speeches, the scribes, in Constance A. Cook’s words, were “the most powerful ritualist[s] and minister[s] in the king’s service.”56 As patrons of inscribed bronze vessels, not only their names but also their titles appear in the inscriptions. The reference work Index to Bronze Inscriptions (Jinwen yinde 金文引得) notes 83 appearances of scribes for the Shang dynasty, 4 for either the Shang or Western Zhou, 220 for the Western Zhou, 11 for the Spring and Autumn period, and 5 for Warring States times. In addition, it registers three inscriptions that refer to makers of records for the Shang and twenty-five for the Western Zhou57—though none of these inscriptions mentions a scribe or maker of records as being in charge of the actual practice of writing. Instead, they are representatives of the king who speak on behalf of their ruler either during court ceremonies or when sent out as envoys. Serving in this unique capacity, they were men of great stature who often had their accomplishments recognized and were given permission to have them recorded on inscribed bronze vessels. This is not to downplay the ample presence of documents at the Western Zhou royal court and their association with these high officials. Following King Zhao’s 昭 (r. 977/75–957 BCE) disastrous military campaign southward that resulted in defeat and the death of the 53

Legge 1895, vol. 3, 549–551, 558; Karlgren 1950a, 70–71. As argued by Shaughnessy 1980–1981. 55 Huang Huaixin, Zhang Maorong, and Tian Xudong 1995, 464–465; Shaughnessy 1980–1981, 59. 56 C. Cook 1995, 250. 57 There is no question that, in many other inscriptions, scribes or makers of records do not identify themselves with their title, but the actual number of these inscriptions is impossible to determine. 54

OFFICES OF WRITING AND READING

83

king,58 large numbers of official appointments were given to members of the elite beyond the royal clan. Meanwhile, the eastern part of the realm appears to have slipped from royal control; “inscribed vessels from the middle and late Western Zhou have been found almost exclusively in the western Wei River capital region.”59 In this time of crisis, a series of social, political, ritual, and administrative reforms led to a more complex bureaucracy. At the same time, all aspects of the production of inscribed bronze vessels became increasingly standardized: the design of the vessel, the calligraphy, the phrasing of the text.60 The most conspicuous evidence of this increasingly bureaucratic ritual, political, and administrative order are the appointment inscriptions that begin to appear in mid–Western Zhou times with the reign of King Mu 穆 (r. 956–918 BCE), the successor to King Zhao. In these ceremonies of royal appointment, the king (or sometimes a high-level aristocrat) issued a profoundly formulaic “charge” (ming 命, also written as ling 令, “order”) or “bestowal” (ci [易>]賜) with which he commanded the appointee to a certain position and bestowed on him the insignia and paraphernalia appropriate to his rank. The charge or bestowal was read from a bamboo document (ce 冊), a copy of which was given to the appointee; this text then served as the basis for the inscription of a bronze artifact that was cast in his name and with which he could henceforth sacrifice to his ancestors and inform them about his accomplishments.61 Not only the inscription (and the vessel as a whole) but also the underlying ceremony were strictly codified.62 To date, six late Western Zhou inscriptions provide the most comprehensive picture of the ceremony: on the Song-ding 頌鼎 tripod (ca. 825 BCE?), which is repeated on at least three ding tripods, five gui tureens and their lids, and two hu 壺 vases and their lids; the Fengding 鼎 tripod (809 BCE); the Huan-pan 盤 water basin (800 58

Li Feng 2006, 93–102; Shaughnessy 1999, 322–323. Shaughnessy 1999, 323–328 (quotation on 325); see also Li Feng 2000. 60 Rawson 1990, vol. 2A, 93, 125. As noted by Rawson (1999, 438–439), “a strong centralized control of ritual seems to have been in place”; for bronze design, a “static repertoire” came into being, “limited and reiterated” and of “persistent sameness”expressing an aesthetic ideology that embraced the bronze object as well as the wording and calligraphy of the inscription. 61 Jinwen yinde contains forty-two bronze inscriptions that use the phrase “issue the charge/order from a bamboo document” (ce ming/ling) and seven more that use “issue the bestowal from a bamboo document” (ce ci). Kane 1982–1983; Chen Hanping 1986; Wong Yin-wai 1978; Falkenhausen 1993, 156–167. 62 Chen Hanping 1986, 28–31. 59

84

CHAPTER TWO

BCE), which is repeated on at least one ding tripod; the Shanfu Shanding 善夫山鼎 tripod (789 BCE); and the two separate Qiu-ding 逑鼎 tripods (786 and 785 BCE), which are repeated on two and ten ding tripods, respectively.63 The Feng-ding inscription (809 BCE) comprises ninety-seven characters: It was the nineteenth year, the fourth month, after the full moon, the day xinmao. The king was in the Zhao [Temple] of the Kang Palace. He arrived at the Great Chamber and assumed his position. Assisted to his right by Intendant Xun, [I,] Feng entered the gate. [I] assumed [my] position in the center of the court, facing north [toward the king]. Scribe Liu presented the king with the written order. The king called out to the scribe of the interior, X,64 to announce the written bestowal to [me,] Feng: “[I bestow on you] a black jacket with embroidered hem, red kneepads, a scarlet demi-circlet, a chime pennant, and a bridle with bit and cheek-pieces; use [these] to perform your service!” [I] bowed with my head touching the ground. [May I] dare in response to extol the Son of Heaven’s greatly illustrious and abundant blessings and on account of this make for my August Deceased Father, the Elder Y, and his wife Zheng [this] precious tripod! May [I enjoy] extended longevity for ten thousand years! May sons of sons, grandsons of grandsons, forever treasure [this tripod]! 唯十又四年月既望辛卯。王在周康昭宮。格于大室即位。宰訊右 入門立中廷北嚮。史留受王令書。王呼內史留冊賜 玄衣純黹 赤市朱黃鑾旂攸勒，用事。 拜稽首。敢對揚天子丕顯魯休，用 作朕皇考釐(?)伯鄭姬寶鼎。其眉壽萬年。子子孫孫永寶。65

This text is in large part identical to the Song, Huan, and Shanfu Shan inscriptions. This fact testifies to a standard institutional writing practice as well as to archival records at King Xuan’s 宣 (827–782 BCE) court, as the inscriptions date from 825(?), 809, 800, and 789 63 For the text of the Feng-ding inscription, see Chen Hanping 1986, 26; Chen Peifen 1982, 17. For the Song-ding inscription, see Shirakawa 24.165–168 (no. 137), with the translation of the gui inscription in Shaughnessy 1999, 298–299. For the Shanfu Shan-ding inscription, see Shirakawa 26.357–361 (no. 154), with the translation in Shaughnessy 1997a, 74–76. For the Huan-pan inscription, see Shirakawa 29.590–595 (no. 177). For the Qiu-ding (also called Zuo-, Mai-, or Lai-ding by some scholars) inscriptions, see the large number of essays in Wenwu 2003.3, Kaogu yu wenwu 2003.3, and Zhongguo lishi wenwu 2003.4. Shaughnessy (1999, 298) dates the Song-gui (and by implication the other Song vessels) tentatively to 825 BCE. The remaining inscriptions are fully dated (year, month, day). See also Shaughnessy 1991, 285, table A16. 64 The name of this scribe of the interior is unclear. 65 Chen Peifen 1982, 17–20; Chen Hanping 1986, 26. Here and below, I provide the bronze inscription characters in their accepted transcription and interpretation by the modern editors.

OFFICES OF WRITING AND READING

85

BCE and the bronze vessels were in the possession of different individuals. The two Qiu inscriptions, at 281 and 316 characters, are far longer in their royal speeches and the final “statement of purpose” by the appointee, Qiu, but they still relate the same sequence of events and mention by name the same scribe who fifteen years earlier appeared already in the Huan-pan inscription. Furthermore, right before the statement of purpose “on account of this make…” (which was not part of the appointment ceremony but was later affixed to the text as it was prepared for inscription), they mention a concluding step in the ceremony that is also related in the Song and Shanfu Shan inscriptions; I quote the second Qiu-ding:66 [I,] Qiu, bowed with my head touching the ground. [I] received the bamboo slips and suspended them from my girdle before exiting. In return, [I] submitted a jade tablet. 逑拜稽首受冊佩以出，反入堇圭。

The vast majority of mid– and late Western Zhou appointment inscriptions mention neither the initially written order (lingshu 令書) nor the appointee’s exiting with the inscribed charge; they simply provide the text of the royal appointment charge, often introduced by “the king says” (wang yue 王曰).67 However, it is clear that the charge was always issued in an elaborate, strictly orchestrated and standardized court ritual, and that this ceremony is represented most extensively in the Song, Shanfu Shan, and Qiu inscriptions. The textual performance that emerges from these standardized descriptions—note the uniformity not only of the performances but also of their inscribed records—is a complex interplay between oral and written textual presentation, and it includes high-ranking scribes (or makers of records) in both. First, prior to the appointment ceremony, a written order was prepared at court. After the ritual participants formally assumed their positions, a scribe handed the order to the king. The king then called out to a second scribe to read the written order to the appointee. The document was given to the appointee, who attached it to his girdle. After the ceremony, this written charge served as the basis for the bronze inscription. The appointment ceremony was thus 66 For a detailed study and full translation of the Qiu bronze inscriptions, see Falkenhausen 2006; for the present passage, he offers a different, equally plausible translation. 67 For a detailed discussion of this phrase and its ceremonial implications, see Luo Tai [Lothar von Falkenhausen] 2006, 363–364.

86

CHAPTER TWO

focused on the presentation and handover of a written document that, however, was delivered as the king’s direct speech and with the full ritual force and imposing dignity of the royal ceremony. The written document was important, but it was its ritual performance, with the king personally present, that sealed its authority. It is at this point that the function of the high-ranking scribe becomes apparent: he is the chief ritualist and royal representative who reads out loud the document—a text presumably prepared by his staff of low-ranking scribes—in order to announce the charge to the appointee. Yet, as Lothar von Falkenhausen has shown, even a text like the Feng-ding inscription may be telling only half the story. In the magnificent inscription of 373 characters on the Qiu-pan water basin that was excavated in 2003, together with the tripods mentioned above, in Yangjiacun 楊家村 (Mei xian 眉縣, Shaanxi), the royal speech is, in fact, given in response to a speech in which Qiu extols his and his ancestors’ accomplishments. The Qiu-pan is one of only two texts known to date that include both speeches—the other one being the inscription on the Da Ke-ding 大克鼎 tripod 68 —while numerous Western Zhou bronze inscriptions relate either one or the other. Yet the Qiu-pan’s royal speech is clearly related to its counterparts in the two tripod inscriptions that describe in detail the appointment ceremony. Considering the very high degree of standardization in mid– and late Western Zhou court ritual and its textual expressions both oral and written, it seems unlikely that the Qiu-pan reflects a singular case; in fact, traces of similar speech exchanges have already been identified.69 Most importantly, Qiu’s extensive speech cannot have been an impromptu performance (if that would have been at all permissible in the highly formalized appointment ceremony). It includes—in this respect similar to the famous pan water basin of Scribe Qiang—not only a list of Qiu’s ancestors but also a list of the Zhou kings whom they had served. Without question, such lists were backed by documents in the royal archives from which Qiu’s text was then carefully prepared in writing. Finally, in another expression of Western Zhou royal ritual and textual culture, parallel to the rise of the appointment ceremony and its representation in bronze vessels, inscriptions of legal contracts ap-

68 69

See Shirakawa 28.490–511 (no. 167) and the analysis in Luo Tai 2006, 357–359. Luo Tai 2006.

OFFICES OF WRITING AND READING

87

peared in larger numbers.70 One characteristic feature of these inscriptions is that they meticulously list the names and titles of the officials who served as witnesses at the time of the legal agreement. The same logic seems to underlie the inscriptions that give more detailed accounts of appointment ceremonies: especially from the appointee’s perspective, it was important that his court-sanctioned bronze vessel included the names and titles of the officials who delivered the appointment (and the right to a bronze vessel). Through their texts, such bronze vessels forever contained and displayed the origins of their own existence—the official, bureaucratically verified event of the appointment ceremony—and enlisted both the king and his highest ritual officials as witnesses. Conclusion The early passages from the Poetry and the Documents as well as numerous examples of Western Zhou bronze inscriptions show the highranking scribes of the royal court in precisely the functions attributed to them in the Rituals of Zhou. As representatives of the king, these officials—especially the scribe of the interior, who ranks the highest in the Rituals of Zhou and appears with particular prominence and frequency in mid– and late Western Zhou inscriptions—went on diplomatic and military missions, led important royal rituals, and announced the king’s proclamations. They also were the men in command of the official court documents and their ritual presentation but apparently not personally involved in their menial preparation. The latter duty, and also the maintenance of what must have been sizable royal archives, will have fallen to clerks and storehouse keepers. Thus, the Rituals of Zhou account of scribes as chief ritualists and royal representatives, on the one hand, and of menial clerks, on the other, accords well with the available Western Zhou sources. While the clerks, working on perishable stationery, have left no direct traces of their activities, circumstantial evidence for the presence of archives and the use of writing for administrative, legal, economic, and other pragmatic purposes is incontrovertible. Late Warring States and early imperial manuscripts—the earliest evidence we have on wood, bamboo, and silk—show the continuity and extensive proliferation of such writing over time. 70

U. Lau 1999; Schunk 1994; Skosey 1996.

88

CHAPTER TWO

Evidence for high-ranking scribes as royal dignitaries comes directly from Western Zhou bronze inscriptions as well as from the early layers of the Documents and the Poetry. All terms used for scribes in the Rituals of Zhou also appear in bronze inscriptions, but only some of them can be found in later texts such as the Zuo Tradition, the Discourses of the States, the Ceremonial Rituals, and the Records of Ritual, among others. On the other hand, these later texts show scribes frequently involved in a range of more specifically religious practices; during the Warring States, in fact, the term shi referred not merely to clerical officials but also to genuine religious specialists. Yet this is not how shi officials (or makers of records) appear in the early passages from the Poetry and the Documents or in Western Zhou bronze inscriptions—or in the Rituals of Zhou. In short, the account of scribes in the latter text matches their characterization in Western Zhou, rather than Warring States and early imperial, sources. The case of the scribes, however limited it may be, suggests that at least some knowledge of actual mid– and late Western Zhou offices was available to the authors of the Rituals of Zhou, and that they championed this knowledge over more recent or contemporary practices. Without access to the numerous Western Zhou bronze vessels that after some three millennia have finally been excavated from tombs and storage pits, the authors of the Rituals of Zhou not only listed the correct Western Zhou offices but also managed to avoid conflating earlier with later information. They were able to distinguish ancient, long-discontinued practices from contemporaneous ones and decided to give an account that they knew was far removed from their own time. Nevertheless, as has often been pointed out, this is by no means true for all of the Rituals of Zhou; there simply is too much Warring States (or even early imperial) thought and language interwoven into the texture of this ideal governmental blueprint, and the fictional nature of the Rituals of Zhou arises from the conflation of diachronic phenomena. The complexity of this situation invites us to rethink some of our stereotypes in dating the Rituals of Zhou or, for that matter, any other early text. The text as a whole must be a late composite, but its parts are not forged. Whoever compiled or used the text in Warring States or early imperial times thought of it as describing an ideal of perfect and comprehensive order. Pending further research, the account of the scribes may be a peculiar exception—yet an exception that speaks in a unique way to the nature of the Rituals of Zhou as a text. The knowledge about the

OFFICES OF WRITING AND READING

89

scribes testifies to an institutional memory that must have existed in the form of archives. Yet unlike most other cultural artifacts, the archives, preserving the bureaucratic order of old, were the products of scribes. Transforming these scribal products into a new one, the Rituals of Zhou came to express the ideal order of government and more: the text emerged as the ideal account of institutional scribal memory itself that was instrumental to conceptualizing any such order. Although the Rituals of Zhou was a late Warring States or early imperial artifact of writing, the memory of scribes enshrined in it was that of high antiquity; its authors, it appears, imagined themselves as the true successors of the ancient scribes they knew so well. In the end, their text stands as a supreme example of technical administrative writing, devoid of history, moral claims, or explicit argument. Its own imperial history as a Classic, since Han times supported by its association with the Duke of Zhou, emerged from mapping a unified bureaucratic order onto the imagined ideal of the Western Zhou—an ideal past, remembered and imagined, in which the authors of the text had already found themselves. *** Appendix: The Offices of Writing in the Ministries of State, Education, War, and Justice As earlier in this essay, the list of offices contains direct translations from the Rituals of Zhou while also summarizing, expanding, and explaining the text for the individual offices. Ministry of State (tianguan) • The chief steward (zaifu 宰夫) exerts control over eight offices of functionaries; the sixth of these eight is that of the scribes (shi 史) in charge of drafting government documents. • The accountant (sikuai 司會) conducts audits of the fiscal records and records the fiscal information of various administrative units. • The manager of writings (sishu 司書) is in charge of the tables for the various administrative codices as well as for geographical maps. He registers the inflow and outflow of goods. His office belongs to that of the si kuai.

90

CHAPTER TWO

• The keeper of consumables (zhinei 職內) records the flow of consumable goods that are coming in and being distributed across the various agencies. • The keeper of silks (zhibi 職幣) controls and records the silk that is used by various agencies. • The administrator of the interior palace (neizai 內宰) keeps the population registers and charts of those living within the palace. • The scribe of female affairs (nüshi 女史) is in the service of the queen and writes down her charges. • The supervisor of palace women’s work (dianfugong 典婦功) marks and records the value of the goods produced by the palace women. • The supervisor of silk (diansi 典絲) monitors the inflow of silk and marks and records its value. Ministry of Education (diguan) • In the office of the grand minister of education (da situ 大司 徒), writing is mentioned as one of the “six arts” (liu yi 六藝), which also include the arts of ritual, music, archery, charioteering, and calculating. • The township head (zhouzhang 州長), in the first month as well as at the major seasonal sacrifices, assembles the populace and reads out loud the laws in order to encourage the people toward morality and to warn them against unruly behavior. • The head of the ward (dangzheng 黨正) reads out loud the laws to his people at the beginning of the four seasons as well as on the occasion of the major spring and autumn sacrifices; moreover, he keeps records of their virtuous conduct. • The master of the community (zushi 族師) reads out loud the laws to his people on the relevant ritual occasions; he also keeps written records of their attainment in the polite arts. • The village assistant (lüxu 閭胥) reads out loud the laws to his people and keeps records of those who distinguish themselves through their virtuous conduct. • The remonstrator (sijian 司諫) records the moral conduct and principles he finds in the common people, and he distinguishes who is able to serve in official functions. • The arbitrator (tiaoren 調人) solves disputes among the people; when they cannot be resolved, he produces a written record.

OFFICES OF WRITING AND READING

91

• The marriage monitor (meishi 媒氏) records the names and dates of birth of all children surviving the first three months of their lives; he also records marriages. • The treasurer of the markets (quanfu 泉府) writes price tags on goods. Ministry of War (xiaguan) • In the office of the minister of war (da sima 大司馬), the scribes in the second month of summer count the war carts and the infantry, and they read out loud the written registers to match them with the actual troops. The officials all carry their insignia, and they write down their own services and designations. • The manager of rewards (sixun 司勳) is in charge of determining merit. He inscribes the names of meritorious persons onto the king’s great standard (taichang 太常). At the grand winter sacrifice, he calls upon the person in charge of the banner to proclaim the written text to the king and to the spirits. • The horse appraiser (mazhi 馬質) records the quality of a horse’s teeth and hair, as well as its market value. • The surveyor (liangren 量人) records land formations and road lengths on maps and stores the maps. • The chief of the tiger braves (huben shi 虎賁氏) is the chief of the military bodyguard surrounding and protecting the king. When roads become impassable and there are matters like a military draft, he puts the draft into writing and sends it into the four directions. • The provisioner (gaoren 槀人) is in charge of receiving material from the officer in charge of metals, and he gives them to the craftsmen who manufacture bows, crossbows, and arrows. He records the quality grades of these weapons to determine the food provisions given to the craftsmen. Ministry of Justice (qiuguan) • The vice minister of justice (xiao sikou 小司寇) reads out loud the criminal records of the imprisoned and then determines the crime and punishment according to the law. • The chief judge (shishi 士師) proclaims at court, with the wooden clapper, the “five prohibitions” regarding the royal palace, the officials, the capital, the countryside, and the army. He writes them

92

CHAPTER TWO

down on tablets and suspends them from the road gates all over the city. • The judge of the domain (fangshi 方士) brings the death penalty cases to the chief judge to decide, as well as the name of the one who has judged the case. • The litigation judge (yashi 訝士) reads out loud the vows (threats) and prohibitions at all grand ritual occasions when the masses are assembled. • The audience monitor (chaoshi 朝士) requires inscribed twopart tallies to regulate disputes over debts and only then listens to the case. He also punishes those who charge excessive interest for purchasing goods on credit. When a person wishes to take revenge on an enemy, if he has first reported his case in writing to a judge, then killing the enemy is not a crime. • The population registrar (simin 司民) is in charge of registering the population. All those growing teeth (i.e., children of seven or eight months) and older are recorded on tablets. The records include their place of origin, their gender, and their age. Every year, those who have been born and those who have died are recorded. • The enforcer of agreements (siyue 司約) is in charge of the written agreements of the state and the multitudes. All major written agreements are inscribed onto the ritual vessels of the ancestral temple; the minor written agreements are written on red charts on (lacquered?) ritual objects. • The protector of corpses (qushi 蜡氏) is in charge of any remaining bones from corpses. If somebody dies on the road, he gives orders to bury him and to set up a marker right there. The marker is to be inscribed with the day and month. • The nest remover (checushi 硩蔟氏) is in charge of removing the nests of inauspicious birds. On wooden tablets, he writes the appropriate name of the ten days, of the twelve hours, of the twelve months, of the twelve years, and the twenty-eight lunar stations. He suspends the tablets above the nests and then removes the nests. • The senior messenger (da xingren 大行人) receives and communicates with foreign guests. Nine years after the king has pacified the states and lords, the senior messenger assembles the music masters and scribes, compares the written characters, and distinguishes the sounds of the language. After eleven years, he certifies the jade tallies, unifies weights and measures, accomplishes the rituals involving sac-

OFFICES OF WRITING AND READING

93

rificial animals, unifies the computation tools, and regulates rules and standards. • The junior messenger (xiao xingren 小行人) is in charge of the ritual records of the guests from the feudal states. For each of the states, he writes in a register what benefits and what harms the multitudes; what smoothly follows and what goes against ritual customs, governmental affairs, instruction and rulership, punishments and prohibitions; the various kinds of matters that go against the royal orders; the various kinds of calamities; and the circumstances of happiness, kinship, and peace. These five matters he distinguishes for every state, responding to the command of the king and letting him know the state of affairs in all under Heaven.

CHAPTER THREE

THE MANY DUKES OF ZHOU IN EARLY SOURCES Michael Nylan, University of California at Berkeley This essay traces the evolution of portraits of the Duke of Zhou (Zhou gong 周公) in the period before the elevation of the Rituals of Zhou (Zhouli 周禮) to secure it canonical status in late Eastern Han (25–220 CE) via the systematizing works of Zheng Xuan 鄭玄 (127–200 CE), which portray the duke as a moral exemplar second only to Confucius (Kongzi 孔子) himself.1 Early in their training, nearly all readers of classical Chinese come to regard the duke as the ultimate inspiration for Confucius, thanks to Analects 7/5, in which the sage laments his recent failures to meet the duke in his dreams. Readers of classical Chinese soon learn also about the duke’s reputed authorship of a number of canonical works, including the Line Texts of the Classic of Changes (Yijing 易經), at least two odes, and possibly even the Approaching Elegance (Erya 爾雅) word list and a mathematical classic.2 Fewer readers, however, are familiar with the contradictory portraits of the duke offered to account for the consolidation of the “great peace” (taiping 太平) in early Western Zhou (ca. 1046–771 BCE) via the institution of rites, music, and punishments—portraits that query the actual motives of this legendary figure. Some Zhanguo (475–222 1 According to tradition, a version of the Zhouguan first appeared at the court of Liu De 劉德, King Xian of Hejian (r. 155–129 BCE), stepbrother of Han Wudi 漢武 帝 (r. 141–87) and a famous bibliophile. The Zhouli was briefly awarded the status of an imperially sanctioned Classic (jing 經) at the end of Western Han, and three scholars—Du Zichun 杜子春 (early Eastern Han), Zheng Xing 鄭興 (fl. 30 CE), and Jia Kui 賈逵 (30–101)—wrote commentaries for it, but its status was finally secure only after Zheng Xuan 鄭玄 (127–200) treated the Three Ritual Classics (sanli 三禮) as a single Classic. See Nylan 2001, chap. 4. The figure of the duke was also readily available for jests: students who mocked their dozing teacher were told that he was in wordless communication with the duke. See Hou Hanshu 80A.2623. (NB: As all translations used in this chapter are my own, references to translations are for the convenience of readers.) 2 Odes 155, a plaint, and 235, a hymn written in praise of King Wen 文王. For the Duke of Zhou as author of the Erya, see Loewe 1993, 33, 95; the mathematical classic is the Mathematical Classic of the Zhou Gnomon (Zhou bi suan jing 周髀算經).

THE MANY DUKES OF ZHOU

95

BCE) and Western Han (206 BCE–8 CE) texts reveal a great ambivalence toward the Duke of Zhou, with Han texts regularly citing the duke’s attempts to arrogate imperial prerogatives. 3 Such disparate pasts invoked in competing traditions command our attention—and not only because these rival visions of the past inevitably informed the early reception of both the language and content of the Rituals of Zhou itself.4 The broader questions that this essay raises include the following: What role did the figure of the Duke of Zhou play in rhetorical constructions, why, and for whom? What effect was conveyed by the various representations of the duke’s performative speeches? Did the various portraits of the duke make some things happen while preventing others? Even at this remove, it is obvious that the words attributed to the duke were at least as important in the early centralizing states as any deeds the duke supposedly performed. A mid– to late Western Han compilation, the Records of Ritual (Liji 禮記) provides a window onto these complexities in an opening passage from a chapter devoted to the Spirit Hall (mingtang 明堂) supposedly built by the Duke of Zhou shortly after the dynastic founding ca. 1046 BCE: Formerly, when the Duke of Zhou gave audience to the vassal princes in their various places in the Spirit Hall, the Son of Heaven stood with his back to the axe-embroidered screen, facing south. The “Three Lords” were in front of the steps, in the middle, facing east, as this was the most honorable position [for subjects]. The lords one rank below took their places in the eastern part of the east steps, and they faced west… [and so on down three more ranks, before the passage describes the leaders of the Nine Yi and the Eight Man].

3

For the stories of Huo Guang 霍光 (d. 68 BCE) and others who arrogated similar powers, see below and also Loewe 2000, 170–174; Loewe 2004, chap. 10, esp. 340ff. Conceivably, the duke’s example was invoked even before Han by ministers who assumed the ruler’s powers. See Wenxuan 52.16b, citing links between the duke’s actions and the usurpation of power by Tian Chang 田常 (fl. 481 BCE) in Lu. For the tradition that Zhou gong is responsible for the Nine Punishments (jiu xing 九刑), see Shangshu guwen shuzheng 尚書古文疏證 1.29b. 4 Considerable overlap, for example, exists between the portrait of the Duke of Zhou provided by the Remnant Zhou Documents (Yi Zhoushu 逸周書) and the implied portrait of the duke as compiler of the Zhouli, not only in relation to general content but also in specific vocabulary items, including “Eight Fundamentals of Governing” (ba zheng 八政) and “Six Protections” (liu wei 六衛).

96

CHAPTER THREE 昔者周公朝諸侯于明堂之位：天子負斧依南鄉而立；三公，中階 之前，北面東上。諸侯之位，阼階之東，西面北上 5

The passage fails to stipulate who presides over this scene as Son of Heaven: the Duke of Zhou or King Cheng 成王 (1042–1021 BCE), the young ruler whom the duke served as regent. It was the duke who gave audience to the assembled dignitaries, we are explicitly informed, but the “Three Lords” (san gong 三公) group of advisors was usually thought to have included the duke himself. Could a single Duke of Zhou have occupied two ritual places at once?6 As this sort of basic confusion is rife in pre-Han and Han sources describing relations between the Duke of Zhou and King Cheng, a survey of the early debates concerning the enigmatic figure of the duke seems necessary. As this essay will attest, pre-Han and Western Han texts left unanswered a host of questions concerning the duke’s loyalties, so that Eastern Han classicists still felt the need to address controversies over the duke’s motivation. Had the duke acted treacherously when he “assumed the position” of a king (dai wei 代位), “received the Mandate” (shou ming 受命), and “ascended the throne” (ji wei 即位) after King Wu’s untimely death? What did it mean when the duke acted as “overseer” (yin 尹) for all of the empire after the conquest of Shang?7 Did the duke “seize control of the government” or “act as regent,” 5 “Mingtang wei” (Legge 1885, vol. 2, 29). Commentators, proposing improbably precise layouts for the scene described here, believe that the phrase zuojie 阼階 refers to the middle of three sets of south-facing steps. Aside from the “Mingtang” chapter, the most important chapter in the Liji to discuss the duke’s reputation is “King Wen’s Sons” (“Wen wang shizi” 文王世子). However, the Shizi 尸子 “Mingtang” chapter (shang 6), by contrast, speaks of a Mingtang during Shun’s reign (i.e., more than a millennium earlier). 6 This explains the impulse to emend the passage, as is done, for example, by Huang Huaixin, Zhang Maorong, and Tian Xudong in their Yi Zhoushu huijiao jizhu 逸周書彙校集注 (hereafter Yi Zhoushu), 761. 7 For yin, some Yi Zhoushu editions substitute jun 君, which would make the duke “ruler” of the Zhou. See Li Zhenxing 1994, vol. 1, 267. The duke “took the position of the Son of Heaven” (ju Tianzi zhi wei 居天子之位) according to the Liji’s “Mingtang” chapter, and Shiji 3.132 and Hanshu 99 both speak of the Duke of Zhou taking power from King Cheng “to rule the realm” (攝行政當國). Tradition holds that King Wu died three, six, or ten years after the Shang conquest. The Duke of Zhou appears in ten chapters in the received Modern Script Documents and in two chapters in the postHan, pseudo-Kong version of the Documents; the duke is always linked to two other chapters, the “Da gao” 大誥 and “Jiu gao” 酒誥, and to two chapters no longer extant: the “Giving in the Grain” (“Gui he” 歸禾) and “Auspicious Grain” (“Jia he” 嘉禾) chapters. See Nylan 2001, chap. 3. As shown below, talk of the Three Examiners (san jian 三監) came relatively late, perhaps to offset the impression that the duke was overseeing (i.e., managing) imperial affairs alone in early Western Zhou.

THE MANY DUKES OF ZHOU

97

since both translations render the binomial phrase she zheng 攝政? Such queries may have been less hostile to the degree that the duke was viewed as mentor and precursor of another uncrowned king from Lu 魯, Confucius, whose influence gradually rose during the four centuries of the Han.8 Two hypotheses are advanced here to account for these controversies, telltale signs of which remain in a number of early texts, despite the likelihood of continual and substantial editing of texts before the Song period.9 Matsumoto Masaaki 松本雅明 laid out the first hypothesis, that the duke’s ascension to the throne following the death of his brother King Wu became problematic for the first time in mid– Zhou times, ca. 600 BCE—four centuries after the duke’s regency— when policy makers intent upon obviating succession struggles advocated a strict system of primogeniture to replace an earlier model that allowed brothers, as well as sons, to inherit the throne.10 Building upon Matsumoto’s theory, I propose a second hypothesis, arguing that still later debates on topics as various as reclusion and the institution of rites and music required ongoing reassessments of the duke’s role in the centuries after 600 BCE.11 Such debates were almost certainly complicated—like the present analysis—by the fact that the term zhougong 周公 was a title for a high-ranking minister at the Zhou court rather than a name strictly reserved for Ji Dan 姬旦, the first Duke of Zhou; evidently, in some cases, conflation of originally separate lives resulted.12 Informed speculation—but speculation nonetheless—suggests that portrayals of the duke’s interest in administration, as described fully in two Documents (Shangshu 尚書) chapters— 8

See Nylan and Wilson, forthcoming, esp. chap. 3, for further details. On the scale of editorial changes in Han times, see Kalinowski 2003. The introduction to Tian Xiaofei 2005 provides a good overview of related issues. 10 See Matsumoto 1968. Tamaki (2006) unfortunately limits his inquiry to what he deems “Confucian” sources. The Zhou Proclamations certainly depict a consortial form of government, wherein regents, advisors, and allies participate in state deliberations. 11 Kryukov (2000), like Tamaki (2006), sees the duke’s legend evolving in the hands of “proto-Confucians” and “Confucians.” I prefer, following Matsumoto, to look well beyond the ahistorical “schools” construction. 12 For example, another set of Dukes of Zhou and Shao ruled jointly when King Li of Zhou (r. 857–828 BCE) was exiled by his own subjects in 843 BCE, as noted in Shiji 4.143, 145. This joint rule generated the term gonghe 共和 to describe powersharing during Western Zhou. The legend that the Duke of Zhou went to Chu 楚 (i.e., far to the south) may have arisen because the personal name of one of the many Dukes of Zhou was Chu. See Chunqiu jingzhuan yinde 232/Cheng 12/1 jing, 1 Zuo; Lunheng jijie 2.25. But cf. Zhu Fenghan 2006 for a late Western Zhou bronze. 9

98

CHAPTER THREE

“Establishing Rule” (“Li zheng” 立 政 ) and “Officers of Zhou” (“Zhouguan” 周官)13—prompted the eventual attribution of a second text by the name of Zhouguan to the duke, as did the centralizing efforts credited to the duke in the Remnant Zhou Documents (Yi Zhoushu 逸周書).14 Two texts in particular—the Documents and the Yi Zhoushu— dominate present discussions about the Duke of Zhou, given the number of times that the first Duke of Zhou speaks and acts in them, while a third text, the Archivists’ Records (Shiji 史記), proves of interest mainly because of the startling omissions and elisions in its narratives. The Documents is a canonical compilation whose records purportedly go back to the very dawn of historical time. Competing versions of the Documents circulated into the Zhanguo, Han, and post-Han periods,15 and in pre-Han and Han times doubts regarding the reliability of the corpus centered specifically on the chapters relating to the Zhou conquest,16 if extant texts reflect earlier realities. Most scholars have believed the earliest parts of the Documents to be a group of Proclamations (gao 誥) whose re-created addresses describe the drawn-out process whereby the Zhou rulers conquered the Shang, overcame the initial resistance to their rule, and finally achieved an era of great peace and prosperity that lasted for about a century in fact, and millennia longer in cultural memories.17 No fewer than five Proclamations are ascribed to the Duke of Zhou: the “Great Proclamation” (“Da gao” 大誥), the “Proclamation to Kang” (“Kang gao” 康誥), the “Proclamation on Wine” (“Jiu gao” 酒誥), the “Proclamation to Duke Shao” (“Shao gao” 召誥), and the “Proclamation at Luo” (“Luo gao” 洛誥). These chapters are generally considered “wholly trustworthy and authentic” (i.e., reflecting genuinely early Western Zhou traditions), 18 even if they were probably written down centuries later, 13 An “Archaic Script” (guwen 古文) chapter, the “Zhouguan” is usually dated to the fourth century CE, but certain passages in the pseudo-Kong traditions may reflect older formulas. 14 And possibly Analects 18/10. 15 See Matsumoto 1966. 16 E.g., Mencius 7B/3, on the “Wu cheng” 武成 chapter; Fayan zhu 5/7–8, regarding the Documents in general and the “Jiu gao” in particular. 17 It seems likely, however, that those Proclamations were transcribed centuries after the Zhou conquest, around the start of Eastern Zhou in 771 BCE. The precise dating of the Proclamations is not central to the argument here. 18 See Gu Jiegang 1926, 201–202. However, Gu’s student He Dingsheng (1928) found that probably only the “Da gao” is a Western Zhou text (a conclusion generally ignored).

THE MANY DUKES OF ZHOU

99

around the end of Western Zhou, as now seems increasingly likely.19 Another Documents chapter, “Metal Coffer” (“Jin teng” 金滕), apparently of still later date (early or mid-Zhanguo?), reads like a defense of the duke designed to dispel charges of disloyalty relating to these Proclamations from the duke’s critics, who may have included the great Confucius himself (see below). Scholarly consensus today has a majority of the Yi Zhoushu chapters being compiled in late Zhanguo, after the “Metal Coffer” chapter, and some even in Qin (221–210 BCE) or early Western Han, judging from their style of argumentation, as well as the type and range of activities that the Yi Zhoushu credits to the Duke of Zhou.20 Its late portrait of the duke-as-planner, laid out in ten chapters out of its total of seventy,21 appeared when issues of timely action and “acquiring people” preoccupied policy makers in the Six Kingdoms, and persuaders had come to rely heavily upon numbered rubrics, such as the Nine Pretexts, Three Limits, and Ten Treacheries, when making their pitches. The Yi Zhoushu casts the duke as a loyal public servant whose governing style foreshadows that of the celebrated prime minister Guan Zhong 管仲 (d. 645 BCE).22 In Han times, the authority of the 19

Kai Vogelsang (2002) counts 1,282 instances of discrepancies between the language of the Proclamations and the extant bronze inscriptions. 20 On the dating of the Yi Zhoushu, see Loewe 1993, 229–233; Huang Peirong 1976; McNeal 2000. Huang Peirong speaks of a “core text” (urtext?) of the Yi Zhoushu dating to about the fourth century BCE, with chapters added later, some well after unification in 221 BCE. Gu Jiegang (1998, 20) also dates a number of chapters to Han. Zhu Fenghan and Xu Yong (1996) argue that the Yi Zhoushu represents Western Zhou materials rewritten in Eastern Zhou style, but they fail to convince. McNeal (192) notes that Xu Shen 許慎 (ca. 100 CE) was the first person we know of to have called the text by this three-character title; earlier, the text, evidently with other sources, was called simply the Zhou Writings/Documents (Zhoushu 周書). Another “ Zhoushu” or “Zhou Record” 周志 was found in 279 CE at Jizhong, as noted by Zhu Xizu (1960, 29–30). 21 The Yi Zhoushu has the Duke of Zhou speaking in thirteen chapters (pian 篇). In other chapters, the duke makes a brief appearance, as in the “Conquest of Yin” (“Ke Yin” 克殷) and “Spirit Hall” (“Mingtang” 明堂) chapters; see Yi Zhoushu 375–376, 759–760. In still other chapters, it is a matter of dispute who is speaking as king. Gu Jiegang (1998, 21) and Guo Weichuan (2001, 259) argue that the king is the Duke of Zhou; in chapter 44, “Taking the Measure of the Capital” (“Du yi” 度邑), King Wu speaks to the duke, urging him, according to the Yin (i.e., current) form of succession, to accept the throne after his death. On the significance of this passage, see Guo Weichuan 2001, 258, which, however, generally assumes the Yi Zhoushu chapters to be accurate, and possibly contemporaneous, reflections of Late Shang (ca. 1200–ca. 1046 BCE) and early Zhou events, rather than post facto plausible reconstructions. 22 Lüshi chunqiu zhuzi suoyin (hereafter Lüshi chunqiu) 4.3/19/4 makes the comparison explicit.

100

CHAPTER THREE

Yi Zhoushu was far greater than it is today; the compilation was believed, thanks to its pseudo-archaistic style, to contain genuine writings of Western Zhou date.23 Some late Western Han thinkers even combined the twenty-eight chapters of the Modern Script Documents with the seventy or seventy-two chapters of the Yi Zhoushu into a single corpus, presumably because the Yi Zhoushu comments and expands upon the Documents even as it subtly revises the latter text’s portraits of the past. 24 Nevertheless, the Documents and the Yi Zhoushu in some sense occupy two poles of the received tradition: whereas every phrase in the Documents is the subject of multiple commentaries, some of them as early as Han, all but one of the few extant commentaries to the Yi Zhoushu, that of Kong Chao 孔晁 (mid–third century CE), date to millennia later and thus are unlikely to represent much more than “reasonable conjecture.”25 Sima Tan 司馬談 (d. 110 BCE) and Sima Qian 遷 (145?–86? BCE), compilers of the Shiji around 100 BCE, had to adjudicate between opposing traditions when compiling their “Basic Annals” and “Hereditary Houses” sections. The Shiji, in consequence, provided a balanced account that tipped readers off to the most serious charges leveled against the Duke of Zhou while showcasing the duke’s authorship of canons and commentaries and his purported role in creating the Zhou rites and music.26 Significantly, however, the Shiji’s authors “did not use the occasion to applaud the glories of the Zhou dy23 See McNeal 2000 and n. 20 of this essay. For the term “Zhou Writings/Documents” and its relation to the Yi Zhoushu and other early texts, see Yanaka 1999. 24 Possibly, the Yi Zhoushu—now nearly forgotten—represents a commentary, expansion, or critique of the Documents canon, since (a) nearly all its chapters begin with one or another of the standard opening formulas employed in the Documents (e.g., wang ruo yue 王若曰, wei wang X si X yue ji sheng po 維王X 祀X 月既生魄 ); (b) the preface of the Yi Zhoushu clearly relates to the “Preface to the Documents” attributed to Kong Anguo 孔安國, which gained circulation no earlier than the late second century BCE; (c) the Yi Zhoushu expands yet revises the portraits of many of the Western Zhou figures mentioned in the Documents; and (d) the Yi Zhoushu, chap. 28, seems to expand upon the “Great Plan” (“Hong fan” 鴻範) chapter of the Documents. 25 For Qing philological studies based on “reasonable conjecture,” rather than appeals to “early witnesses,” see Vogelsang 2007. 26 I speak of “the rites and music,” though it is not clear that Sima Qian saw the early Zhou rites and music as a unitary system. For the Duke of Zhou as author, see, e.g., Shiji 4.133. An additional quandary provoked by the Documents and made explicit in the Shiji is whether King Cheng or the Duke of Zhou wrote the “Zhouguan” 周官 (presumably the “Documents” chapter rather than the book now known as the Rituals of Zhou).

THE MANY DUKES OF ZHOU

101

nasty.”27 But within a century after the Shiji, thinkers formulating a new brand of classicism sought to cleanse the duke’s record of all the aspersions cast against him, so that he, who had once wielded all the power behind an imperial throne, might provide a more perfect model for contemporary aspirants to political power. The new classicism promoted in the final decades of Western Han ultimately prevailed at the court in Eastern Han thanks mainly to the lasting influence of Yang Xiong 揚雄 (d. 18 CE) and Liu Xin 劉歆 (d. 23 CE). It was then adopted—with minor modifications and elaborations—in the works of Zheng Xuan, which traced nearly all the enduring achievements of Central States culture back to the Duke of Zhou’s seven-year tenure as virtual king.28 The marked attention to the duke’s civilizing role, evident in texts like the Treatises on Salt and Iron (Yantie lun 鹽鐵論), continued even after the collapse of support for Wang Mang’s regime (r. 9–23 CE), shifting interest from the Zhou receipt of the mandate to the duke’s effectiveness as regent.29 The popularity of the new classicism notwithstanding, Wang Chong 王充 (27–ca. 100 CE)—a fervent admirer of Yang Xiong—in his Balanced Discourses (Lunheng 論衡, comp. ca. 80 CE) lambasted “the experts in old writings” (guwen jia 古文家)30 for their mishandling of legends about the Duke of Zhou with respect to such issues as whether it was proper for King Cheng to bury the good duke with all the honors due a king, thereby putting him on a par with the predynastic founders of Zhou. Wang’s rhetoric is intriguing insofar as it testifies to the existence of unresolved controversies surrounding the moral valence of the figure of the duke in Ru 儒 circles ca. 100 CE, long after the first major push toward the new classicism. Even a con27 See Loewe 2004, 325. Curiously, the Shiji in its closing appraisal of the “Basic Annals of the Zhou” praises the Duke of Shao but is silent on the topic of the Duke of Zhou. 28 Liu Xin allegedly discovered the text of the Zhouli in the imperial library and considered it a blueprint for Wang Mang’s takeover. Loewe (2004, 325) observes: “in his comments to his chapter on the kings of Zhou, the author of the Shiji contented himself with correcting certain misapprehensions… Perhaps we must wait until the years of Eastern Han [to find] a need, or perhaps a habit, of invoking such examples as models worthy of emulation.” 29 Majima (2002) traces this shift. 30 Lunheng jijie 55.373–375. Wang clearly means by this term the Documents in either old editions or “new-old” theories, and he never once refers to “experts in New Script classics” (jinwen jia 今文家). Gu Jiegang (1998, 27) notes that such “archaic” theories never existed in pre-Han and Western Han but can be traced to Wei Hong 衛 宏, advisor to Guangwu (r. 25–57), and Jia Kui 賈逵 (30–101) in Eastern Han.

102

CHAPTER THREE

temporary of Zheng Xuan, Zhao Qi 趙岐 (d. 201 CE), in his commentary to the Mencius, saw the need to defend the duke against recurring charges that the duke exhibited none of the foresight expected of a sage when dealing with members of his own family.31 The Duke of the Documents Chapters Commentators have come to no consensus about the Proclamations chapters in the Documents with respect to even such fundamental issues as which legendary figure speaks on behalf of the king in a specific proclamation.32 Since the Tang period (618–907 CE), classical scholars have freely acknowledged the difficulty of reading the “Luo gao” in particular and the Proclamations in general.33 In early Western Han, however, the disagreements about the Proclamations seem to have been fewer, although this impression of greater consensus may simply reflect the small proportion of Han texts that have survived down to the present. The Shiji credits the Duke of Zhou with the authorship of five Proclamations, in addition to several other frequently cited Documents chapters, including “Catalpa Timber” (“Zicai” 梓材), “Many Officers” (“Duoshi” 多士), “No Ease” (“Wuyi” 無逸), and “Lord Shi” (“Junshi” 君奭).34 Serious study of the Documents precludes attempts to locate in such texts some kernel of historical 31 See Mengzi 孟子 (Sibu beiyao 4.11a–b). Critics suggested that the duke was either mistaken—or worse—when he judged the character of his elder brother. Zhao Qi suggested that Zhou gong had no choice, as a loving brother, but to give his elder brother the benefit of the doubt, until he engaged in open rebellion. Zhu Xi (Sishu zhangju jizhu 四書章句集注 4.247) concurs. 32 For modern debates, one may consult Karlgren, Yu Xingwu, Shaughnessy, and Nivison, to take four prominent examples. Many of these debates began in Song when commentators, beginning with Hu Hong 胡宏 (1105–1155), Wu Yu 吳棫 (d. 1155), Zhu Xi 朱熹 (1130–1200), and Cai Shen 蔡沈 (1167–1230), argued that the “Kang gao” was proclaimed by King Wu rather than by the Duke of Zhou. Qu Wanli in his Shangshu jishi 尚書集釋 in the main follows these Song commentators’ readings, even where they seem forced and contra earlier readings. 33 See Han Yu’s 韓愈 “Jin xue jie” 進學解 (Dongya tang Han Changli jizhu 東雅 堂昌黎集註 12.5b), which singles out the Zhou Proclamations and the “Pan Geng” 盤 庚 chapters as exceptionally difficult. Both Wang Anshi and Zhu Xi confessed to having particular difficulties with the “Luo gao.” For Wang, see Chen Li’s 陳櫟 Shi jizhuan zuanshu (SKQS 5.13a); for Zhu, see Nalan Xingde’s 納蘭成德 Tongzhi tang jingjie 通志堂經解 ce 15, 8910. 34 The Shiji attributes the five Proclamations (gao) to Zhou gong in Shiji 4.132– 133; cf. Shiji 37.1590. The “Shao gao” is not attributed to the Duke of Shao in Shiji 33.1519 or in Shiji 34, the “Hereditary House” of Lord Shao.

THE MANY DUKES OF ZHOU

103

reality, though it is worth noting in passing that Western Zhou bronze inscriptions and the Classic of Poetry up to now suggest that the Duke of Zhou in that period had not yet attained the monumental stature that he would later enjoy; perhaps he was less important to his contemporaries than the Duke of Shao (Shao gong 召公).35 The figure of the Duke of Zhou posed a string of moral dilemmas, as the duke allegedly (1) set up new systems of rites and music, an activity normally seen as the prerogative of a true king;36 (2) built a capital city, whose very name of Perfected Zhou (Chengzhou 成周) proclaimed his own right to such kingly prerogatives as a co-founder of the Zhou dynastic fortunes; (3) accepted the tribute and gifts due a king, not only from other lords, including the Duke of Shao (nominally his co-regent), but also from foreign envoys, such as the Yueshang 越嘗; (4) personally offered sacrifices to the High Lord (Shangdi 上帝);37 (5) undertook punitive campaigns on his own initiative;38 and (6) was buried with all the honors due a king.39 In addition, the duke, though enfeoffed in lands some thousands of miles away from the capital, left the state of Lu to his son to administer, preferring 35 Students of the Documents concede that “we cannot know for sure” the early biographical details of the life of the duke, nor can we reconstruct at this remove what really happened in early Western Zhou. See Nivison 1995, 190; also Shangshu jishi 157, which records the Shiji’s authors’ opinion that, even if the proclamation was made in the king’s name, “it was really the Duke of Zhou’s idea.” Seeking a kernel of historical truth in late sources, a group of PRC scholars argue against the view, promoted by Gu Jiegang, that the Duke of Zhou “proclaimed himself king” (cheng wang 稱王), however. See Yang Chaoming 2002. On the bronze inscriptions, see Shaughnessy 1997; Nivison 1995; Ma Chengyuan 1998; Zhu Fenghan 2006. Mao Ode 262 (“Jiang Han”) names the Duke of Shao (it does not mention Zhou gong) as chief supporter of the Zhou royal house; cf. Ode 265. The Duke of Zhou is mentioned in only two odes: Mao 157 (a complaint from the “Guofeng” section) and Mao 300 (in the Lu Hymns). The so-called “Confucius on the Odes” (“Kongzi shi lun” 孔子詩論) (the title was chosen by the modern editors) included in the Shanghai manuscript cache also focuses on the generosity and effectiveness of the policies of the Duke of Shao, linking them to the “Sweet Pear” (“Gan tang” 甘棠; Mao 16) and “Green Clothes” (“Lü yi” 綠衣; Mao 27) odes. See Shanghai bowuguan, vol. 1, 145; Chao Fulin 2003. 36 A number of Han texts, including the White Tiger Hall Discussions (Bohu tong 白虎通), attribute the invention of posthumous names (potentially critical of hereditary rank) to the Duke of Zhou. See also the “Shifa jie” 謚法解, Yi Zhoushu, chap. 54. 37 As the Liji’s “Mingtang” chapter puts it, “Lu…observed the royal ceremonies” (是故魯王禮也。). 38 A vassal is not supposed to undertake campaigns or make policy decisions on his own. See Shangshu dazhuan zhuzi suoyin 5.9/18/8. 39 King Cheng gave the Duke of Zhou the same burial rites that he accorded King Wen, as he “did not dare treat the Duke of Zhou as a minister” (不敢臣周公). See Shiji 33.1522.

104

CHAPTER THREE

to remain in or nearer the center of power right up until his death at a ripe old age.40 It is small wonder, then, that a great many persons, including some purported sages, suspected him of treason, as recounted in the Documents. The Documents, after all, depicts the Duke of Zhou acknowledging quite selectively the achievements of his elder brother, King Wu, the actual founder of the Zhou.41 All these acts of commission and omission probably complicated positive assessments of the duke’s role in the establishment and consolidation of Western Zhou power—all the more so if King Cheng was by no means a mere toddler, as in some legends, but rather someone old enough to participate in military campaigns early in his reign, as a few early bronze inscriptions suggest.42 The duke ascends the Documents stage in the “Metal Coffer” chapter, whose events are dated to a time about two years after the conquest of Shang by Zhou. King Wu lies gravely ill. In consequence, two of his chief advisors decide to consult the former kings through divination about the dire prospects of the dynasty.43 The Duke of Zhou dissuades them from ascertaining the will of the ancestors, insisting that “it would not be right at this point to distress our former kings” (未可以戚我先王). As soon as the advisors leave the scene, however, the duke “on his own initiative” (zi 自) builds three altars, at which he sacrifices and prays to the very former Zhou kings whom the two advisors had proposed to contact. The duke’s prayers to his own ancestors, including his own father, are couched in a very curious, almost comical appeal: the duke offers to have King Wu’s grave illness transferred to himself, even if the transfer entailed the duke’s death, on the triple grounds that he, the Duke of Zhou, has been the better son to 40 The duke gives audience to vassal lords in many texts, including the “Mingtang wei” chapter of the Liji quoted above. Several of the texts the duke supposedly authored were critical of the powers-that-be—hence, what Nivison (1995, 190) calls the “insinuation in unfriendly Eastern Zhou thinkers.” 41 Matsumoto (1966) insists on this point (see below). 42 Shaughnessy 1997, 107. 43 The advisors are generally taken to be Taigong Wang 太公望 (by a pun, “Grandfather’s Expected [Advisor]”) and the Duke of Shao 召公, neither a brother nor a son of Zhou kings, as it happens. Taigong Wang is assumed to have been an official of the last king of Shang, who then served King Wen of Zhou while he was Western Protector, laying the groundwork for the Zhou conquest. Shao gong was a relative of the Zhou founders who shared the cognomen of Ji 姬, though later traditions, such as that recorded in Wang Chong’s Lunheng (comp. ca. 90 CE), made Shao gong a half-brother to the Duke of Zhou. Guo Weichuan (2001, 263) believes that tradition is verified by the “Shao gong” chapter of the Documents.

THE MANY DUKES OF ZHOU

105

King Wen; that he, the duke, is more capable of serving the spirits well, being more intelligent and more talented than his elder brother, King Wu; and that King Wu should not be removed, lest the perpetual sacrifices to the ancestors be disrupted by supporters who (inexplicably?) credit Wu with putting the Zhou dynasty on a sound footing. By such pleas did the Duke of Zhou seek to wrest a “new mandate” and a “new life tenure” (both xin ming 新命) from the ancestors for his elder brother. Meanwhile, according to the “Metal Coffer” chapter, the duke swore to secrecy the few who were witnesses to his negotiations with the ancestors in Heaven. The very next day the king’s condition improves markedly, but somehow King Wu dies not long afterward anyway. The surviving brothers of the Duke of Zhou—entirely ignorant of the duke’s noble impulse to sacrifice himself for the good of the Zhou ruling house— suspect the duke of having designs upon the young son of King Wu: “The duke will do no good to the young son” (公將不利於孺子) they believe, at which point the Duke of Zhou proceeds to execute or exile his brothers, depending on the degree of their guilt. In justification, he claims that he “will have no way to make his reports to the former kings” (我無以告我先王) unless he “lays down the law” (我之弗辟), implying that the Zhou house would otherwise collapse in turmoil. The rest of the chapter has the Duke of Zhou conquering the rebels, then pleading with King Cheng upon his triumphal return that the king “not destroy his [the duke’s] nest” 44 in retaliation for the duke’s harshness in meting out justice to the king’s own uncles. With all due speed, Heaven moves to dispel any doubts in the young king’s mind about the duke’s loyalty, so that by chapter’s end, Heaven and the “Metal Coffer” have joined to “display the virtue of the Duke of Zhou” (zhang Zhougong zhi de 彰周公之德). The events in the “Metal Coffer” both precede and postdate those covered in the next Documents chapter, wherein the Duke of Zhou makes his first great public speech. Therefore, the “Metal Coffer” chapter, in effect, acts as framing device for all the succeeding Proclamation chapters relating the duke’s deeds, and while there is little doubt that the “Metal Coffer,” a late chapter, was written to defend the duke, the defense never whitewashes three of the more dubious aspects of the duke’s character and deeds: (1) the duke, determined to monopolize all communications with his powerful ancestors in 44

The plea appears in Mao Ode 155, “The Owl,” according to the Mao tradition.

106

CHAPTER THREE

Heaven, prevents his fellow regents, Taigong Wang and the Duke of Shao,45 from contacting the ancestral powers; (2) the duke does not hesitate to claim considerably more virtues and talents than his elder brother King Wu, not to mention a better rapport with his ancestors, the former kings, especially King Wen; and (3) the duke also persuades his reluctant co-regents that he alone must leave the capital to quell the rebellion, lest he otherwise lose the support of the former kings. (The two co-regents are left behind to dutifully undertake damage control during the crisis.) Significantly, then, the “Metal Coffer” story concocted about the duke’s “secret virtue” (yin de 隱德) does not fully absolve him from the implicit charges of arrogance, attention-grabbing, and ruthlessness, despite the final resounding approbation of both the young king and Heaven. The “Metal Coffer” chapter excites our curiosity, predictably enough. How did the Duke of Zhou actually perform when in virtual command of the dynasty’s fortunes, speaking for the king, according to the received traditions?46 The succeeding Proclamations chapters offer a consistent portrait of the duke addressing the great multitudes of lords, knights, and administrators who are under the direct command of Zhou (as with the conquered Yin 殷 masses under their leaders) or in the states allied with it (some of which are related by blood or marriage to the Zhou royal house).47 In these formal addresses, the duke employs a variety of humble phrases that draw attention to the youth and inexperience of King Cheng.48 He also makes frequent mention of the wrath of the High Lord, who may swiftly exact retribution for every misstep by the Zhou. The former kings received the Mandate, whose commands the duke must now proceed to execute.49 Notably, in expressing his abject obedience to the will of the “former kings” (xianwang 先王) and “previous generations” (xian [zu] 先[祖]), the duke appeals to the 45

See n. 35. For the phrase wang ruo yue 王若曰, see n. 85 below. 47 Chunqiu jingzhuan yinde, 427/Zhao 28/5 says that King Wu enfeoffed fifteen of his brothers (older and younger) and forty members of the Ji clan. 48 See “Da gao” for “our young and immature person, who has inherited infinite lands and great responsibilities” (嗣無疆大歷服) (Shangshu tongjian 27.0028–38; Legge 1895, vol. 3, 362). Gu Jiegang (1998, 37) remarks that such phrases do not refer to the youth of King Cheng; instead, they are “usual [polite] phrases used by the Son of Heaven to refer to himself.” 49 The “Da gao” characterizes Heaven in this way: “Heaven sends down harm to our house, not delaying the slightest [its harsh punishments]” (天降割于我家，不少 延！) (Shangshu tongjian 27.0018–26; Legge 1895, vol. 3, 362). 46

THE MANY DUKES OF ZHOU

107

lofty figure of King Wen, his dead father, whose spirit informed his own upbringing and mature reflections. (By contrast, he seldom refers to King Wu, perhaps because Wu was not thought to be equally powerful; after all, he died soon after conquering Yin.)50 It was King Wen, the duke insists, who left him the precious turtles to facilitate communication between the duke and the most potent spirits in Heaven.51 Such direct communications with King Wen, as well as the good king’s example, convinces the duke that he must summon the necessary courage, despite the evident dangers, “to complete King Wen’s merit” (成文王功).52 Not coincidentally, the most powerful metaphors in the Proclamations tell of fathers who plan and sons who bring those plans to fruition;53 the father-son pair is equated with King Wen and the duke far more plausibly than with King Wu and his rightful heir, King Cheng. Through such equations, acts that were typically condemned in Chinese history—acts of fratricide and arrogation of political powers by nominal subordinates—came paradoxically to be 50 As Olivier Venture has shown in an unpublished paper (2005), references to the former kings that include Wen and Wu on Western Zhou bronze inscriptions tend to be late. The “Da gao” mentions King Wen eleven times (with two additional references to the “former cultivated person(s),” qian ning ren 前寧人, that probably refer to King Wen) and King Wu only once. Aside from the “Luo gao,” where King Wu (usually after King Wen) is mentioned six times, no proclamations ascribed to the duke mention King Wu, with the possible exception of “Kang gao,” where a line talks of “my elder brother exerting himself” (寡兄勗), which some take to refer to King Wu and some to the Duke of Zhou himself. The “Luo gao” may mention King Wu repeatedly because (1) it is addressed by the duke to King Wu’s son; and (2) according to traditional accounts, the duke made this pronouncement only months before handing over the reins of government to King Cheng, and he was possibly in the presence of that king when he made his speech. Only rather slighting references are made to King Wu in the “Jun shi” chapter. See Shangshu tongjian 36.0444, 0456, 0471; Legge 1895, vol. 3, 482. 51 See, for example, the “Da gao” (Shangshu tongjian 27.0091–0100; Legge 1895, vol. 3, 365), where the duke says that King Wen has “left to me [us?] the great precious turtle [for divination], so as to bring Heaven’s light [to me]” (寧{= 文}王遺我 大寶龜，紹天明). 52 “Da gao” (Shangshu tongjian 27.0288–0307; Legge 1895, vol. 3, 369). King Wen provided an example while alive of divining and following the oracles, as the “Dao gao” says (Shangshu tongjian 27.0328–0333; Legge 1895, vol. 3, 369). He also worked tirelessly, as the duke notes in the “Da gao” (Shangshu tongjian 27.0370– 0376; Legge 1895, vol. 3, 370). Another example of King Wen’s excellent policies are reported in the “Kang kao” (Shangshu tongjian 29.0074–0082; Legge 1895, vol. 3, 383), which says that he “did not dare despise the widows and widowers, he employed those who were of our group, he revered all those worthy of reverence, he struck awe in the hearts of those who needed it,” and so on. 53 In the “Da gao,” several metaphors derive from this theme.

108

CHAPTER THREE

praised as miraculous reversions to the virtuous course that provided the basic rationale for the conquest by the Zhou overlords.54 Throughout these Proclamations, it is the Duke of Zhou who commands center stage, insofar as it is he who “first lays the foundations” (其基作民明辟) for the new Zhou dynasty, in both the institutional and physical senses, and who then endeavors to complete and perfect the Zhou legacy as well. It is he who first quells the rebellions that arise in almost all quarters of the empire following the untimely death of King Wu,55 it is he who forbids the Zhou princes and officers to take their solemn duty to offer sacrifices as a cynical pretext to overindulge in wine,56 and it is he who supervises the other high-ranking officers, including the Duke of Shao, in building the new city at Luo 洛, while putting in place the major dynastic institutions, prohibitions, and rituals.57 Let us tarry just a moment over the Documents description of the Duke of Zhou at Luo, which shows him arriving at the site of the new city after the Duke of Shao has already carried out preliminary surveys and requisitioned the necessary labor force: The Duke of Zhou surveyed the plans [drawn up by Shao gong] for the new city. On the dingsi day, three days later, he offered two bulls as victims in the suburbs, and on the following day of wuwu, he sacrificed to the gods of the soil in the new city, using one bull, one goat, and one pig. After seven more days, on the morning of jiazi, the first day of the sixty-day cycle, the Duke of Zhou then held court and gave written charges to the various Yin groups, and to the elders of the Hou, Dian, and Nan states. As soon as he had ordered the Yin people [to undertake 54 The passage from Xunzi, “The Efficacy of Ru” (“Ru xiao” 儒效), that I quote at the end of this chapter makes the duke’s unfilial and unbrotherly behavior epitomize filial and fraternal duty. A Zuozhuan passage makes the “Duke of Zhou’s virtue” “the reason why the Zhou became kings” (周公之德與周之所以王也). See Chunqiu jingzhuan yinde, 346/Zhao 2/1 Zuo. The Gongyang commentary is uninterested in the Duke of Zhou, except when a temple dedicated to him in Lu burns down. NB: The “Ru Xiao” chapter is the one pre-Han source attesting to the duke’s “claims to be king/calling himself king” (cheng Wang 稱王), but Lu’s hereditary prerogative to conduct royal sacrifices to the gods, including the suburban sacrifice, clearly derives from the duke’s status, as does the duke’s burial with royal honors, and hence receives no criticism in the early sources. 55 See Matsumoto 1968. The “Da gao” mentions only unrest in the western lands and the rebels to the east, but later legends multiply the dangers confronting the dynasty. 56 This is the subject of the duke’s “Proclamation about Wine” (“Jiu gao” 酒誥). 57 See the opening lines to the “Kang gao” (Shangshu tongjian 29.000–0080; Legge 1895, vol. 3, 381), which are clearly misplaced and probably represent an alternative opening narrative for the “Luo gao” (as Su Shi 蘇軾 argued) or the “Shao gao” (as Nivison [1995, esp. 184–185] argues).

THE MANY DUKES OF ZHOU

109

the construction work], they set about doing their work on a grand scale. The Senior Protector [the Duke of Shao] then went out with the leaders of the various states to collect rich presents, and when he [Shao gong] entered [the area] again, he presented them to the Duke of Zhou, saying, “I bow and kowtow, setting these out to the king-favored duke. You are to make the proclamations to the various Yin groups and you are to manage the affairs.” 周公朝至于洛則達觀于新邑營。越三日丁巳用牲于郊，牛二。越 翼日戊午乃社于新邑，牛一，羊一，豕一。越七日甲子周公乃朝 用書命庶殷侯甸男邦伯。厥既命殷庶，庶殷丕作。太保乃以庶邦 冢君出取幣，乃復入，錫周公。曰，拜手稽首。旅王若公。告誥 庶殷。越自乃御事。58

The unsurpassed power and prestige of the Duke of Zhou underpins this unctuous speech uttered by his co-regent. (More than once the Proclamations give the impression that King Cheng is too pusillanimous himself to finally “fix the Mandate” [ding ming 定命] for Zhou.)59 The Duke of Zhou then purportedly responds with the following proclamation: “I, Dan say this: Let a great city be constructed here, and may the king from this place function as co-adjutor to Heaven.… Dwelling in the new city, let the king now earnestly cultivate the virtue of reverence” (旦曰：「其作大邑，其自時配皇 天。 宅新邑，肆惟王其疾敬德。王其德之用，祈天永命。」). Thereupon the duke proceeds to instruct King Cheng that the building of the new city at Luo will allow him “from this time henceforth” (其 自時) to be commander of all. Completion of the city, in other words, will inaugurate the real reign of King Cheng, who will be “newly appointed [as king]” (xin pi 新辟). At this, the Duke of Shao once more pays obeisance to his co-regent, announcing his readiness to “maintain and receive his majesty’s dread command and brilliant virtue” (保受 王威命明德) and belittling his own contributions as but a minor “assist” (gong 供) to the prayers offered to Heaven by the Zhou.

58

This passage is now in the “Shao gao” (Shangshu tongjian 32.0087–0189; Legge 1895, vol. 3, 423–429). If the character zi 自 (self) is retained, as it is in many editions, the sentence still more emphatically insists that the Duke of Zhou take charge. As the Chinese text does not specify the speaker here, there was some debate over who is bowing, kowtowing, and speaking. Nivison (1995) makes a good case that the speech is uttered by the Duke of Zhou, contra Qu Wanli (Shangshu jishi 174n13). Nivison curiously asserts, however, that the “traditional understanding” of the “Shao gao” ascribes it to the Duke of Shao. See Shiji 4.133, 33.1519; and the preface to the Documents. 59 The king’s hesitation is registered in “Da gao” and in “Luo gao.”

110

CHAPTER THREE

The “Proclamation at Luo” chapter features a series of illuminating exchanges between King Cheng and the Duke of Zhou, wherein the duke continues to instruct the king, despite his growing maturity. It is striking that King Cheng repeatedly bows and kowtows to the duke while he humbly begs for this instruction,60 until eventually the king issues an invitation for the two of them to continue to share all power and responsibility for the dynasty together.61 Only after this invitation does the duke instruct the king as to which orders to give and how (“Let the king instantly give orders…Let him also command…” [今王 即命 惟命曰]).62 And finally the duke goes so far as to threaten the new king with ruin should the king flout his teachings.63 In this “Proclamation at Luo” chapter, then, the duke does not mince words when chiding the king; he threatens the king with errors, with insults, and with the loss of authority.64 Meanwhile, the duke never downplays his own unstinting efforts on behalf of the throne.65 King Cheng, for his part, seems anxious enough to retire to his old capital, leaving the duke in charge of sacrifices and ceremonies in the new capital. Twice, however, the king reminds the duke that the Mandate has come to him, the king, through the joint efforts of both King Wen and King Wu, in a subtle rebuke.66 Yet the duke concludes by claiming that it will be the good government of the new city of Luo under his direction that

60

Later in the “Luo gao,” the duke says that the king sent him two flagons of lack millet wine, with the message that he once again bowed and kowtowed to the duke. See Shangshu tongjian 33.0634–0655; Legge 1895, vol. 3, 449. 61 In the opening exchanges in that chapter, the king speaks of “we two people sharing [the task of] stabilizing [the rule]” (我二人共貞), “since the duke has made provision for me for ten thousand years and more” (公其以予萬億年敬天之休) (Shangshu tongjian 33.0118–0129; Legge 1895, vol. 3, 438). 62 See “Luo gao” (Shangshu tongjian 33.0171–0190; Legge 1895, vol. 3, 439). 63 The duke says that, if he fails, the consequences will be like a fire that “blazes up and eventually cannot be extinguished” (Shangshu tongjian 33.0213–02225; Legge 1895, vol. 3, 440). 64 “Luo gao” (Shangshu tongjian 33.0304–0309; Legge 1895, vol. 3, 441): “Affairs will be troubled by errors and insults [to you]!” (惟事其爽侮); cf. “If you do not bestir yourself in such things, you will not continue long, surely” (汝乃是不蘉，乃時惟 不永哉) (Shangshu tongjian 33.0325–0335; Legge 1895, vol. 3, 442). 65 “Luo gao” (Shangshu tongjian 33.0308–0316; Legge 1895, vol. 3, 442): “Do you, my son, extend [or “proclaim” or “bestow”] my unwearied efforts” (乃惟孺子頒 朕不暇). 66 “Luo gao” (Shangshu tongjian 33.0434, 0663; Legge 1895, vol. 3, 444–445, 452). In the concluding lines of the chapter (Shangshu tongjian 33.0757–0765; Legge 1895, vol. 3, 452), a passage that feels tacked on says that the duke then conscientiously fulfilled the decree of Kings Wen and Wu.

THE MANY DUKES OF ZHOU

111

will finally succeed in establishing the king as “the new leader of the four quarters” (為四方新辟).67 The rhetoric of the Proclamations section goes well beyond a formal doctrine of “ministerial indispensability” (the phrase is Nivison’s)68 to a recognition of the duke’s paramount authority. Other chapters in the Zhou section of the Documents only serve to underscore this picture of the duke’s unique status. Alone of the early Western Zhou legendary figures, the Duke of Zhou appears as exemplary author and chronicler of the past;69 the duke repeatedly exhorts members of the Zhou governing elite to consult great men’s biographies as “mirrors” (jian 鑑) of historical patterns.70 He it is who decides that his younger brother Kang is to be invested with one-half of the Yin lands formerly in the hands of the rebels. He it is who freely interprets for others the will and intentions of the High Lord, the chief god of Shang, as well as of Tian, or “Heaven,” the chief deity of Zhou.71 He it is also who seeks to assimilate the Shang people, practices, and advisors to Zhou ways, so that the Yin people may more speedily be resigned to serving the Zhou cause.72 At the same time, no attempt is made to downplay the ruthlessness of the duke when executing the rebel leaders, since the duke is shown exhorting his king to apply the capital punishments to all evildoers; by the duke’s logic, such people merely “hasten their own guilty verdicts” (惟民自速辜).73 Pardons, in the duke’s view, are to be reserved for inadvertent mistakes.74

67

“Luo gao” (Shangshu tongjian 33.0581–0584; Legge 1895, vol. 3, 447). Nivison 1995. 69 The Duke of Zhou composed Mao Ode 155; he is also given as the author of the “Offices of Zhou” (“Zhouguan”) chapter of the Documents and parts of the Changes, not to mention the many chapters described here. 70 “Jiu gao” (Shangshu tongjian 30.0515–0523; Legge 1895, vol. 3, 409–410), where min 民 is more likely to refer to the “king’s men” than to commoners. Cf. “Kang gao” (Shangshu tongjian 29.0181–0189; Legge 1895, vol. 3, 386). Two chapters (“Jiu gao” and “Duo shi”) attributed to the duke employ the phrase “from Tang the Victorious down to Di Yi,” producing a sweeping overview of history, as does part of “Lord Shi” (Shangshu tongjian 36.0171–0291; Legge 1895, vol. 3, 477–478). 71 See “Kang gao” (Shangshu tongjian 29.0107–0110; Legge 1895, vol. 3, 385). 72 The “Kang gao” reveals this intention in many passages, for Prince Kang is to use the Shang-Yin penal code and customs. 73 “Jiu gao” (Shangshu tongjian 30.0495–0499; Legge 1895, vol. 3, 409). At the close of the “Jiu gao,” the Duke of Zhou asks that those who continue to indulge themselves be put to death (Shangshu tongjian 30.0599–0613; Legge 1895, vol. 3, 411). 74 “Kang gao” (Shangshu tongjian 29.0316–0330; Legge 1895, vol. 3, 388). 68

112

CHAPTER THREE

Other chapters in the Documents do little more than embellish this picture drawn by the Zhou Proclamations. The “Catalpa Timber” chapter, for instance, continues the theme of coordinating the efforts of many disparate groups in society by placing the primary emphasis on securing the loyalties of the people and ministers and great families, thereby putting the sovereign last.75 It meanwhile paints a vivid picture of the ideal ruler as one who models himself on his ministers, faithfully executing their policy proposals, so as to “lead the people to the enjoyment of plenty and peace” (肆王惟德用和懌先後迷民). The true sovereign is to use every tool in his repertoire to attract the lords’ affection to himself, for this will please his august forebears, the former kings.76 The same instructions are delivered in a rather heavyhanded fashion in the “No Ease” chapter, which admonishes the king “from this day forward” (今日) to eschew the wanton pleasures that will incur the blame of Heaven above and the people below; only an idiot of a young king would choose to ignore such dire warnings. Similarly, the “Lord Shi” chapter, supposedly composed to dispel the “displeasure” (出于不祥) felt by the Duke of Shao toward the Duke of Zhou, has the latter freely admitting that he is having some difficulty in setting the dynasty on a right course.77 But this admission serves only to buttress the chapter’s main argument that, with so many dangers pressing round the throne, it is too dangerous a moment to usher in any real change within the governing elite. Besides, the Zhou have always had strong ministers assisting the rulers’ advance toward full possession of the Mandate.78 Thus, King Cheng’s failure to ascend the throne cannot be “blamed on” (無我責) the first Duke of Zhou79— hence the cunning plea by the Duke of Zhou that the Duke of Shao believe him when he, the Duke of Zhou, says that the dynasty’s fate rests on the two of them continuing to work together in perfect harmony toward their common goal. (Contrast King Cheng’s plea to share the government with the duke.) And indeed, it is the Duke of 75

See the opening paragraph of “Zicai” (Shangshu tongjian 31.000–24; Legge 1895, vol. 3, 413). 76 “Zicai” (Shangshu tongjian 31.0115–0118; Legge 1895, vol. 3, 416) and (Shangshu tongjian 31.0231–36; Legge 1895, vol. 3, 418). 77 “Jun shi” (Shangshu tongjian 36.0125–0134; Legge 1895, vol. 3, 477). Legge thinks that the duke is complaining here of his inability to correct King Cheng, while Shaughnessy (1997, 111) has the Duke of Zhou admitting here that he should not be ruling in King Cheng’s stead. 78 “Jun shi” (Shangshu tongjian 36.0446–0453; Legge 1895, vol. 3, 482). 79 “Jun shi” (Shangshu tongjian 36.0496–0505; Legge 1895, vol. 3, 482).

THE MANY DUKES OF ZHOU

113

Zhou who is finally credited with “Establishing [Zhou] Rule” (立政) in the traditional readings, thanks to his excellent judgment in staffing the court’s offices. Still, the Proclamations fail at many points to specify who is acting for whom and for what reason. Readers may surmise that the duke there acts and speaks with a view to securing his own power, justifying his frequent resorts to force as the only way to strengthen the foundations of Zhou rule. At other points, the writings and speeches ascribed to the duke tend to gloss over the issue of precisely who in the Zhou royal line has received the Mandate from Shang-Yin, King Cheng alone, the duke himself, or all the members of the entire Ji clan born to positions of authority and nominally following the king’s lead.80 Moreover, on not a few occasions, readers cannot ascertain who is where. To take one example, the final passage in the “Proclamation at Luo” chapter does not make clear whether it is the king or the Duke of Zhou acting in his name who is offering sacrifices in the new city of Luo. And, as soon as canon, commentary, and prefaces are read together, the confusions proliferate: First, it is clearly the Duke of Zhou who executes or exiles the rebels, according to the “Metal Coffer” and Proclamations chapters, but the preface to the Documents talks unambiguously of King Cheng “smiting his uncles” (伐管叔， 蔡叔).81 Second, King Cheng bows and kowtows to the duke, and the duke responds with the same gestures, though such polite expressions of obeisance more typically occur when a minister has an audience with a king or when equals meet on formal (and potentially hostile) ground.82 Third, the duke is not averse to using the “royal we” (zhen 朕) of himself, even when reporting on his activities to King Cheng, nor is he averse, apparently, to dubbing himself “a young son” (xiaozi 小子), presumably to humbly indicate his relative inexperience and gross ineptitude in governance.83 This sort of practice makes the true referents of many such expressions (including “weak man,” chong ren 沖人) virtually impossible to discern.84 Fourth, the commentators often cannot decide whether the phrase “The king has said to the effect” 80

The “Jiu Gao” (Shangshu tongjian 30.0045; Legge 1895, vol. 3, 401) provides but one instance where the duke says that the Yin Mandate came to “us” (wo 我). 81 See the preface to the “Kang gao” (Shangshu tongjian 51.0720–0728). 82 See Li Zhenxing 1994, vol. 1, 296–302, on bowing and kowtowing. As Li shows, five times in the “Luo gao” the duke and the king exchange bows and kowtows. 83 See n. 39. For one unambiguous case where the Duke of Zhou calls himself a “young son,” see “Jun shi,” par. 3 (Legge 1895, vol. 3, 477). 84 See Guo Weichuan 2001, 259.

114

CHAPTER THREE

(wang ruo yue 王若曰) refers to King Cheng or the Duke of Zhou (Shao gong is never deemed the speaker of such phrases, however).85 And fifth, while King Cheng resides in the old capital of Zhou (i.e., Hao 鎬), the Duke of Zhou resides in a new capital at Luo whose name, “Fulfilling [or Perfecting] Zhou” (Chengzhou 成周), transparently implies that it is the duke’s actions—not the king’s—that have secured or will secure the Mandate.86 When one adds to this the contents of the “Lord Shi” (“Jun Shi” 君奭) chapter in the Documents, which takes as its starting point the suspicions of the Duke of Shao about the Duke of Zhou’s motives, we can understand why some early and modern scholars concluded that the Duke of Zhou “continually” (chang 常) proclaimed the “king’s Mandate,” monopolized power, and even failed to report as subordinate to King Cheng.87 Defenders of the duke’s legend derived from the various Documents the all-important timetable for the duke’s initiatives, to settle all charges and countercharges regarding the duke’s role in early Western Zhou. The timetable proposed for the duke’s initiatives is as follows, counting from the death of King Wu: Year 1: rebellion by Wu Geng 武庚, Guanshu 管叔, and Caishu 蔡叔. Year 3: victory over the remnant Yin subjects. Year 4: the duke located at Yan 奄.88 Year 4: established Kangshu 康叔 as marquis in Wei.89 Year 5: [began to?] build Chengzhou.90 85 Zheng Xuan’s commentaries to the “Da gao” specifically say that in the phrase wang ruo yue, the “king” is Zhou gong. See Li Zhenxing 1994, vol. 1, 247–258, for an overview of this issue. There are three possible translations for wang ruo yue: (1) “the king has said to the effect”; (2) “as king, he has said”; and (3) “the king said approvingly.” The third translation was suggested by Luo Zhenyu (1914), who argued that ruo means nuo 諾 (to approve). This interpretation is followed by both Shaughnessy (1997) and Falkenhausen (2006/Luo Tai 2006). 86 Note, meanwhile, that later legend makes it Zhou gong who places the Nine Tripods symbolizing Zhou rule in Luo, the city he oversees. See Shiji 4.132. 87 For this phrasing, see Dushu oushi, fu 讀書偶識附 6b; cf. Dushu oushi, fu, 11a (only 1 juan). 88 Either the duke or King Cheng says in the “Duoshi” chapter (Shangshu tongjian 38.0008–0011; Legge 1895, vol. 3, 461) that “formerly we came from Yan,” a statement that has led most commentators to locate Yan somewhere near the old Zhou capital. Shaughnessy (1997, 123) argues that Yan is equivalent to Gai 蓋, an area near Qufu 曲阜, “that is, within the state of the Lu which was the fief of Zhou’s eldest son Qin.” 89 The Bamboo Annals dates this to the third year of King Cheng, after the defeat of Wu Geng, a Shang prince. 90 However, the Shiji puts this in the seventh year, evidently thinking the construction occupied only a few months. See Li Zhenxing 1994, vol. 1, 310.

THE MANY DUKES OF ZHOU

115

Year 6: instituted rites and composed music (during which time away from Hao he presumably had the leisure to author the Zhouli and plan the building of the Spirit Hall).91 Year 7: completed the city of Luo, after which the duke purportedly “returned the government to King Cheng” 成王.

Defenders of the duke invariably reiterated their belief that the duke eventually “returned the government” (fan zheng 反政), while the duke’s detractors pointed to the speed and number of his initiatives as a sure sign of his overweening ambitions. The Duke of Zhou in the Yi Zhoushu According to tradition, King Wen succinctly formulated the central problem for the nascent Zhou dynasty before the conquest of ShangYin: how is the Zhou ruling line to protect the fledgling dynasty so as to ensure that its rule continues over generations to come?92 The Duke of Zhou supposedly advised King Wen that the last Shang king was losing support because he “insults his ministers, makes the common people suffer, and treats the vassal lords badly” (欺侮群臣，辛苦百 姓，忍辱諸侯), staffing his offices with the very worst sort of people. To provide the perfect counterexample, King Wen should implement the Five Goods (wu xiang 五祥), which are (1) the ruler’s care in choosing officials to represent him; (2) the officers’ subsequent adherence to bureaucratic models and rules; (3) continuing efforts to increase agricultural productivity; (4) the weeding out of all corruption from administration; and (5) the profound reassessment of the sources of the people’s difficulties. Moreover, the duke continued, a true king intent on building the dynasty’s fortunes should untiringly elevate to high rank the benevolent, the wise, the courageous, the talented, the cultivated, and the managers, for putting these sorts of men in high office will encourage other honest and hard-working people to devote their efforts unstintingly to ensuring the dynasty’s well-being. Additionally, in order to maximize the state’s resources, the officials should see that the farmers and bureaucrats alike strictly observe the

91 This timetable tallies with Shangshu dazhuan 5.12/20/11–12 (on “Luo gao”) but contradicts Shangshu dazhuan 5/8/17/22–25, which makes the duke’s success in building Luo the necessary precondition for his decision to turn to the larger task of instituting new rites and music, after three years of hesitation. 92 “Feng bao,” Yi Zhoushu 207.

116

CHAPTER THREE

seasons for agricultural activities while allowing no piece of land to remain uncultivated. The enumeration of these goals in the Yi Zhoushu offers us our first glimpse of governing policies in predynastic Zhou. We see a king and his son working in close concert to achieve a new set of aims quite unlike those said to preoccupy the duke in the Documents chapters. (Notably, King Wen does not consult Fa 發, his elder son known to history as King Wu, and King Wu, after the Shang conquest, continues the family tradition by consulting again and again with the Duke of Zhou.)93 The people and wealth of the realm are to be mobilized toward the achievement of a common goal,94 with the ends apparently justifying the means, even when those means are not high-minded or pretty. For example, the duke urges the king to include among his methods “bribing [rival states’] prayer masters” (神巫靈寵以惑之; 厚 其禱巫), so that the plans of the Shang overlords may be known, and “broadcasting talk of one’s virtues” (liuliu de 流德), the better to advertise the present evils of the moribund Shang-Yin.95 If this sort of trap is not quietly and carefully laid for the enemy, including one’s liege lord, Zhou power will quickly collapse—or so the duke asserts. “From power, one may then show sympathy” (cong chuan nai wei 從 權乃慰) to the vanquished.96 Thus, each generation is to instruct the next in the sorts of plots and counterplots designed to strengthen the ruling house.97 A second early strategy session, this time arranged between the future King Wu and the Duke of Zhou after the death of King Wen, outlines a series of aggressive moves called “soft strikes” (rou wu 柔武), which take aim against ineptitude and indolence within the Zhou court. Arguably, the very best war is the one in which one never needs to use force, as such military classics as the War Techniques of Master Sun (Sunzi bingfa 孫子兵法) explain, so to implement such moves is ideal. 93 King Wu also consults with Shao gong 召公 and Bi gong 畢公, according to “He wu” 和寤 (Yi Zhoushu 350). 94 “Feng bao,” Yi Zhoushu 211–212. 95 “Feng bao,” Yi Zhoushu 219. Later commentators (e.g., Zhu Youzeng 朱右曾, 1846) hasten to say that the sage would never use such hostile plans to overturn another’s state; he merely wants his own descendants to know that unscrupulous rivals use such methods (“Feng bao,” Yi Zhoushu 221). This moralizing is hard to accept, given the tenor of the entire Yi Zhoushu text, which many have compared to the Guanzi. 96 “Feng bao,” Yi Zhoushu 223. 97 “Feng bao,” Yi Zhoushu 224.

THE MANY DUKES OF ZHOU

117

The most serious lapse of judgment would be to ignore the timing of agricultural activities, but unjust litigation, a preoccupation with the harem, reliance on powerful favorites, and the king’s love of travel all invite disaster. The wise ruler is advised therefore to “always take impressive character (de) as the basis, duty as the technique, trust as the motivating factor, victory as the heart’s desire, resoluteness as the plan, and restraint as a success…in order to harmonize and equalize the villages and hamlets, so as to correct the people’s miseries” (故必以德 為本，以義為術，以信為動，以成為心，以決為計，以節為勝 和均里以匡辛苦。).98 By such methods, gradual but inexorable victory over one’s rival will be achieved, without resorting once to expensive and dangerous military actions. King Wu has reason enough for concern: he is sore afraid that the Shang king has heard the rumors of Zhou’s disaffection. It is therefore left to the Duke of Zhou to stiffen his resolve to launch his expedition against Shang. This the duke accomplishes through two rousing exchanges recorded in two Yi Zhoushu chapters—“The Great Opening to War” (“Da Kaiwu” 大開武) and “The Secondary Opening to War” (“Xiao Kaiwu” 小開武)—urging the king to stay the course and complete the undertaking begun by their deceased father, King Wen. One clever speech finds the duke comparing war to farming: “[Founding a dynasty] is like preparing the fields for farming. If one works hard at tilling but then fails to hoe, only weeds will take root there; and, when autumn comes, there will be no harvest. The beasts will feed on them [the starving] and the people will have reaped a famine. And then who will mourn them when they are gone?” (若農之服田。務耕而不耨， 維草其宅之。既秋而不穫。維禽其饗之。人而獲饑。去誰哀 之。). 99 Reassuring King Wu that Zhou enjoys possession of the Mandate, the duke insists that his brother need only display due diligence in planning the insurrection for the Shang to taste defeat. Preparations then need entail no marshaling and drilling of troops, activities that would only further arouse the suspicions of the Shang king. Rather, the Zhou should concentrate on putting their own house in order,100 a task, by the duke’s account, that requires the king not to favor any one branch of his relatives or one set of clients over another, 98

“Rou wu,” Yi Zhoushu 269–270. “Da Kaiwu,” Yi Zhoushu 286. 100 The surprising weakness of the Shang-Yin house, which fell after only one month’s assault by Zhou, is invariably attributed to the Shang ruler’s inability to put his own house, court, and administration in order. 99

118

CHAPTER THREE

while ignoring baseless rumors circulating in the back palace. The king is meanwhile to see that he and his court avoid any appearance of guilt, since rival states routinely exploit Nine Pretexts to justify their unlawful invasions, chief among them neglect in offering sacrifices to the gods, lapses in virtue, failures to hire the most talented men for office, and mismanagement of the administration of the realm.101 With staffing and oversight issues resolved, the king must finally amend his own conduct, lest he be charged with one or more of the Ten Excesses (shi yin 十淫), because lapses at court (especially those in the king’s person) are apt to be multiplied and magnified in the conduct of the “people below” (xia min 下民). Long before force becomes necessary, the king must accurately gauge the limitations and constraints of his situation: the patterns of the stars, the terrain of the land, and the physical capacities of his troops.102 A continual refrain in the duke’s counsel enjoins Fa to note the change of seasons and abide by the rules of timely engagement, lest the would-be king invite disaster and increase the people’s miseries and feelings of alienation.103 The panic felt by the future King Wu has been mollified only in part, unfortunately; he remains virtually immobilized by his fear of spies. What if the bad last king of Shang has discovered the Zhou plot to overthrow him? When King Wu seeks counsel from the Duke of Zhou yet a second time, the duke calmly announces that the time for insurrection has arrived.104 Accordingly, the king must immediately open hostilities by encouraging trade, lowering the price of grain, and disbursing the necessary funds to secure allies, so as to amass the necessary human resources at every level of society for the upcoming battles.105 Through such advice, the duke gradually reveals his strategic 101

“Da Kaiwu,” Yi Zhoushu 280. “Xiao Kaiwu,” Yi Zhoushu 290, where these are identified as the Three Limits (san ji 三極). 103 In this respect, according to the commentators, the duke’s speeches, especially those in “Xiao Kaiwu,” resemble the “Hong fan” chapter of the Documents. (Sometimes it is difficult to tell who the speaker is in these chapters, the king or the Duke of Zhou, but since the king is consulting the duke, I have taken the major speeches to be the duke’s.) 104 “Feng mou,” Yi Zhoushu 314–315. Cf. the opening lines of the following chapter, “Wu jing,” Yi Zhoushu 320. 105 In “Da ju,” Yi Zhoushu 417, the duke outlines a plan by which the former Shang counselors are to be induced to enter Zhou service through a reward program. The duke is taken as a model of frugality in the Han Feizi 20/40/23–26, and his frugality is modeled on the seasons, where periods of leanness precede and prepare for the abundance of summer. 102

THE MANY DUKES OF ZHOU

119

plans for the attack, which will allow the Zhou “to spread its wings” (fu yi 傅翼) after a massive victory.106 As before, the king has only to catch the resolute tone of the Duke of Zhou’s words for him to bow to their wisdom. “Ah, how true!” (yun zai 允哉) is his response.107 This stress on timely action, comparatively muted in the Documents, is quite pronounced in the Yi Zhoushu.108 One of the duke’s many injunctions reads: “If it is not the right time, then there can never be model talk!” (非時罔有恪言).109 The good ruler invariably “lays plans in response to the exigencies of the time” (維日望謀), knowing that regrets for lost opportunities will not markedly improve a situation once it has turned unfavorable.110 Unfortunately, it is never easy to determine the most opportune time for decisive action. To perceive unfolding realities as they really are, the would-be ideal ruler must first undertake to “settle” (ding 定) his mind, equanimity being the precondition for useful insights about timing in relation to all major initiatives.111 Then, from his lofty and detached vantage point, as if from Heaven itself, the ruler must watch the progress of the agricultural year, so that he brings all his activities into compliance with the changing seasons, as mandated by the cosmic order, “for only then can good rule flourish” (政乃盛行).112 The king must learn, for instance, that “in the third month of summer, in streams and marshes, do not cast nets, in order to allow fishes and turtles to mature fully” (夏三 月川澤不入網罟，以成魚鱉之長。). In the end, productivity in his realm depends as much upon “discerning the benefits of yin and yang” (別時陰陽) (i.e., determining the cycles for each activity) as upon adjusting farming methods to varying types of soil or using irrigation when needed.113 In this way, “living creatures will not lose what is appropriate to them; the myriad things will not lose their natures; people will not fail in their affairs; and Heaven will not lose its seasons” (萬物不失其性。不失其事。人天不失其時。).114 The duke continues in this vein:

106

“Wu jing,” Yi Zhoushu 323. As in “Da ju,” Yi Zhoushu 433. 108 Yanaka (1999) emphasizes this aspect of the Yi Zhoushu narrative. 109 “Xiao Kaiwu,” Yi Zhoushu 294. 110 “Feng mou,” Yi Zhoushu 317–318. 111 “Baodian,” Yi Zhoushu 298. 112 “Baodian,” Yi Zhoushu 308. 113 “Da ju,” Yi Zhoushu 414. 114 “Da ju,” Yi Zhoushu 431. 107

120

CHAPTER THREE Mountains, forests, thickets, and marshes are to be relied upon [for their resources]. Craftsmen and artisans are to be put to work on state projects, so that their talents may create achievements. Merchants and traders are to hasten to the markets, so that they may join forces [by being collected together]… If there are slopes, ditches, roads, pathways, hemp patches, hills, and mounds that cannot be planted with grain, plant them with trees suitable for timber. In spring, distribute dried goods [seeds?]; in summer, greens and leafy foods. In autumn, distribute fruits and vegetables; in winter, firewood and kindling. Thus one may relieve impoverishment and difficulties. Consolidate the strengths of the people and have them act as instructors to one another. Rely on the advantages of their land to provide all the people’s material goods… This we call “humane and imposing virtue.” 山林藪澤以因其。工匠役工，以政其材。商賈趣市，以合其用 陂溝道路藂苴丘墳不可樹榖者樹以材木。春發枯槁，夏發葉榮， 秋發實蔬，冬發薪烝，以匡窮困。揖其民力，相更為師。因其土 宜，以為民資 此謂仁德。115

Of course, timely action has yet another dimension, as commentators to the Yi Zhoushu are fully aware: each person in the state must carry out his specific responsibilities at his or her societal level, if the state is ever to be unified and power centralized on the model of the cosmos.116 Once again, the Yi Zhoushu vision of good government is strongly shaped by Zhanguo concerns. The Yi Zhoushu repeatedly refers to the great utility to be had from inducing fear in others, in contrast to the Documents, where the Duke of Zhou is more apt to express anxiety lest his own conduct and that of King Cheng fall short of the high standards set by their immediate forebears.117 For instance, the “Treasured Canons” (“Baodian” 寶典) chapter lists “filial duty” (xiao 孝) as the first of nine virtues to instill in the ruler’s subordinates, because “only if sons are truly afraid will they not plot to create chaos [i.e., topple the throne]” (子畏哉, 乃不亂 謀).118 Rewards and punishments are to be doled out in such a way that the “correct order” (zheng 正) results and the people “can be put to use” (wei yong 維庸), at which point the ten sources of treachery

115 “Da ju,” Yi Zhoushu 427, before the ellipsis; 428–429 after. The Duke of Zhou cites as precedent for his stress on timeliness the “Prohibitions of Yu” (禹之禁; apparently a sort of “Yueling” 月令 essay, otherwise unknown), on 430. 116 See “Baodian,” Yi Zhoushu 312. 117 See, e.g., “Li zheng” (Shangshu tongjian 31.0579–0584; Legge 1895, vol. 3, 519). 118 “Baodian,” Yi Zhoushu 299.

THE MANY DUKES OF ZHOU

121

can be identified and nipped in the bud.119 The duke’s evident reliance on harsh punishments may seem excessive to those who imagined that his rule of “great peace” in the name of King Cheng epitomized compassion and benevolence (ren 仁), but deterrence—“killing one person to benefit the empire” (殺一人而以利天下)—turns out to be the revised definition of “benevolence.”120 Pardons are to be reserved when seeking allies in preparation for battle.121 Still, generous rewards offset such draconian measures, in the full knowledge that loyalty can be bought. By the carrot and the stick, then, Zhou’s new subjects will be coaxed to submit to its rule.122 No less crucial to the early Zhou success, according to the Yi Zhoushu, is the duke’s division of the subject population into uniform mutual-aid and mutual-surveillance groups along lines very similar to Shang Yang’s 商鞅 famous reforms in the mid–fourth century BCE— although Shang Yang would be excoriated in late Western Han sources by the classicists and Confucians (Ru 儒) claiming to represent the Zhou Way.123 Of course, the Yi Zhoushu puts a kindlier construction on these policy measures, stressing the mutual aid facilitated by mutual subservience; rigorous organization of the subject population is necessary if the government is to care for the “old, weak, sick, and ill” (lao ruo ji bing 老弱疾病) and those without families.124 Nevertheless, under the duke’s reforms, the penetration of central government directives into the lives of villagers is to extend even to the family administration of marriages and mourning, to the management of domestic animals, to the close supervision of market activities, and to the training of local militias. Put another way, the Yi Zhoushu virtually eliminates the mid-level political structure vested in the vassal 119 “Baodian,” Yi Zhoushu 299–304. For “put to use,” see Yi Zhoushu 322. Such language, which appears frequently in the Yi Zhoushu, flatly contradicts Xunzi’s description of the duke as one who used punishments rarely, if at all. See Xunzi, “Zhongni,” 7.1 (trans. in Knoblock 1990, 59). 120 See “Taizi Jin,” Yi Zhoushu 1089, where this phrase is ascribed to King Wu. 121 “Wu jing,” Yi Zhoushu 323. Cf. “Baodian,” Yi Zhoushu 308, where Heaven’s way of giving birth and then killing things becomes the model for the ruler’s punishments and favors. 122 See “Huang men,” Yi Zhoushu 297–298. 123 For examples, see the arguments in the Yantie lun, which was written several decades after the debates it purports to transcribe verbatim; see Loewe 1993, 477–481; Nylan, forthcoming, chap. 22. The Han Feizi is one of many texts to show that a link between Zhonggong and Guan Zhong carried rhetorical force. See Han Feizi 37/119– 20/25–1. 124 “Da ju,” Yi Zhoushu 417, 419–420, 422, 424.

122

CHAPTER THREE

lords in the early Zhou period, opting instead for an anachronistic portrayal of ideal Zhou rule wherein the key to good rule evidently lies in the direct interactions between the throne and its commoners, so as to circumvent the power of the local lords—as under the Qin and early Western Han. The Yi Zhoushu’s portrait of the Duke of Zhou is emphatically that of a loyal and brilliant minister—with the duke’s subservience signaled by his frequent identification of himself as “subordinate” and “minister” (chen 臣) in place of the pronouns or noun phrases used often in the Documents to indicate the royal “we” (e.g., zhen 朕 and wo 我).125 But this is hardly the only reason that the Yi Zhoushu’s activist minister busily engaged in aggregating power at the capital looks more like a Guan Zhong 管仲, a Han Feizi 韓非子, or even a Shang Yang than any faithful minister appearing in the early Documents’ chapters. For instance, the Yi Zhoushu’s Duke of Zhou, like Han Feizi, suspects the reputedly pure and brave of traitorous intent.126 Furthermore, the respect that the Yi Zhoushu’s duke has for agriculture and the military as the twin engines of state power tallies perfectly with the sentiments expressed in Lord Shang’s treatise entitled “Agriculture and War” (“Nong zhan” 農戰).127 In contrast to the Documents, which casts the Duke of Zhou’s mighty endeavors to “establish good government” (li zheng 立政) as expressions of filial devotion to his ancestral line,128 the Yi Zhoushu firmly harnesses the duke’s merits to service the centralizing project directed from the capital for the particular benefit of the living person of the king.129 Whereas the Documents’ duke spoke of appeasing the gods and ancestors in Heaven, the newly minted duke sees in sacrifice yet another opportunity for manipulation. In a limited sense, then, the Yi Zhoushu’s duke paves the way for the 125

See, e.g., “Baodian,” Yi Zhoushu 310. That the Yi Zhoushu credits King Wu, rather than the Duke of Zhou, with not only the conquest of Shang but also the later stabilization of Zhou rule is made abundantly clear in “Da ju,” 413: “It was King Wu who conquered the Yin, pacified the state, and settled the people.” The entire “Great Booty” (“Shi fu” 世浮) chapter (Yi Zhoushu chap. 40) ascribes to King Wu all the achievements required to establish Zhou rule. Many texts link the duke to Guan Zhong, including Hanshi waizhuan 4/3, which discusses types of loyalty. 126 “Baodian,” Yi Zhoushu 303. 127 See Shangjun shu, chap. 3. 128 One thinks of the “Jin teng” line, “If I do not rule these men, I shall not be able to report to the former kings” (Shangshu tongjian 26.0288–0299; Legge 1895, vol. 3, 358). The main thrust of the “Shao gao” is that the king should trust the advice of the “old and the wise” (a description that fits the Duke of Zhou perfectly). 129 See, e.g., “Shi fu,” Yi Zhoushu 443.

THE MANY DUKES OF ZHOU

123

famous theory of “Heaven and Humans Are One” (tianren he yi 天人 合一) attributed to Dong Zhongshu 董仲叔 (ca. 179–ca. 104 BCE), while embracing policies identified with the Legalists.130 Ideally, to trace the changing assessments of the Duke of Zhou’s character and achievements, historians would compare the two different portraits summarized above with the Great Commentary to the Documents (Shangshu dazhuan 尚書大傳), a Han compilation ascribed to Fu Sheng 伏勝 (fl. 221–ca. 175? BCE) at the beginning of Western Han. Curiously, however, many of the key sources relating to the duke in the Modern Script Documents find no counterpart in that Great Commentary at all.131 No commentary exists, for example, for the influential “Lord Shi,” “Charge to Bi” (“Bi ming” 畢命), and “Establishing Rule” chapters, and the extant commentary contains only a few choice anecdotes relating to the Duke of Zhou or his famous pronouncements, nearly all of which are known from other early compilations. No less interesting is the failure of the Great Commentary, in company with other Western Han texts, including Mr. Han’s Outer Tradition of the Poetry (Hanshi waizhuan 韓氏外傳), to make little if any effort to devise a logically consistent portrait of the duke, aside from insisting on the duke’s unfailing subservience to the young King Cheng, a point of great interest to the sponsors of the new imperial ideology.132 Given the frequency and ferocity of Western Han debates weighing the political advantages accruing from meritocratic claims versus hereditary privilege in the new realm, it should come as no surprise that the Duke of Zhou is hailed as sponsor of both competing positions, sometimes even within the same text. Two early texts, the Springs and Autumns Annals of Mr. Lü (Lüshi chunqiu 呂氏春秋, comp. 239? BCE) and Mr. Han’s Outer Tradition of the Poetry, provide variants on a single story recording such debates and placing them in the mouths of Taigong Wang and the Duke of Zhou (there an advocate of hereditary privilege):

130

For a new interpretation of Dong’s imperial theory, see Nylan 2008. For Dong’s changing place in the Han scholastic imaginaire, see A. Cheng 1998–1999. 131 Altogether the duke appears in eleven chapters of the Shangshu dazhuan text, which probably does not in its entirety date to the time of Fu Sheng. Zheng Xuan (127–200) is the major commentator for it. 132 For example, Shangshu dazhuan 6.2/28/1–3 specifically states, “The emperor [i.e., King Wu(?)] ordered Zhou gong to become regent… The Duke of Zhou assisted the young ruler” (帝命周公踐阼 周公輔幼主).

124

CHAPTER THREE Taigong asked the Duke of Zhou, “How do you intend to rule Lu?” The Duke of Zhou replied, “I will honor those already in honorable places, and treat my close kin with special affection (zunzun qinqin).” Taigong remarked, “If Lu adheres to this policy, it will grow weak.” The Duke of Zhou asked Taigong, “So how do you intend to rule Qi?” Taigong replied, “Promote the worthy and reward [only] the deserving.” The Duke of Zhou remarked, “In later generations, your princes will certainly be assassinated!” 太公問周公何以治魯。周公曰，尊尊親親。太公曰，魯從此弱矣 。周公問太公何以治齊。太公曰，舉賢賞功。周公曰，後世比有 劫殺之君矣。133

If this passage has the duke placing a high priority on kinship ties as key to good rule, another anecdote recorded, with variations, in the same two texts underscores the wisdom of “promoting the worthy” (ju xian 舉賢) by tying that slogan to the duke’s activities: During the seven years that the Duke of Zhou occupied the place of the Son of Heaven, there were ten poor gentlemen to whom he gave presents and whom he treated as teachers. There were twelve men to whom he made return presents and regarded as friends, and forty-nine from poor dwellings in mean quarters to whom he gave precedence in interviews. There were a hundred good men whom he advanced at regular times; there were a thousand teachers and ten thousand officers. 周公踐天子之位七年，布衣之士所執(贄)而師者十人。所友見者十 二，窮巷白屋子[所]先見者四十九人，時進善者百人，教士者千， 宮朝者，萬人。134

Then a third description in the later of the two texts, the Outer Tradition of the Poetry, seamlessly weaves together the passages supporting the two opposing stances in a single narrative of the good duke’s activities.135 Rhetorical malleability is a hallmark of political maneuvering in late Zhanguo and early Western Han (times of major political upheavals), when those in power naturally sought to put themselves forward as men able to ride the waves of change. The duke, if his model was to 133

Compare Lüshi chunqiu 11/57/9–12 and Hanshi waizhuan 10/25 (numbered after Hightower 1952), where the order of speeches is changed; see also Huainanzi 11/2a. This story appears in many compilations, including the Garden of Persuasions (Shuiyuan 說苑). The phrase “treating kin with special affection” (qin qin 親親) is linked to the duke in Zuozhuan (Xi 24/2 Zuo) as well. 134 Compare Lüshi chunqiu 15.3/83/11–12 and Hanshi waizhuan 3/31, where the numbers are different. 135 Hanshi waizhuan 8/32.

THE MANY DUKES OF ZHOU

125

remain relevant to succeeding generations of persuaders, would also have to appear as a master of transformations.136 Thus, a typical formulation from that period highlights the duke’s ability to assume high rank and just as calmly to relinquish it, while retaining his imposing authority and charisma: Confucius said, “In olden days, when he served King Wen, the Duke of Zhou had nothing arbitrary in his conduct and there was nothing selfwilled in the affairs he managed. It was as though his body could not bear the weight of his clothes nor his mouth utter words [so diffident was he]. When he received anything directly from King Wen, he was cautious, as though he feared he would drop it. Truly one can say of him that he was capable of acting as a son should. “When King Wu died, King Cheng was young, so the duke took over the work of Kings Wen and Wu. He then occupied the place of the Son of Heaven and controlled the government of the empire, took measures to deal with the unrest among the Yi and Di, and punished Guanshu and Caishu severely for their crimes. Holding King Cheng in his arms, he received homage from the lords. Punishments and rewards, he instituted and decided without advice from others. His prestige shook Heaven and Earth; his gestures terrified the empire. Truly one can say of him that he was capable of acting as a martial figure should. “But when King Cheng reached adulthood, the Duke of Zhou turned the rule over to him and, facing north [in the position of a loyal subject], was subservient to him. He asked permission before acting and never once wore a self-aggrandizing air. Truly one can say of him that he was capable of acting as a subject should. Really, a man capable of three such transformations in his own person is capable of adapting himself to the changing times.” 孔子曰，昔者周公事文王，行無專制，事無由己，身若不勝衣， 言若不出口，有奉持於前，洞洞焉若將失之，可謂[能]子矣。武王 崩，成王幼，周公承文武之業，履天子之位，聽天子之政，征夷 狄之亂，誅管蔡之罪，抱成王而朝諸侯，誅賞制斷，無所顧問， 威動天地，振恐海內，可謂能武矣。成王壯，周公致政，北面而 [臣]事之，請然後行，無伐矜之色，可謂[能]臣矣。故一人之身， 能三變者，所以應時也。137

From such an accomplished master of changing social roles, it was but one more step to make the duke into the full-blown god known from the late Western Han apocryphal texts attached to the Documents. In those apocrypha, the Duke of Zhou easily outshines his brother, 136

Other legendary figures who fulfill this role include Liuxia Hui 柳下惠, who is less famous today than the duke today, but who may have been equally famous in Western Han. 137 Hanshi waizhuan 7/4 (mod.).

126

CHAPTER THREE

King Wu, attracting to his refined person all manner of auspicious signs, including great pairs of phoenixes, a miraculous “time plant” (ming jia 蓂筴) marking off the days, blue-green dragons, and auspicious cloud formations, not to mention the large turtle rising forth from the waves of the Yellow River on whose back red and green inscriptions foretold both the ascendancy of Zhou and the later dispensation that would end in the Han triumph, thanks to the patronage of the uncrowned king Confucius.138 By such rhetorical moves would the Duke of Zhou, once virtual king of the early Western Zhou, figure simultaneously as alter ego of the faithful mid- to low-ranking men in service to the throne (shi 士), in whose company Confucius included himself,139 and as the image of the true god and savior of Central States culture. Conclusion This essay has surveyed several pseudo-historical constructions put upon the single figure of the Duke of Zhou over the course of the centuries spanning mid-Zhou to late Western Han. It has relied mainly on texts dating from before late Western Han, when the Rituals of Zhou briefly attained canonical status, and even longer before Zheng Xuan’s commentaries to the Rituals of Zhou forged the Rituals into one of the three Rites canons. The foregoing analysis suggests, first, the protean character of the monumental figures within storytelling traditions in China and, second, the very large areas of overlap between strands of thought later identified as opposed (e.g., Legalist vs. Confucian). And while readers may find it astonishing, given the duke’s later exalted status, to find so much evidence attesting to a more troubling story, the Yi Zhoushu is precisely the sort of text where we would expect to find such evidence. Considered only briefly for canonical status, the Yi Zhoushu almost certainly underwent fewer editorial revisions than the Documents after Han times.

138 Yasui and Nakamura 1971–1994, esp. 87–88, 103, citing Shangshu zhonghou 尚書中候. 139 As a group, the shi were originally restricted to the lower-ranking members of the aristocracy with the right to bear arms, but by Han times the shi included officers below Rank 18 and even the lettered among commoners. For the brilliance of the duke’s virtual reign redounding to the credit of the shi, see Zhanguo ce zhuzi suoyin 136B/68/8.

THE MANY DUKES OF ZHOU

127

Both the Duke of Zhou and Confucius provided a “usable past” amenable to unceasing renewal and renovation. Clearly, the revised Han portraits of the duke in such para-canonical works as the Yi Zhoushu and the Great Commentary to the Documents—especially the latter—endeavor to cast each deed ascribed to the Duke of Zhou as faithful service rendered to a larger political abstraction exemplifying the new political virtue of “acting in the public interest” (gong 公), where that interest was equated with centralizing efforts. The Qin and Han courts, busily engaged in their own centralizing efforts, resorted to such portraits of ideal men as ready props for imperial power.140 Doubtless the elevation of the Duke of Zhou was coupled in many minds during the classical era with that of another celebrated figure of Lu, Kongzi, who was credited with predicting (and perhaps helping) the rise of the Han. Much was made of the fact that the two men supposedly lived some five hundred years apart, since, by prevailing theories of the time, sages arose at just such intervals. The duke apparently handed Heaven’s Mandate over to the next sage in his home state.141 No wonder the Confucius of legend tied the success or failure of his bold political program to the frequency with which the duke appeared in his dreams.142 These two men meanwhile figured as the principal authors of numerous compilations that came to be included in the Five Classics corpus. In essence, then, both men from Lu were “plain kings” (suwang 素王), nominal subjects who exerted kingly power and authority in their own lifetimes and beyond. Both legendary figures drew criticisms for precisely this elevation, for each purportedly performed potentially treasonous acts, including instituting new systems of rites and music for later generations and pronouncing judgments on bad or inept hereditary rulers. 143 (How cunning, then, to put serious critiques of the duke’s conduct into the mouth of Confucius in pre-Han accounts.)144 It could hardly escape 140

So speculates Majima (2005, 2). Shiji 130. 142 Analects 7/5. 143 See, e.g., Zuo, Ding 11, fu Zuo ii. Mencius 3A/4 speaks of Chen Liang 陳良, a native of Chu, who delighted in (yue 悅) “the Way of Zhou gong and Zhongni [i.e., Kongzi]” (周公仲尼之道). See n. 36 above for the Duke of Zhou as inventor of posthumous names. 144 See the Shizi, xia/22: “Long ago, the Duke of Zhou returned the government [to King Cheng]. Confucius criticized this, saying, ‘Was Zhou gong not a sage? On account of his abdication of the empire, he failed to act on behalf of the masses’” (昔周 公反政，孔子非之，曰：「周公其不聖乎！以天下讓，不為兆人也。」). Cf. 141

128

CHAPTER THREE

the notice of Han historians, then or now, that the good duke’s name was regularly invoked whenever talk turned to irregular successions and regencies (male or female). The most famous of the Han episodes concerns the linkage forged between the Duke of Zhou and the socalled usurper Wang Mang, the subject of another essay in this volume.145 But similar associations pepper the standard histories for the classical period146—hence the necessity for Xunzi 荀子 (ca. 335–238 BCE), the most supple persuader to be numbered among the classicists employed by Qi,147 to boldly address the most troublesome aspects of the duke’s story, converting them to virtuous deeds: Since he [the Duke of Zhou]…once had the empire but then he no longer held it [after returning the rule to King Cheng], the duke’s actions did not constitute a “usurpation.”… The transfer of power took place in an orderly, methodical fashion within an appropriate span of time. Thus, even though his cadet branch of the family supplanted the main line, this did not constitute an “irregularity,” nor did the younger brother’s execution of an older brother [Guanshu] constitute a “violent coup.” And though ruler and minister exchanged positions, this did not constitute an “act of treason.” Since for the sake of the peace of the empire, the duke had carried on the tasks of Kings Wen and Wu,… the whole known world lived together in peace as though but one. No one but a sage could have managed this! This is what we mean when we talk of an “achievement of a great classicist 儒.”148 周公 鄉有天下，今無天下，非擅也。 變勢次序〔之〕節然也。 故以枝代主而非越也。以弟誅兄而，非暴也。君臣易位而非不順 也。因天下之和，遂文武之業 天下厭然猶一也。非聖人莫之能 為。夫是謂大儒之效。

Mozi yinde 65/39/60, which shows Confucius besmirching the reputation of Zhou gong while teaching his own disciples to rebel. I will not discuss coupled critiques of Zhou gong and Confucius, such as that offered by Ban Si 班嗣 (late Western Han), who perceived the life promoted by the Duke of Zhou and Confucius to be a straitjacket rather than a “great Way” (da dao 大道). For Ban Si, see Hanshu 100A.4205. 145 See, e.g., Hanshu 84.3426–30, 99B.4087, 4099. Michael Puett’s essay (chapter 4 in this volume) explores this topic in depth. 146 For example, the Duke of Zhou comes up in discussions beginning with Chen She’s rebellion, with the Huainan rebellion, and with the events of 91–90 BCE, when an heir was killed; with Huo Guang; with Dou Xian’s 竇憲 (d. 92 CE) regency in Eastern Han (Hou Hanshu 23.795), with the regent Liang Ji 梁冀 (d. 152 CE), and so on. For Zhou gong and rebellion, see Shiji 58.2090, 118.3077. 147 There is some evidence that the strongest defenders of the duke’s loyalty were drawn from the ranks of the classicists of Qi and Lu, where the duke functioned, after all, as “local hero.” 148 Xunzi, “Ru xiao” chapter; trans. in Knoblock 1990, 69.

CHAPTER FOUR

CENTERING THE REALM: WANG MANG, THE ZHOULI, AND EARLY CHINESE STATECRAFT Michael Puett, Harvard University In this chapter I address a basic problem: why would a text like the Rituals of Zhou (Zhouli 周禮), which purports to describe the administrative structure of the Western Zhou 周 dynasty (ca. 1050–771 BCE), come to be employed by Wang Mang 王莽 (45 BCE–23 CE) and, later, Wang Anshi 王安石 (1021–1086) in projects of strong state centralization? Answering this question for the case of Wang Mang, however, is no easy task. In contrast to what we have later for Wang Anshi, there are almost no sources to help us understand precisely how Wang Mang used, appropriated, and presented the Zhouli. We are told in the History of the Han (Hanshu 漢書) that Wang Mang employed the Zhouli, but we possess no commentaries on the text by either Wang Mang or one of his associates. In fact, we have no full commentary until Zheng Xuan 鄭玄 (127–200 CE), who was far removed from the events of Wang Mang’s time and was concerned with different issues. Even the statements in the Hanshu about how Wang Mang used the Zhouli—referred to as the Offices of Zhou (Zhouguan 周官)—are brief. We are told that Wang Mang changed the ritual system of the time to follow that of the Zhouguan,1 that he used the Zhouguan for the taxation system,2 and that he used the Zhouguan, along with the “The Regulations of the King” (“Wangzhi” 王制) chapter of the Records of Ritual (Liji 禮記), to organize state offices.3 I propose to tackle this problem in a way that is admittedly highly speculative. I will discuss the argument of the Zhouli in relation to other claims being made about state organization in early China. This will still not, of course, explain how figures in the court of Wang Mang were reading and appropriating the Zhouli, but it will at least 1

Hanshu (“Jiaosi zhi”) 25.1265. All references to the dynastic histories in this chapter are to the Zhonghua shuju editions. 2 Hanshu (“Shihuo zhi”) 25.1180. 3 Hanshu (“Wang Mang zhuan”) 99.4136.

130

CHAPTER FOUR

help us speculate on the cultural resonance that the Zhouli might have had in the Han, why groups might have attempted to appropriate the Zhouli, and what implications such appropriations might have had at the time. The Organization of the Zhouli The Zhouli opens with the following claim: It is the king who establishes the state, distinguishes the quarters and rectifies the positions, structures the state and aligns the fields, sets up the offices and designates the functions. He thereby serves as the pivot for the populace. 惟王建國，辨方正位，體國經野，設官分職，以為民極。4

This statement is then repeated at the start of each new section of the text, and each section of the text describes how the king set up the administrative structure for one of the six divisions of the state. The first of these divisions concerns the officials in charge of administering the realm: He thereupon institutes the official for Heaven, the minister of the state, to employ and take charge of his subordinates and to supervise the regulation of the territories, so as to assist the king in ruling the territories and states. 乃立天官冢宰，使帥其屬而掌邦治，以佐王均邦國。5

The text then lists the officials under the minister, along with a short description of their functions. These are the “heavenly officials” (tianguan 天官). The next section concerns the “earthly officials,” charged with educating and training the populace. The structure is the same as for the heavenly officials: It is the king who establishes the state, distinguishes the quarters and rectifies the positions, structures the state and aligns the fields, sets up the offices and designates the functions. He thereby serves as the pivot for the populace. He thereupon institutes the official for Earth, the minister of the multitude, to employ and take charge of his subordinates 4 Zhouli (“Tianguan”) (Chinese University of Hong Kong, Institute of Chinese Studies Ancient Chinese Text Concordance Series [hereafter ICS]) 1.0/1/3. Here and throughout, my translations have greatly benefited from the excellent translation by E. Biot 1851. 5 Zhouli (“Tianguan”) ICS 1.0/1/3–4.

WANG MANG AND THE ZHOULI

131

and to supervise the teaching of the territories, so as to assist the king in pacifying and training the territories and states. 惟王建國，辨方正位，體國經野，設官分職，以為民極。乃立地 官司徒，使帥其屬而掌邦教，以佐王安擾邦國。6

And then the officials in charge of rituals are mentioned: It is the king who establishes the state, distinguishes the quarters and rectifies the positions, structures the state and aligns the fields, sets up the offices and designates the functions. He thereby serves as the pivot for the populace. He thereupon institutes the official for Spring, the minister of cult, to employ and take charge of his subordinates and to supervise the rituals of the territories, so as to assist the king in bringing harmony to the territories and states. 惟王建國，辨方正位，體國經野，設官分職，以為民極。乃立春 官宗伯，使帥其屬而掌邦禮，以佐王和邦國。7

The text continues with the officials for Summer (in charge of war) and the officials for Autumn (charged with the administration of justice and punishments). (The section on the officials for Winter is missing and was later replaced by the “Artificer’s Record” [“Kaogong ji” 考工記].) And that is the argument of the Zhouli. What is immediately striking about the text is what is absent. Completely lacking here is any concern with what might be called legitimation—with, for example, training the emotions and dispositions of the populace, with rooting state practice or ideology in cosmological patterns, or with seeking support for the state from divine powers. Absent, in other words, are the very things that so many other texts from early China (and elsewhere) dealing with political organization emphasize. The move of the text is instead simply to focus on the ruler, who serves as the pivot for the populace by establishing the center and organizing all activities under a clear hierarchy of officials. The goal is to take any practice of potential interest to the state, define an official to oversee it, and define the hierarchy within which that official will operate. Why would a text argue in this way, and what resonance might such claims have had in the Western Han? And why would they become so influential in later East Asian history? 6 7

Zhouli (“Diguan”) ICS 2.0/15/23–24. Zhouli (“Chunguan”) ICS 3.0/32/17–18.

132

CHAPTER FOUR

The Liji To begin to answer these questions, let us turn to some contrasting visions. I will examine another set of texts that would also become important in very late Western Han politics, namely several of the chapters of the Records of Ritual (Liji 禮記).8 Elsewhere I have argued that several chapters of the Liji—“The Movements of the Rites” (“Li yun” 禮運), “The Meaning of Sacrifice” (“Ji yi” 祭義), “The Method of Sacrifice” (“Ji fa” 祭法)—try to build a vision of statecraft through a particular reading of ritual. More specifically, the chapters argue that sacrifice creates familial feelings toward those outside one’s immediate family: deceased persons come to be thought of as ancestors, and the ruler comes to be thought of as the father and mother of the people and also as the Son of Heaven. In other words, participants in the ritual system come to see themselves as linked to the ruler and to parts of the natural world in chains of constructed genealogical continuity. It is an argument that assumes a relatively decentralized form of governance in which power is based on particular dispositions inculcated in the populace through participation in rituals.9 One of the most powerful examples of this can be seen in the “Li yun” chapter. The chapter opens with a dialogue between Confucius and his disciple Yan Yan 言偃. Confucius is lamenting his inability to put into practice the way of the ancients or the great figures of the Three Dynasties of Antiquity: The practice of the Great Way and the illustrious figures of the Three Dynasties—these I have not been able to reach. But my intent is to do so. 大道之行也，與三代之英，丘未之逮也，而有志焉。10

In particular, Confucius is regretting the fact that people think only in terms of their immediate kin, resulting in endless competition between these kin groups. According to Confucius, in early antiquity, everything was shared: In the practice of the Great Way, all under Heaven was public. They selected the talented and capable. They spoke sincerely and cultivated 8

For a study of the uses of the ritual Classics, see Qian Xuan 1996. Puett 2005, 2008. 10 Liji (“Li yun”) ICS 59/1.23–24. My translations from the Liji here and throughout have been aided greatly by those of Legge 1885. 9

WANG MANG AND THE ZHOULI

133

peace. Therefore, people did not only treat their own kin as kin, and did not only treat their own sons as sons. 大道之行也，天下為公。選賢與能，講信脩睦，故人不獨親其親 ，不獨子其子。11

This Great Way has since been, in part, lost: Now, the Great Way has become obscure. All under Heaven is [divided into] families. Each treats only its own kin as kin, only their own sons as sons. 今大道既隱，天下為家，各親其親，各子其子。12

Confucius’s narrative places the point of discontinuity at the emergence of hereditary rule with the Xia dynasty. At this point, rulership comes to be defined as being within a family, and political power comes to be defined as competition between families—for dynastic rule at the highest level and for political influence and high ministerial positions at the next level of the hierarchy. To control the ensuing competition, ritual is used to try to control the populace: Ritual and propriety are used as the binding. They are used to regulate the ruler and subject, used to build respect between the father and son, used to pacify elder and younger brother, used to harmonize husband and wife, used to set up regulations and standards, used to establish fields and villages, used to honor the courageous and knowledgeable, taking merit as personal. Therefore, schemes manipulating this arose, and because of this arms were taken up. 禮義以為紀；以正君臣，以篤父子，以睦兄弟，以和夫婦，以設 制度，以立田里，以賢勇知，以功為己。故謀用是作，而兵由此 起。13

According to Confucius, six figures have been able to use ritual properly to counteract the decline: Yu, Tang, Wen, Wu, Cheng, and the Duke of Zhou were selected because of this. These six rulers were always attentive to ritual, thereby making manifest their propriety, thereby examining their trustworthiness, making manifest when there were transgressions, making the punishments humane and the expositions yielding, showing constancy to the populace. If there were some who were not following this, they were removed from their position and the populace would consider them dangerous. This was the Lesser Peace. 11

Liji (“Li yun”) ICS 59/1.24. Liji (“Li yun”) ICS 59.9/1.27–28. 13 Liji (“Li yun”) ICS 59.9/1.28–30. 12

134

CHAPTER FOUR 禹、湯、文、武、成王、周公，由此其選也。此六君子者，未有 不謹於禮者也。以著其義，以考其信，著有過，刑仁講讓，示民 有常。如有不由此者，在埶者去，眾以為殃，是謂小康。14

The key, therefore, is the correct use of ritual: Yan Yan asked again, “Are the rites of such urgency?” Confucius said, “Rites are what the former kings used to uphold the way of Heaven and regulate the dispositions of humans.” 言偃復問曰：「如此乎禮之急也？」孔子曰：「夫禮，先王以承 天之道，以治人之情。」15

When Yan Yan asks Confucius to explain, Confucius provides a narrative of the invention of ritual by the sages. I have summarized the full narrative elsewhere,16 but the key I want to emphasize in the argument here is that the inventions of the sages allowed humans to move beyond the primitive world in which they originally lived, to survive the natural world, and to have a ruler to organize the world. But those same inventions also resulted in a destruction of the original unity of humanity. Thus, rituals were created to allow for a re-creation (but never a perfect one) of the ideal sets of relationships that existed in deep antiquity. As the “Li yun” argues, the consequence of these rites of sacrifice—families developing proper filiality through sacrifices to ancestors, the ruler sacrificing to Heaven and thus defining himself as both the Son of Heaven and the father and mother of the people—is that the entire realm comes to function again as a single family: Therefore, as for the sage undertaking to treat all under Heaven as one family and to treat the central states as one person, it is not something done overtly. He always knows their dispositions, stimulates their sense of propriety, clarifies what they feel to be advantageous, and apprehends what they feel to be calamitous. Only then is he capable of bringing about [the unity of the populace]. 故聖人耐以天下為一家，以中國為一人者，非意之也，必知其情 ，辟於其義，明於其利，達於其患，然後能為之。17

In short, sacrifice allows the sage to consolidate his rule by molding the dispositions of the populace, leading the people to think of the realm as a single family. The sage is thus able to rule effectively, since 14

Liji (“Li yun”) ICS 59/9.1/28–32. Liji (“Li yun”) ICS 9.2/60.1 16 Puett 2005. 17 Liji (“Li yun”) ICS 62/9/22. 15

WANG MANG AND THE ZHOULI

135

he is seen, not as wielding arbitrary power, but rather as the father and mother of the people. In other words, ritual, when it is effective, is not seen as a tool to control the populace (in which case, as the beginning of the text makes clear, it just leads people to scheme against it) but rather comes to function covertly, through the dispositions of the populace. This, according to the text, was the key for the rule of figures like the Duke of Zhou. The Argument of the Zhouli It is striking to note how radically distinctive the Zhouli approach is. Unlike the sacrifice chapters of the Liji, the concern of the Zhouli is neither to build up genealogical connections through ritual nor to develop a notion of statecraft based upon the dispositions of the people. The text rather operates from dramatically different premises. The opening concern of the text is that the world is inherently disordered, and the goal of the ruler is thus to designate a series of functionaries who will bring about order in the economic, administrative, and ritual realms. In the realm of ritual, the concern is to construct, not genealogical connection, but rather administrative order: to place the activities of the ritual specialists within a defined organizational structure. Both the Zhouli’s opening claim and how it is to be put into effect thus differ from the principles expressed in the sacrifice chapters of the Liji. In making this point, I am not necessarily claiming that, taken as a whole, the arguments of the Zhouli fully conflict with the arguments concerning rituals found in the Liji (although for many of the chapters of the Liji I think they do). Indeed, it is immediately clear how later commentators who wished to read these texts as related could do so: both the Zhouli and the chapters of the Liji mentioned above are based upon a claim of centering. In both cases, the ruler centers the realm and thereby creates order in an otherwise fractured and discontinuous world. But, unlike most of the chapters of the Liji, the Zhouli is well suited to support a strong centralized state. As we saw, the sacrifice chapters of the Liji focus on ritual as a means of refining and encouraging certain attitudes among the populace. Although this does not preclude a centralized form of statecraft, it does imply that a successful form of governance would be based not upon strong centralized

136

CHAPTER FOUR

institutions but rather upon the populace coming to think of the ruler in familial terms. In contrast, the Zhouli proceeds from the claim that successful governance is based upon taking any given human activity and giving it a proper place within an institutionalized order. If the sacrifice chapters of the Liji construct chains of continuity from the refined familial tendencies of humans, starting with the family and ultimately encompassing the ruler and the larger cosmos, the argument of the Zhouli proceeds by beginning with the ruler and then working out from there, ultimately (ideally) encompassing everything. Familial dispositions and genealogical continuity are of no interest. Debates over Political Order in the Late Warring States Period Having laid out briefly some of the main claims of the Zhouli and of the sacrifice chapters of the Liji—two sets of texts that became important at the end of the Western Han—it may now be helpful to situate these claims in a historical context. Many of the debates about statecraft that developed at the end of the Western Han can best be understood in relation to the forms of state centralization that developed in the late Warring States period (403–221 BCE), came to a head in the reign of the First Emperor of Qin 秦 (r. 221–210 BCE), and were consolidated during that of Emperor Wu 武 of the Han 漢 (156–87 BCE). The forms of centralized statecraft that began developing in the late Warring States period were clearly recognized at the time to be unprecedented. This sparked a huge debate as to whether or not such institutions were legitimate and, if not, precisely how they could be legitimated. It is quite possible that the Zhouli itself,18 as well as several chapters of what would later be collected into the Liji, date to this same period.19 Although, as we have seen, both texts came into prominence 18 Kang Youwei 康有為 famously argued that the Zhouli was probably a forgery by Liu Xin 劉歆 (46 BCE–23 CE), a figure closely associated with Wang Mang. Subsequent work has demonstrated that in fact at least much of the material of the Zhouli—and thus possibly the Zhouli itself—dates from the Warring States period. For an excellent overview of the proposed dates for the composition of the Zhouli, see Boltz 1993. For more specialized studies, see Karlgren 1931; Broman 1961; Qian Xuan 1996; Jin Chunfeng 1993. 19 Even if the Liji itself was probably not compiled until the Western Han (202 BCE–9 CE), many of the chapters certainly are much earlier. For discussions of the dating of the text, see Riegel 1993. More recently, one of the chapters of the Liji, the “Black Robes” (“Ziyi” 緇衣) was discovered in the Guodian tomb, which was sealed

WANG MANG AND THE ZHOULI

137

at the end of the Western Han dynasty, the contents of both were probably composed much earlier. As for the Zhouli, Mark Edward Lewis has argued: “Despite its reputation as a ritual text, a Confucian classic, and collection of implausible offices, the Zhouguan is closely linked to the major legal and administrative reforms of the Warring States period.”20 But if the Zhouli and many chapters of the Liji date to this period, there is little evidence to suggest that their arguments had a significant impact at the time. Despite their many differences, both of these texts were associated with a return to the Zhou—a position that was losing favor at the time and would only continue to do so over the ensuing two centuries. The texts that were becoming prominent were those that supported unprecedented forms of centralized statecraft. A few examples of such texts will be helpful. Cosmological Order One such text was the Springs and Autumns Annals of Mr. Lü (Lüshi chunqiu 呂氏春秋), a text composed at the court of the Qin. Completed around 240 BCE, when the Qin were clearly in a position to conquer the remaining states, the text was an argument for one way of building an imperial order. The essential move was one of pure inclusivity: taking any intellectual position that existed and giving it a place within a larger, unified order. And the basis for this unified order was a claim concerning the larger order of the cosmos: the cosmos was a unified totality, and if the earthly realm were modeled on this cosmic order, then it would be a unified totality as well.21 The first section of the work organizes the many intellectual visions for ordering the self and state that had developed during the Warring States period and places them on the grid of the seasons, with spring being used for self-cultivation, summer for education, and so on. The claim was that each method was correct but partial; the goal around 300 BCE. This has sparked a general rethinking of the dates of many of the Liji chapters. For the Guodian find in general, see Hubeisheng Jingmenshi Bowuguan 1997, and for the implications of the find for our understanding of the Liji compilation, see Li Xueqin 1998. For a discussion of how the “Ziyi” was transformed into a chapter of what would ultimately come to be seen as one of the Classics, see Shaughnessy 2006, 63–130. 20 Lewis 1999b, 42. 21 On the Lüshi chunqiu, see S. Cook 2002; Sellmann 2002.

138

CHAPTER FOUR

was thus to give each view a place and to call on each to be implemented at the moment when such activity would be proper within the larger cosmic cycle. The Lüshi chunqiu itself, therefore, claimed to provide the system for unifying all other positions, and the basis for doing so was a claim concerning the order of the cosmos. The postface (xu yi 序意) to the Lüshi chunqiu makes the reasoning clear. To quote from the excellent translation by Knoblock and Riegel: On the day of the new moon, a good man asked about the twelve Almanacs. The Marquis of Wenxin replied: “I have succeeded in studying what the Yellow Sovereign used to instruct the Zhuanxu sovereign: ‘There is a great circle above and a great square below. If you are able to make them your model, you will be as father and mother to the people.’ You have probably heard about the ancient age of purity. This was due to following the model of Heaven and Earth.” 朔之日，良人請問《十二紀》。文信侯曰：「嘗得學黃帝之所以 誨顓頊矣，爰有大圜在上，大矩在下，汝能法之，為民父母。蓋 聞古之清世，是法天地。22

The proper ordering of the state is one that is modeled on cosmic patterns. Such an argument certainly invites comparison with the Zhouli. To begin with, as we have seen, the Zhouli may well be a late Warring States text as well, and there are at least hints that it could be associated with the state of Qin. As Mark Edward Lewis has argued: “The reference to the sacrifices to the five di 帝, which recur under many offices, are particularly significant, because this was the most important cult in Qin. Thus accounts of cults reinforce the impression that the text was greatly influenced by Qin.”23 Like the Lüshi chunqiu, the Zhouli is an attempt to take everything that exists and define a place for it within a comprehensive system. Indeed, in many ways the Zhouli shares much less with texts like the “Li yun” that call for rulers to build up support covertly through the training of the dispositions of the populace than with texts like the Lüshi chunqiu that attempt to provide a fully systematic explanation of the workings of the state. Moreover, the organization of the Zhouli is based upon a model that sounds much like the first section of the Lüshi chunqiu: the political administration is divided into six realms, called Heaven, Earth, Spring, Summer, Autumn, and Winter. Indeed, 22 23

Lüshi chunqiu ICS 12.6/62/6–8. Translation by Knoblock and Riegel 2000, 272. Lewis 1999b, 44.

WANG MANG AND THE ZHOULI

139

Lewis has argued that the Zhouli should in fact be read as a text that, like the Lüshi chunqiu, is based upon a cosmological claim: “the rise of correlative cosmology in the late Warring States led to new models of governance based on imitation of natural patterns…. The Zhouguan reworked the old practice of listing Zhou officials in the light of this new theory of government, and thereby provided a model of the state that incorporated current cosmological theory, state cults, and administrative practice.”24 Lewis sees two bases for such a cosmological reading of the Zhouli. The first is that “the Zhouguan’s government operates on the principle that every office has a double function: administrative and religious. The authors of the text may not have recognized this distinction, but we must move through our own categories to reach an understanding of the alien world that produced such a work.”25 The second is that “the officers are organized to function as a symbolic reproduction of the structure and working of the cosmos. In this way the text offers one of the earliest and most elaborate versions of the idea, central to Chinese civilization, that the world is fundamentally congruent with a bureaucratic order.”26 Elaborating on this point, Lewis argues: “The text is arranged according to principles of numerology and ritual calendrics that emerged to prominence in the late Warring States period. The composition of the Zhouli was thus a ritual act that conjured into existence a graphic image of the state as cosmic mandala.”27 But there is a difference here between the arguments of the Zhouli and those of texts like the Lüshi chunqiu. Although it is certainly true that many of the administrative positions in the Zhouli involve activities that we would define as both political and ritual in nature, the text makes no claim about the nature of the spirits being sacrificed to. On the contrary, the sense seems to be simply that there are ritual specialists out there—and the Zhouli will explain how to order them. Whether or not their rituals are efficacious (either because of their effect on the spirits or because of their effect on the dispositions of the participants) is something on which the text expresses no interest. Similarly, although the administrative framework of the Zhouli does indeed involve a nomenclature of Heaven, Earth, and the seasons, here again the text makes no claim that there exists a cosmic pattern 24

Lewis 1999b, 47. Lewis 1999b, 43–44. 26 Lewis 1999b, 44. 27 Lewis 1999b, 45. 25

140

CHAPTER FOUR

upon which the political realm should be modeled. Indeed, the Zhouli seems to use these as nothing more than useful organizing principles: it does not claim that Heaven and Earth function as a totality or that the political world should be modeled on such a larger cosmic pattern. In saying this, I am not arguing that Lewis’s argument is incorrect: on the contrary, in Chinese history the Zhouli would often be read precisely along these cosmological lines. My point is that the text makes no such claims itself: unlike the Lüshi chunqiu, the Zhouli does not lay out its argument in cosmological terms. It makes no attempt to claim that the cosmos is an ordered whole and that a properly structured political order should be based upon these cosmic principles. Instead, the entire argument is based upon the ruler establishing an order and then organizing the world accordingly. Instead of seeking an order in cosmology, the Zhouli simply does so administratively, by giving everything that exists a place in the administrative hierarchy. Indeed, I would argue that part of the power—the counterintuitive power—of the Zhouli is that it makes no claims about legitimating state power, about the dispositions of humans, about the normative relationships of humans to the divine and cosmic order. In other words, it takes no position in the debates raging at the time concerning the nature of the cosmos, the relationship of the cosmos to the state, the nature of divine powers and the proper way for a ruler to relate to them, and so on. But, importantly, it also does not rule out any position on these questions: one can give the text a cosmological reading, a reading in terms of divine powers, and so forth.

The Sagely Creation of Order Given the lack of interest in cosmology in the Zhouli, we might search for possible analogues for the Zhouli instead in texts that would later be classified as “Legalist” or “Huang-Lao”—texts like the Writings of Lord Shang (Shangjun shu 商君書) and Han Feizi 韓非子. As with the Zhouli, the goal of these texts is to call on the ruler to create an institutional order in which the duties and functions of each office are clearly delineated. And, also like the Zhouli, these texts make few appeals to legitimation—apart from an implicit claim that the system

WANG MANG AND THE ZHOULI

141

invoked will work—and certainly no claims to a cosmological foundation for governance. This attempt to create order through the development of a proper institutional structure is seen powerfully in the portions of the Han Feizi that build upon the text of the Laozi 老子. Appropriating the language of the Laozi, the Han Feizi argues that the ruler should set up an organizational system and then practice noninterference, allowing the system to operate on its own. In other words, by setting up a proper institutional order, the sage is able to create a world in which humans spontaneously act as the sage wants them to act, while the sage practices nonaction and noninterference: “[The sage] sets up a correct order and resides within it, causing everything to become settled by itself” (正與處之，使皆自定之。).28 The ruler, therefore, is “empty, staying behind with stillness, never exerting himself” (虛以靜 後，未嘗用己。).29 The Zhouli certainly has no such Laozian language, but there are some significant similarities here. The Zhouli too is concerned with the ruler’s construction of a system in which the duties of each office are defined clearly, after which there would be no need for the ruler to interfere with the activities of the state. Moreover, since there are no cosmological norms on which institutional order should rest, it is entirely up to the ruler to set up the order according to the requirements of the times. And, since times change, what is required for one time will differ from what is required for another. As the Shangjun shu argues: “if a sage can thereby strengthen the state, he does not model himself on antiquity, and if he can thereby benefit the people, he does not accord with rites” (是以聖人苟可以強 國，不法其故；苟可以利民，不循於禮。). 30 The sage must be completely free from following past precedent: “Therefore, the knowledgeable create laws, and the stupid are regulated by them. The worthy alter the rites, and the unworthy are restrained by them” (故知 者作法，而愚者制焉。賢者更禮，而不肖者拘焉。).31 The only criterion for judging a sage, therefore, is the degree to which his creation of a new order accurately reflects what is necessary at the time:

28

Han Feizi ICS 8/10/30–31. Han Feizi ICS 8/11/3. 30 Shangjun shu (“Gengfa”) ICS 1/1/12–13. 31 Shangjun shu (“Gengfa”) ICS 1/1/22–23. 29

142

CHAPTER FOUR In the time of Shennong, the males plowed and the people were fed; the women weaved and the people were clothed. Punishments and administration were not used, but everything was put in order. Armored soldiers were not raised, but he reigned as king. After Shennong died, people used strength to overcome the weak and used the many to oppress the few. Therefore, Huangdi created (zuowei) the propriety of ruler and minister and of superior and inferior, the rites of father and son and of elder and younger brother, and the union between husband and wife and between wife and mate. In the interior he employed knives and saws, and in the exterior he used armored soldiers. This is because the times had changed. Looking at it from this perspective, it is not that Shennong is above Huangdi; the reason that his name is honored is that he fit the times. 神農之世，公耕而食，婦織而衣，刑政不用而治，甲兵不起而王 。神農既沒，以強勝弱，以眾暴寡，故黃帝作為君臣上下之義， 父子兄弟之禮，夫婦妃匹之合；內行刀鋸，外用甲兵。故時變也 。由此觀之，神農非高於黃帝也，然其名尊者，以適於時也。32

The Han Feizi makes much the same argument: In the earliest times, when the people were few and the birds and beasts numerous, the people could not overcome the birds, beasts, insects, and snakes. Then there appeared a sage who created the building up of wood to make nests so as to hide the masses from harm. The people were pleased with him and made him king of all under Heaven, calling him the “One Having Nests.” The people ate fruits, berries, mussels, and clams; they were so rank, rancid, bad, and foul-smelling that they hurt their stomachs, and many of the people became sick. Then there appeared a sage who created (zuo) the boring of wood to get fire so as to transform the rank and rancid food. The people were pleased with him and made him king of all under Heaven, calling him the “Fire Man.” In the time of middle antiquity, all under Heaven was greatly flooded, and Gun and Yu opened channels [for the water]. In the most recent period of antiquity, Jie and Zhou were oppressive and chaotic, and Tang and Wu campaigned against them. Now, to have the building up of wood and the boring of wood in the time of the Xia would certainly have made Gun and Yu laugh, and to have the opening of channels in the time of the Yin and Zhou would certainly have made Tang and Wu laugh. As such, to exalt the way of Yao, Shun, Tang, Wu, and Yu in the present age would certainly make the new sages laugh. This is why sages do not try to cultivate the ancient ways and do not model themselves on constancy. 上古之世，人民少而禽獸眾，人民不勝禽獸蟲蛇。有聖人作，搆 木為巢以避群害，而民悅之，使王天下，號曰有巢氏。民食果蓏 蜯蛤，腥臊惡臭而傷害腹胃，民多疾病。有聖人作，鑽燧取火以 32

Shangjun shu (“Huace”) ICS 18/23/1–4.

WANG MANG AND THE ZHOULI

143

化腥臊，而民說之，使王天下，號之曰燧人民。中古之世，天下 大水，而鯀、禹決瀆。近古之世，桀、紂暴亂，而湯、武征伐。 今有搆木鑽燧於夏后氏之世者，必為鯀、禹笑矣；有決瀆於殷、 周之世者，必為湯、武笑矣。然則今有美堯、舜、湯、武、禹之 道於當今之世者，必為新聖笑矣。是以聖人不期脩古，不法常。33

In short, both the Shangjun shu and the Han Feizi celebrate their lack of concern with precedent and give free rein to the sage to create anew as necessary. But here, of course, is a major tension. On the one hand, these texts are committed to the claim that circumstances change, and that the ruler must therefore be free to create a completely new order, unrestricted by past practice or precedent—hence the celebration in these texts of radical innovation on the part of the ruler. But the texts are also committed to the claim that, once this order has been created, the ruler must stop being active and instead practice noninterference. The problem, of course, is that, since times change, the ruler must always be prepared to become active once again and to create anew yet again. Ironically enough, this unresolvable tension may in part explain the appeal of these texts—precisely because it allows the texts to be appropriated in different ways by different figures. Ministers would tend to support such texts’ assertions of the need for clear procedures and regulations—and thus for a noninterfering ruler. But in periods of radical transformation rulers would tend to appeal to such texts because of their strong affirmation of the need for a highly active ruler to innovate. In others words, as problematic as this tension was in practice, the tension was also part of the appeal of the texts to political actors with very different concerns. A crucial difference between the Zhouli and texts like the Shangjun shu and Han Feizi is that, while the Zhouli similarly sees the ruler’s establishment of the state as the condition for the possibility of order, the ruler’s activity is presented not as an act of creation but rather as one of organization. The ruler is an organizer, not a creator. He does not discard precedent and create a new system; he takes what exists and organizes it. Moreover, the way of organizing is always the same: the ruler is the pivot and organizes everything around himself. Thus, there is no hint in the Zhouli that the system described would need to be changed. One is simply organizing what exists, not creating anew.

33

Han Feizi (“Wudu”) ICS 49/145/13–18.

144

CHAPTER FOUR

To return to the example of the ritual specialist: according to texts like the Shangjun shu and Han Feizi a given ritual specialist would be preserved in a new order created by a sage only if that specialist fit into the new order as envisaged by the sage-creator. In other words, the sage would discard anything that needed to be discarded. In contrast, the move of the Zhouli is rather to say that the ruler establishes order by organizing what exists. To take the example of the ritual specialist, the ruler need solicit no opinion on whether that ritual specialist is or is not efficacious, nor is there a concern with whether the ritual specialist does or does not fit into a new order as envisaged by a creator-sage. The concern is rather simply that the ritual specialist is there and thus must be placed within a proper organization. The Zhouli is free from those passages in texts like the Shangjun shu and Han Feizi that celebrate the radical innovations of the creator-sage. This is also why the Zhouli, unlike the Shangjun shu and Han Feizi, works out the full institutional system in excruciating detail. Since the Shangjun shu and Han Feizi and texts like them advocate the creation of systems that are specific to and respond to current situations, it would not make sense for them to present the details of any one system. The Zhouli, however, claims for itself a timeless vision of administration.34 It makes no reference to history, to changing circumstances, to responding to new situations. The implication appears to be that the Zhouli’s prescriptions will work at any time and any place. In short, the Zhouli, like the Lüshi chunqiu, Shangjun shu, and Han Feizi, may well have been written in the context of the emergence of centralized forms of statecraft in the late Warring States period. But, unlike the sacrifice chapters of the Liji, the Zhouli clearly supports a form of overt centralized rule. Whereas the “Li yun,” as we saw, calls upon the ruler to legitimate his rule covertly, by training the dispositions of the populace so that they come to think of the state as a family, the Zhouli shows no interest in such covert operations: the rule of the king in the Zhouli is overt, not hidden. Unlike the Lüshi chunqiu, the Zhouli makes no attempt to legitimate institutions through claims concerning cosmological norms. And unlike the Shangjun shu and Han Feizi, the Zhouli makes no assertions about the duty of sages to create anew and discard precedent when necessary. On the contrary, it claims to provide a blueprint for a system that is applicable for all times.

34

I am indebted to Willard Peterson for emphasizing this point to me.

WANG MANG AND THE ZHOULI

145

What is striking about the Zhouli is that it is, in a sense, outside history and outside claims of legitimation altogether. The power of the text is to say that, whatever exists, here is a blueprint for how the ruler establishes the center and creates a hierarchy in which everything is given a place. One can defend the resulting order through cosmological or sacrificial claims, but the text itself makes no such argument, nor is the text at all concerned with celebrating a creator-sage. The power of the text lies in its absolute commitment to timeless modes of organization rather than legitimation through cosmology, the molding of the people’s dispositions, or the calls for sages to create anew: things exist, and here is how one puts them in order. But, if this was the argument, then it fell on deaf ears. During the imperial period of the ensuing two centuries, the Zhouli was almost completely ignored. The Formation and Consolidation of the Empire The creation of the first empire by the Qin 秦 in 221 BCE inaugurated a distinctive period in Chinese history. Ideas like those in texts such as the Shangjun shu and Han Feizi came fully to prominence. And, intriguingly, the period revealed many of the same tensions concerning active rulership that pervade those texts. One of the dominant goals of the Qin and early Han empires was to expand the bureaucratic system of the state of Qin to control an enormous amount of territory. The intent was to do so through an institutional order based upon a bureaucracy with clear procedures, rules for punishment and reward, and defined duties for all involved. During this period, claims of not only breaking from but in fact superseding the past became increasingly commonplace. What is striking about these claims is that, far from simply asserting, like the Shangjun shu and Han Feizi, the legitimacy of not following past precedent, they celebrated superiority over the past. To give a few rhetorically charged examples, let me first quote some lines from one of the inimitable inscriptions of the First Emperor: It is the twenty-eighth year. The First Emperor has created a new beginning. He has put in order the laws, standards, and principles for the myriad things…. 維二十八年，皇帝作始。端平法度，萬物之紀…

146

CHAPTER FOUR All under Heaven is unified in heart and yielding in will. Implements have a single measure, and graphs are written in the same way…. 普天之下，搏心揖志。器械一量，同書文字… He has rectified and given order to the different customs… His accomplishments surpass those of the five thearchs…. 匡飭異俗…功蓋五帝。35

Such bravado at surpassing the accomplishments of the past would continue over the first century of the ensuing Han 漢 dynasty. As I have argued elsewhere, in the early Han one finds many authors claiming that their works supersede all previous texts—the postface to the Huainanzi 淮南子 being an obvious example.36 A related theme of the period is celebration of the importance of great sages being given the freedom to innovate when necessary, rejecting past precedent. The Huainanzi, again, provides an excellent example. Take the following passage from chapter 13: Great men create and disciples transmit. If you understand from whence standards and order arise, then you can respond to the times and change. If you do not understand the origin of standards and order, you end up in disorder even if you accord with antiquity. The standards and edicts of the current age should change with the times; the rites and propriety should be altered according to changing customs. Scholars accord with those who came before, inherit their practices, rely on their records, and hold fast to their teachings, thinking that there can be no order if it is not thus. This is like placing a square peg into a round hole: they hope to obtain a proper fit and a fixed point, but it is very difficult. 大人作而弟子循。知法治所由生，則應時而變；不知法治之源， 雖循古，終亂。今世之法（藉）〔籍〕與時變，禮義與俗易，為 學者循先襲業，據籍守舊（教），以為非此不治，是猶持方柄而 周員鑿也，欲得宜適致固焉，則難矣。37

The height of this drive toward centralized power, celebration of sagehood, and rejection of using past precedent to legitimize state power occurred during the reign of Han Wudi 漢武帝. As Wudi consolidated the imperial system, he also instituted a sacrificial system that came to symbolize his extreme centralization of power. Building on that of the First Emperor of Qin, the system involved the ruler tak35

Shiji (“Qin Shihuang benji”) 6.245. Puett 2007. For excellent discussions of the Huainanzi postface, see Queen 2001; Murray 2004. 37 Huainanzi (“Fanlun”) ICS 13/122/20–13/120/23. 36

WANG MANG AND THE ZHOULI

147

ing direct control over all land, traveling throughout the realm, personally performing the sacrifices of each local area, and gradually being divinized through the process. The process would end with the ascension of the emperor to the heavens. There was, needless to say, no precedent before the Qin imperium for such a system.38 Critiques of this celebration of innovation and imperial centralization were, of course, frequent. Early in Wudi’s reign, Dong Zhongshu 董仲舒 (ca. 179–ca. 104 BCE) explicitly criticized the Qin for breaking with the past: When it came to the last generations of the Zhou, they definitively brought about the destruction of the Way, and thereby lost all under Heaven. Qin succeeded them. Not only were they unable to change, but they made it worse. They strongly banned the study of cultural patterns and prevented the possession of books. They discarded rituals and what is appropriate, and hated to hear of them. In their hearts they desired to completely extinguish the way of the former kings and to govern only according to their own recklessness and carelessness. 至周之末世，大為亡道，以失天下。秦繼其後，獨不能改，又益 甚之，重禁文學，不得挾書，棄捐禮誼而惡聞之，其心欲盡滅先 王之道，而顓為自恣苟簡之治。39

And for failing to follow the cultural patterns handed down from the earlier kings: When it came to the Qin, things were not like this. They taught the laws of Shen [Buhai] and Shang [Yang]40 and put in practice the theories of Han Fei. They detested the way of Di and the kings, taking greed and cruelty as customary, and did not have cultural patterns or potency to teach and instruct those below. 至秦則不然。師申商之法，行韓非之說，憎帝王之道，以貪狼為 俗，非有文德以教訓於下也。41

Indeed, the attempt by the Han to continue the Qin institutions was comparable to, quoting the famous statement by Confucius, carving rotten wood: Confucius said: “Rotten wood cannot be carved; walls of dung cannot be worked with a trowel.” Now, Han has succeeded Qin. It is like rotten wood or a wall of dung. Although you desire to improve it and put it in order, how is this possible? Laws are promulgated, but crime grows; 38

Puett 2002. Hanshu 26.2504. 40 Shang Yang is the putative author of the Shangjun shu. 41 Hanshu 26.2510. 39

148

CHAPTER FOUR orders are sent down, but deceit arises. This is like using hot water to stop boiling water or carrying kindling to put out a fire. 今漢繼秦之後，如朽木糞牆矣，雖欲善治之，亡可柰何。法出而 姦生，令下而詐起，如以湯止沸，抱薪救火。42

Dong Zhongshu’s solution, of course, was to claim that the new mandate after the fall of the Zhou had been obtained by Confucius, not the Qin, so that the Han should now return to the teachings of Confucius.43 But such arguments were completely out of favor at the court.44 The Return to the Zhou Over the subsequent decades, however, a significant financial crisis developed as a consequence of the extreme centralization of state power under the Han, and several voices emerged calling for a scaling back of the empire. What we see in this period—from about midway through the first century BCE onward—is a change in the court culture, where such calls to reject the legacy of the Qin and to return to something that came before shifted from being a clear minority voice to becoming a major presence in the court debates. Of particular interest to our concerns, however, is that, in the realms of ritual and governance, one of the key issues was not simply to turn to Confucius but rather to return to the practices of the Zhou. The rise of interest in the texts purporting to tell of the ritual and administrative structure of the Zhou occurred in this context. In the thirties BCE, several figures, such as Kuang Heng 匡衡 (chancellor 36–30 BCE) and Zhang Tan 張譚 (imperial counselor: 33–30 BCE), began critiquing the sacrificial system introduced by Wudi, claiming that it differed from the regulations of antiquity.45 They called for a return to the ritual and administrative practices of the 42

Hanshu 26.2504. Hanshu 26.2515. 44 As Sarah Queen (1996) has argued, the view that Dong Zhongshu was a major figure at court is clearly the result of a later reading from the Eastern Han. It is telling in this regard that Sima Qian, writing during the reign of Han Wudi, clearly did not see Dong Zhongshu as a significant figure at court. Sima Qian’s biography of Dong Zhongshu (Shiji 121.3127–3129) is notable for its brevity, mentioning little more than that Dong Zhongshu was a scholar of the Springs and Autumns Annals and that he practiced rain magic. See Puett 2001, 166–168, 262–263n93. 45 Hanshu (“Jiaosi zhi”) 25B.1254. 43

WANG MANG AND THE ZHOULI

149

Zhou. Tellingly, several of the memorials written from this perspective referred not just to the Classic of Documents (Shangshu 尚書) but also to the Liji. Several, moreover, quoted from the Liji, and the quotations given all correspond with our extant text. As Timothy Baker has argued, it seems reasonable to conclude that something like the extant text of the Liji was in existence by the late Western Han.46 In 31 BCE, the court sided with Kuang Heng. The ritual system of Wudi was abolished, and a new system, based upon texts like the Shangshu and Liji, was instituted.47 The new ritual system involved decentralization, an attempt to govern through ritual rather than through imperial institutions, and an explicit claim that it represented a return to the ritual system practiced during the Zhou. This alteration in the ritual system was part of a larger shift that occurred during these last few decades of the Western Han. If the dominant court tradition of the earlier Western Han had been characterized by dramatic claims of superseding the past, the court culture at the end of the Western Han shifted toward calling for restraint, for a return to earlier traditions, and particularly for a return to the Zhou. It is only after this shift in the court culture that significant references to the Zhouli begin appearing in our extant writings. Prior to this period, the Zhouli was referred to only rarely.48 There are, for example, at most two references to the Zhouli in the entirety of the Historical Records (Shiji 史記). (References to the “Zhouguan” in the Shiji refer to the “Zhouguan” chapter of the Shangshu.) Both references appear in the “Monograph on the Feng and Shan Sacrifices” (“Fengshan shu” 封禪書) chapter.49 Although each of these references could refer to the Shangshu “Zhouguan,” Timothy Baker has argued that both probably refer instead to the Zhouli.50 In the first of these, Sima Qian 司馬遷 (ca. 145–ca. 86 BCE) quotes from the Zhouguan concerning sacrifices to Heaven and Earth. However, the quotation does not appear in our extant text of the Zhouli. Later in the chapter, we are told that classicists searched the Zhouguan, as well as the Shangshu and 46

Baker 2006, 164–165. For the late Western Han ritual reform, see Loewe 1974; Kern 2001; Wang Baoxuan 1994; Bujard 2000; Baker 2006; Puett 2002, 307–315. 48 Timothy Baker (2006, 276–293) has provided an invaluable tabulation of all references to the Zhouli, as well as the Liji and the Ceremonial Rituals (Yili 儀禮), in the early histories. Here and below, my summaries are based upon Baker’s findings. 49 Shiji (“Fengshan shu”) 1357, 1397. 50 Baker 2006, 280nn1 and 2. 47

150

CHAPTER FOUR

the “Wangzhi” (later made into a chapter of the Liji), to find information for Han Wudi on the feng and shan sacrifices. Even if these are in fact references to the Zhouli, it is clear that the text was not of major significance at the time. The references show only that Sima Qian associated the text with “classicists” (ru 儒) who were trying to find ancient precedents for rituals, and who were clearly being ignored by the emperor. In short, if the Zhouli is the text being referred to here, these references seem to represent nothing more than how completely marginalized the users of the text were at the court of Han Wudi. As Baker has shown, the Zhouli does not appear in the Hanshu until one reaches the very end of the Western Han. At this point, however, there are several references, including quotations that match our extant text.51 Baker concludes: The large number of references to this text under the term Zhouguan, together with the quotations corresponding to the transmitted version and lack of non-corresponding quotations, clearly confirm that a text similar to the current version was in active circulation by the end of the Western Han. That these references to the Zhouli almost all occur very late in the dynasty, in the Wang Mang period or the two decades preceding that, and [are] essentially all by or related to Liu Xin or Wang Mang clearly show that its popularity lay in that political camp.52

When he was a minister, Wang Mang, while appointing additional specialists of the Classics, also called to court those who had copies of other works, including the lost chapters of the Liji, the ancient text of the Shangshu, and the Mao Odes (Mao shi 毛詩). The Zhouguan was one of the works on the list.53 The reference makes it clear that the Zhouli was not a widely accepted text at the time.54 So, although the Zhouli may well date from the Warring States period, it certainly does not seem to have been considered a major work, and there is a clear sense that Wang Mang was rescuing an obscure text. Following his usurpation, Wang Mang used the Zhouli as a basis for the ritual system,55 for taxation,56 and for organizing state offices.57 Although we 51

Baker 2006, 280–284; Peng Lin 1991. Baker 2006, 163. 53 Hanshu (“Wang Mang zhuan”) 99.4069. 54 Passages such as these, of course, were the ones utilized by Kang Youwei to argue that the Zhouli was fabricated by Liu Xin. For a discussion of Liu Xiang and Liu Xin, see Xu Xingwu 2005. 55 Hanshu (“Jiaosi zhi”) 25.1265. 56 Hanshu (“Shi huo zhi”) 24.1180. 52

WANG MANG AND THE ZHOULI

151

do not have any explicit discussion of the text during this period to give us specifics about Wang Mang’s appropriation of the text, a few clues can be found concerning the culture of the court.58 Wang Mang, very much in keeping with his times, positioned himself as a supporter of classicism rather than the imperial system of the Qin and the legacy of the latter in the early Han. For example, when restoring the “well-field” (jing 井) system of taxation, Wang Mang claimed that the system was practiced by Yao, Shun, and the Three Dynasties and that the Qin destroyed the institutions of the sages when they discarded the system.59 But Wang also championed other texts— like the Zhouli—in addition to those that had been endorsed by figures like Kuang Heng in the generation before. His decision to turn to a text that was clearly not regarded as a significant work—even among classicist scholars—was therefore not an attempt to co-opt the court culture of the day by favoring a text already widely supported. So what was his purpose? Among Wang Mang’s goals were two that were very difficult to accomplish in the period in which he was active. Wang clearly wanted to assert the need for a strong ruler to create a system of government very different from the current one. But, in the context of the time, appeals to a sage-ruler creating a radically new state were very much out of favor. Moreover, it is clear that one of Wang’s goals in creating a new system of government was to establish a powerful central state. But simply returning to the imperial system advocated by Wudi was clearly not an option because that was precisely the system that had been rejected so successfully at court in the previous generation. Both of these goals, of course, would have found strong sanction in texts like the Shangjun shu and Han Feizi. But by the end of the Western Han, texts such as these that asserted the need for a sage-ruler to break from the past and create anew, and specifically to create an order based upon defined roles and functions rather than moral governance, were, as we saw above in the memorials of Dong Zhongshu, very much associated with the Qin and early Han empires. These empires, of course, were being strongly criticized by the end of the Western Han for having broken with antiquity: by far the dominant tendency of the court at the end of the Western Han was to reject the 57 58

244). 59

Hanshu (“Wang Mang zhuan”) 99.4136. See the excellent discussions by Loewe (1974) and Jin Chunfeng (1993, 238– Hanshu (“Wang Mang zhuan”) 99.4110.

152

CHAPTER FOUR

Qin/early Han mode of statecraft and to call for a return to the Zhou. In short, appeals to texts like the Shangjun shu and Han Feizi would have been very unwise in such a context. Yet calls to follow the texts that figures like Kuang Heng in the previous generation had been emphasizing would hardly have supported Wang Mang’s goals. In this context, the Zhouli may have provided precisely the sort of argument Wang Mang needed. As we have seen, the Zhouli possessed the same tension seen in texts like the Shangjun shu and Han Feizi— calling on the one hand for a ruler to establish the state and on the other for a state that thereafter would function without interference from a ruler. But since the Zhouli emphasizes the organizing ruler rather than the creating sage, there is no celebration of having rejected precedent, and certainly no celebration of having created something better than existed in the past. Unlike the claims of radical sagely creation that characterized the reign of the First Emperor, Wang Mang could profess to be following the organizing principles that reigned in the Zhou. In terms of the resulting order that would be created by the ruler, the text also had a potential appeal. In the debates of the Han, the Zhou was associated with a decentralized form of statecraft, in which land was enfeoffed to powerful aristocratic families in perpetuity. However, the Zhouli does not mention how to govern outlying areas. Thus, unlike most texts associated with the Zhou, there is no call for or even discussion of a decentralized feudal arrangement as existed in the Zhou. There is also no discussion of procedures for promotions, rewards, and punishments, or forms of legal ordering of any kind. In other words, the bureaucratic vision that underlay texts like the Shangjun shu and Han Feizi, and that became associated with the Qin and Han empires, is absent from the Zhouli. Nevertheless, nothing in the text argues against a bureaucracy either: the text is simply silent on how the rest of the realm outside the court is to be governed. Thus, while supporting an initially active ruler to create the state, the Zhouli was associated not with the Qin-Han empires but rather with the Zhou. And yet, unlike the Liji, which was so emphasized in the court debates of the thirties BCE, the Zhouli was not based upon a call for legitimation through moral governance or by training the populace not to think of the state as a state—a position associated with at least a claim of decentralization. The ruler is overtly embraced as the figure who establishes order, and the order he establishes is de-

WANG MANG AND THE ZHOULI

153

fined in terms of functions and roles rather than a ritualized familial order. And, conveniently, the text leaves open the question of how to govern the outlying areas of the realm. In short, the Zhouli had the key elements Wang Mang was looking for. Conclusion If the courts of the Qin and early Han empires were distinctive in celebrating innovation, the period beginning in the thirties was distinctive for the opposite reason: a shift occurred that emphasized returning to the past, favoring restraint, and opposing the grandiosity of the earlier imperial courts. But, even within these calls for a return to the Zhou, we have seen two very distinctive approaches, one associated with particular chapters of the Liji and the other associated with the Zhouli. All three of these positions, I might add, continued to exert appeal in later Chinese history as well. The grand imperial claims associated with the Qin and early Han imperial courts would certainly recur. Whereas rulers concerned with building up support during periods of relatively decentralized rule—or in masking a drive toward more centralized rule—would seek support in the sacrifice chapters of the Liji. In contrast, the Zhouli was appealed to by those who sought a more activist form of governance and thus opposed an emphasis on the cultivation of individuals or on the use of rituals to influence the attitudes of the populace, but who for historical reasons found an appeal to antiquity more powerful than a claim to cosmological principles or the potentially divine powers of the ruler.60 With both Wang Mang and Wang Anshi, this was precisely the concern. In short, I would like to argue that part of the power of the Zhouli is precisely that it offers a means of organizing the world simply through the ordering acts of the ruler. Anything that exists can simply be ordered and given a place by the ruler, who thereby becomes a pivot of the realm. In other words, and somewhat ironically, one of the reasons for the Zhouli’s attractiveness was that it offered an approach to political organization based simply on organizing and administering from the center rather than on legitimation through appeals to human nature, creator-sages, or the cosmic order. One institutes and organizes; one does not legitimate or create. I will conclude by simply quoting 60

For a study of the uses of the Zhouli, see Jin Chunfeng 1993.

154

CHAPTER FOUR

the line that recurs throughout the Zhouli: “It is the king who establishes the state.”

CHAPTER FIVE

ZHENG XUAN’S COMMENTARY ON THE ZHOULI Andrew H. Plaks, Princeton University This chapter reports on the findings that emerged in the course of a comprehensive reading of the Zheng Xuan 鄭玄 (127–200) commentary attached to the standard Thirteen Classics, with Commentary and Subcommentary (Shisanjing zhushu 十三經注疏) recension of the Rituals of Zhou (Zhouli 周禮).1 This study was undertaken at the urging of the editors with the idea of commissioning a more focused examination of the largely unquestioned position of Zheng Xuan as the core figure in the long tradition of Zhouli scholarship and its associated political discourse from Han to Qing—an assumption that has gone virtually unchallenged in both traditional and modern scholarship and is taken as a given in a number of the papers from the Zhouli symposia collected here. On the strength of my extensive use of the Zheng Xuan/Kong Yingda 孔穎達 (574–648) zhushu commentary on two other seminal texts of the Three Ritual Classics (sanli 三禮) corpus—the Highest Order of Cultivation (Daxue 大學) and On the Practice of the Mean (Zhongyong 中庸)2—and in the spirit of my lifelong obsession with the ocean of exegetical writings in the scriptural and literary traditions of China and other ancient civilizations, I took on the task of plowing through the entire body of the commonly available Zheng comments on the text, in the hope of bringing to light new insights regarding the commentarial method and intellectual substance of Zheng Xuan’s contribution to the Zhouli textual heritage. My aim was not to construct a comprehensive evaluation of the full range of Zheng’s canonical commentaries within the broader context of Confucian learning, or to uncover any new revelations with respect to the arcane lore of “canonical studies” (jingxue 經學) or the labyrinthine 1 The page numbers cited below refer to the two-register Taiwan reprint of the Shisanjing zhushu compendium (Taibei: Dahua shuju, 1977), vol. 3. 2 See the interpretive apparatus attached to my translation of these two texts for Penguin Classics under the title The Highest Order of Cultivation and On the Practice of the Mean (Plaks 2003), esp. appendix III, 117–127.

156

CHAPTER FIVE

webs of Han (206 BCE–220 CE) “modern text” (jinwen 今文) and “ancient text” (guwen 古文) lineages, but simply to clarify what Zheng Xuan did and did not say in his seminal writings on this particular text.3 The results of this study, to be reviewed below, have proven to be quite a bit thinner than anticipated, both quantitatively and qualitatively. In terms of their sheer volume—as all attentive readers of the Zhouli have no doubt noticed—the Zheng Xuan zhu component of the zhushu recension is not particularly extensive. The vast majority of the Zheng comments printed interlinearly following each bit of text are limited to a scant line or two. Even at those points where Zheng enters into more substantive treatment of a given topic, his remarks rarely run beyond five or ten lines; and only on infrequent occasions does he plunge into lengthy discussions stretching over a page or two to elaborate upon issues that seem to spark his interest. More disappointing still, to all but the most pedantic philologists among us, must be the fact that the overwhelming bulk of the zhu comments attributed to Zheng Xuan are restricted to lexical or phonological glosses of a simple explicative nature, rarely entering into the sort of ideological or institutional discourse of primary concern to most of the participants in the Zhouli symposia. It remained, of course, for later luminaries of Chinese classical learning—from Jia Gongyan 賈公彥 (fl. 627–656) within the context of the early Tang dynasty (618–907) Corrected Interpretations of the Five Classics (Wujing zhengyi 五經正義) enterprise4 to the major Song (960–1279), Ming (1368–1644), and Qing (1644–1912) Confucian proponents of one or another line of argumentation—to draw out and respond critically to some of the more profound issues of Confucian political and philosophical thinking left unexplored in the basic text and its earliest commentarial strata. These later layers of embroidery on the textual ground left by Zheng Xuan form the subject of a number of the prior and present papers of this project, as they pick up and weave these threads into compelling studies of Zhouli-based theory and rhetoric. 3 This study project, carried out during the spring and summer of 2006, involved exhaustively working through all of the Zheng comments I was able to assemble, in both the most common composite printings and in certain separate editions, for example, Zhouli Zheng zhu 周禮鄭注 (comp. Ge Zi 葛鼒; fl. 1633) and Zhouli Zhengshi zhu 周禮鄭氏注 (comp. Huang Pilie 黃丕烈). 4 I choose to translate the term zhengyi in this title in the sense of the Tang/Song term jingyi 經義 rather than the simpler “corrected meanings.”

ZHENG XUAN’S COMMENTARY

157

In what follows, I will present an initial review of those infrequent points encountered in my line-by-line perusal of the Zheng Xuan commentary that may be of relevance to contemporary Zhouli scholarship from a variety of interpretive perspectives. These include certain moments in which Zheng, and other writers before and after him, respond critically to one another’s interpretations, a few definitions and elucidations of specific terms and expressions that I find significant, and those scattered passages in which the master exegete indulges, to varying degrees, in more detailed demonstrations of the unparalleled command of Han ritual lore commonly attributed to him. In a few of these contexts, Zheng may actually have some original insights to express regarding both his general understanding of the concept of ritual and the specific intricacies of certain topics, notably those of sacrificial procedure, court protocol, and calendrics. What is missing, once again, is a direct presentation of Zheng’s thinking on issues of political theory and moral philosophy of the sort that do come into view at certain points in his commentaries attached to the Daxue, Zhongyong, and other treatises in the Ritual Records (Liji 禮記) compendium, in the authoritative Zheng Xuan “glosses” (Zheng jian 鄭箋) on the Classic of Poetry (Shijing 詩經), and in other works of Zheng’s vast exegetical corpus. By rights, an accurate appraisal of Zheng’s contribution to setting the textual parameters for later Zhouli polemics would need to emerge from a comprehensive analysis of the full range of his commentarial oeuvre, touching upon such points as his adaptation of prior sources and his selective use of proof texts. Lacking the technical expertise necessary to undertake a precise mapping of Zheng’s Zhouli glosses on the complex grid of Han and post-Han textual scholarship and intellectual history, however, I have attempted to add at least one comparative perspective to my investigation by contrasting his elucidations on the Zhouli to those found in the standard zhu comments, universally attributed to this same exegete, on a text that covers many of the same topics of the Confucian idealization of the Zhou ritual order: that is, the “Royal Institutions” (“Wangzhi” 王 制) chapter of the Liji.5 All of the following observations depend, of course, on how we deal with the question of identifying Zheng Xuan’s own individual contribution among the tangled strands of what is actually a patch5

I believe that the translation “institutions” gives a more accurate reflection of the contents of this text than do the common “regulations” or “stipulations.”

158

CHAPTER FIVE

work of borrowing, citation, philological erudition, and personal opinion. In the Shisanjing zhushu edition from which all of my examples are taken,6 the zhu comments inserted in double columns after each brief textual unit are left unmarked, there being no need, one assumes, to identify them in a compilation presented explicitly in its attached prefaces as a conjoining of the zhu of Zheng Xuan with the shu of Jia Gongyan. It is not always clear, however, how we are to pinpoint Zheng Xuan’s own voice among the choir of forerunners that he regularly cites (more on this below), and we also face the greater ambiguity occasioned by the appearance of many extensive passages inserted into the Jia Gongyan “subcommentary” that are typically labeled “Zheng Xuan said” (Zheng yun 鄭云) but that are often not duplicated in the zhu sections themselves. Before proceeding to address some of these issues, let us first rehearse what we know and what we do not know about the life and work of Zheng Xuan. Assuming that most readers of this volume on the Zhouli are more or less familiar with the basic facts of the secondcentury exegete’s somewhat atypical life, I will limit myself to a few remarks that may have a bearing on the overall evaluation of Zheng’s contribution to the establishment of the Zhouli exegetical corpus. Our primary source for Zheng’s vita is the biographical entry on his career presented, along with those of his contemporaries Zhang Chun 張純 (d. 56 CE) and Cao Bao 曹襃 (d. 102 CE), in the History of the Later Han (Hou Hanshu 後漢書).7 Both in the fairly predictable outlines of his rise to prominence as a scholar and in his unfortunate implication in the political persecutions of the years 166 and 169 known to history as “sequestration of partisans” (danggu 黨錮)—leading to years of forced retirement, which, according to most historians, afforded him the opportunity to devote all his energies to his most important exegetical writings—the portrait of Zheng given here seems to convey the sense of a man who, though caught up in the swirling intellectual and political currents of his time, remained something of an outsider. With respect to his specific connection to Zhouli learning among the ritual specialists of his day, it may or may not be significant that, aside from a laconic mention of his early achievement as a scholar of the Offices 6 Printed here from the Academia Sinica Website at www.sinica.edu.tw/ ~tdbproj/handy1 (Zhongyang yanjiuyuan Hanji wenxian ziliaoku 中央研究院漢籍文 獻資料庫). 7 Hou Hanshu 35.1193–1216. For a chronological outline of Zheng Xuan’s life, see Wang Liqi 1983. A brief précis of Zheng’s life is given in Coblin 1983, 29.

ZHENG XUAN’S COMMENTARY

159

of Zhou (Zhouguan 周官) in the ranks of the students of a certain Zhang Gongzu 張恭祖, the subsequent summation of his major exegetical writings conspicuously omits any mention of a full commentary on the Zhouli, listing only a prose work written in response to an earlier critical essay on the Zhouli by Lin Xiaocun 臨[=林]孝存, also known as Lin Shuo 林碩.8 This omission may perhaps reflect a deliberate political insinuation or a personal prejudice on the part of Fan Ye 范曄 (398–445), the author of the History of the Later Han, but it does not diminish in the slightest the universal recognition of Zheng Xuan’s central place at the core of Zhouli learning or his pivotal position along the chain of scholars including Jia Kui 賈逵 (30–101) and Ma Rong 馬融 (79– 166), as well as Du Zichun 杜子春 (ca. 30 BCE–58 CE) and the father-and-son pair Zheng Xing 鄭興 (fl. 20–30) and Zheng Zhong 鄭眾 (also called Zheng Sinong 鄭 司 農 , d. 83). 9 This recognition of Zheng’s foundational standing, however, does not necessarily guarantee a uniformly positive evaluation of the substance of his scholarly contribution. On the contrary, the balance sheet on Zheng Xuan is strikingly uneven, with opinions running the gamut from unalloyed praise to rather harsh condemnation. It comes as no surprise that Jia Gongyan loudly proclaims the originality of his champion.10 In his prefatory essay entitled “On the 8

The relevant passages read as follows: “He also received instruction in the Officers of Zhou, Record of Ritual, Zuo Commentarial Tradition on the Springs and Autumns Annals, Mr. Han’s Outer Tradition of the Poetry, and Classic of Documents Ancient Text… The entire range of his commentaries, including the Classic of Changes, Classic of Documents, Mao Odes, Ceremonial Rituals, Records of Ritual, Analects, Canon of Filial Piety, Great Commentary to the Documents, Cyclical Periods [in the Classic of Documents], and Calendrical System [Based on] the Image-Text of the Qian Hexagram, plus his writings On the Seven Astral Bodies of the Celestial Order, Meaning of the Di and Xia Sacrifices in the Ritual System of Lu, Essay on the Six Branches of Learning, Analytical Study of the Mao Odes, A Refutation of the ‘Miscellaneous Interpretations of the Five Canons’ by Xu Shen, and A Reply to the ‘Queries on the Rituals of Zhou’ by Lin Xiaocun, number more than one million words in all” (又從東郡張恭祖受周官，禮記,左氏春 秋，韓詩，古文尚書 [Hou Hanshu 35.1207] 凡玄所注周易，尚書，毛詩，儀 禮，禮記，論語，孝經，尚書大傳，[尚書]中候，乾象歷，又著天文七政論， 魯禮禘祫義，六蓺論，毛詩譜，駮許慎五經異義，荅臨孝存周禮難，凡百餘萬 言。[Hou Hanshu 35.1212]). 9 Zheng Xuan is frequently referred to as the “Latter Zheng” (hou Zheng 後鄭) with respect to the earlier pair. 10 Here I borrow a locution from the Judaic commentarial tradition, in which medieval subcommentators are known as the “arms-bearers” of their champions on the field of exegetical combat.

160

CHAPTER FIVE

Rise and Fall of the Zhouli” (“Zhouli feixing” 周 禮 廢 興 ), Jia Gongyan echoes a phrase from the final reprise (lun 論) of Zheng’s Hou Hanshu biography, crediting him with “a comprehensive grasp of the major canonic texts, encompassing all of the exegetical schools” (kuonang dadian wangluo chongjia 括囊大典，網羅眾家).11 In fact, in one of his first expansions on a textual comment by Zheng Xuan, Jia claims that it is the originality of some of Zheng’s exegetical views in his elaborations upon the original meaning of the classic text that warrants the application of the term zhu to his commentary (as opposed to the term zhuan applied to certain later writings).12 Ruan Yuan 阮元 (1764–1849) repeats this attribution of encyclopedic knowledge in his general introduction to the Zhouli volume of the Shisanjing zhushu, and in his “collation notes” (jiaokanji 校勘記) appended to the end of the first chapter, he singles out the unique paleographic skills of Zheng Xuan, along with those of Du Zichun and the two “Former Zhengs.”13 This sort of praise is, of course, fairly standard rhetoric. However, in the eyes of other critics, it is precisely this syncretic breadth that defines the weakest link in Zheng Xuan’s exegetical method. For example, Ruan Yuan goes on in his basically positive assessment to explain the manner in which Zheng’s tendency to use material cited from apocryphal writings (weishu 緯書) drew the critical ire of Ouyang Xiu 歐陽修 (1007–1072).14 The Song Confucian Ye Shi 葉時 11 The full Hou Hanshu passage reads: “He had a comprehensive grasp of the major canonic texts, encompassing all of the exegetical schools, excising verbose or spurious material, and emending the texts to correct omissions. [Zheng] Xuan was succinct in his philological glosses, though he was criticized quite a bit by others well versed in the field for being too verbose” (括囊大典網羅眾家，刪裁繁誣，刊改漏 失。玄質於辭訓，通人頗譏其繁。). 12 See 1.1b (Tianguan: Zhongzai 冢宰): “The zhu glosses attributed to Zheng Kangcheng note his own views at the end of each [section of the] canon, allowing the meaning of the text to be set forth more clearly. That is why they are called ‘glosses.’ By contrast, those of Master Kong [Anguo] and Wang Su, among others, are referred to as ‘commentary’” (康成注者於經之下自注己意，使經義可申。故云注也。孔 君、王肅之等，則言傳。). 13 Juan 1, 1a. 14 See Ruan introduction 42.3b. Ruan goes on to defend Zheng and Jia from Ouyang Xiu’s attack on the grounds that Zhu Xi 朱熹 (1130–1200), who, Ruan claims, was more deeply versed in ritual texts, gave Zheng and Jia his unqualified praise. Pi Xirui’s 皮錫瑞 (1850–1908) seemingly derogatory remark to the effect that Zheng’s commentary marked the end of Han learning (鄭學盛而漢學衰) may be similarly taken as backhanded praise of Zheng’s final synthesis of the supposedly warring jinwen and guwen schools (see Jingxue lishi 經學歷史 148). Be that as it may, Zheng routinely quotes from a very broad spectrum of earlier texts, including the Chunqiu,

ZHENG XUAN’S COMMENTARY

161

(jinshi 進士 1185) voices a more extreme condemnation of Zheng’s broad-ranging use of related texts, listing his characteristic recourse to extensive citation (again echoing the phraseology of the Hou Hanshu), especially his use of various works of the Han “apocryphal” corpus, as his most destructive failing: One loudly claims that Zheng Kangcheng [=Zheng Xuan] had “a comprehensive grasp of the major canonic texts, encompassing all of the exegetical schools.” The point is that he was skilled in synthesizing different sources. So it is not surprising that he embraces marginalized texts, consistently pushing his interpretations in those directions, while relying on his own personal intuitions and bringing in all kinds of irrelevant and misleading proof texts. This is the reason the meaning of the Zhouli has remained obscure and has called forth conflicting opinions on the part of latter-day Confucian scholars. 鄭康成號為囊括六典，網羅眾家。蓋亦知所折衷矣。胡為不抱遺 經推究終始，而乃慿私臆決旁据曲證，此周禮所以不明，而召後 儒紛紛之議也。15

His animus piqued, Ye now goes on to a shockingly blunt conclusion: What a shame! The manner in which Liu Xin devalued the Zhouli was only a failing of his own time, and the moral norms of the Zhouli continued to stand. But the damage to the Zhouli caused by Zheng Kangcheng is a disruption lasting a thousand years, whereby the moral norms of the Classic have virtually ceased to exist. 吁！劉歆之誣周禮，一時之失，而周禮之法尚在。鄭康成之壞周 禮千載之惑, 而周禮之法幾亡。

These positive and negative evaluations of the overall contribution of Zheng Xuan to the tradition of Zhouli learning leave unanswered a number of basic questions regarding the nature of the Zheng commentary that must be left for further elucidation to jingxue colleagues with a more thorough grounding in the complex filiation of Han intellectual lineages. The first of these concerns the precise relation between Zheng Xuan and the earlier Zhouli exegetes from the first generations Zuozhuan, Guoyu, “Yao dian,” Mengzi, Liji (esp. the “Tangong,” “Quli,” and “Wangzhi” chapters), the Erya, and even the Lisao. 15 Combined Origins of the Ritual Canon (Lijing huiyuan 禮經會元), juan 1, 5a– 8b. Despite this common allegation of a tendency on Zheng’s part to overuse apocryphal sources, I have not noticed an exorbitant number of direct citations of weishu in the Zheng commentary on the Zhouli. For one example, see his quotation from the “Paths to Probe the Mysteries of Heaven” (“Qianzuodu” 乾鑿度) on 1.3a.

162

CHAPTER FIVE

of erudites following the incorporation of the text into the li curriculum by the time of Liu Xin 劉歆 (c. 46 BCE–23 CE), and particularly Du Zichun, Zheng Xing and Zheng Zhong, on whom he draws so extensively and openly.16 In keeping with the personalizing tendency of much of traditional jingxue scholarship, much attention has been paid to the question of whether Zheng Xuan did or did not have any kinship tie to Zheng Xing and Zheng Zhong beyond their shared surname. Jia Gongyan seems to adopt the former position in observing Zheng’s pervasive use of material from the Philological Analysis of the Zhouli (Zhouli jiegu 周禮解詁) attributed to the two Former Zhengs.17 While later scholars tend to cast doubt upon this alleged family connection, there is little uncertainty, in view of the sheer amount of space devoted to explicit citations from this source (especially those attributed to Zheng Zhong), that Zheng Xuan regarded the two of them as at least his intellectual forebears (this may be the literal meaning of Jia Gongyan’s use of the word xian 先 in this context). Of course, there is a certain chicken-or-egg circularity to this issue, since virtually the entire extant corpus of the two Former Zhengs’ comments on the Zhouli, as later reconstituted by Ma Guohan 馬國翰 (1794–1857), seems to be derived almost exclusively from citations adduced in the Zheng Xuan commentary itself.18 As we have seen, Jia Gongyan is at pains to insist that Zheng Xuan’s unique contribution lies precisely in his critical use of the earlier commentarial writings as a springboard for launching his own interpretive views. In an interesting aside in a comment at the beginning of the first chapter, Jia explains the precise rhetorical formula by which Zheng signals the points at which his citations leave off and his own individual voice begins.19 Where Jia Gongyan explicitly uses ad16 This does not mean that Zheng did not use the comments of his exegetical forebears with critical selectivity. Quite frequently he cites the views of Du Zichun, Zheng Xing, and Zheng Zhong only to take issue with them; see 1.3a (Tianguan: Zhongzai 冢宰), 33.16a (Xiaguan: Zhifangshi 職方氏), and many other places. 17 See 1.3a–b (Tianguan: Zhongzai 冢宰): “The ‘two Zhengs’ were both the forebears of Zheng Kangcheng; that is why they are referred to by their official rank and not by name. Du Zichun does not belong to his ancestral line; therefore, his name is indicated” (二鄭皆康成之先, 故言官不言名字. 杜子春非已宗, 故指其名也). See also Feixing 12b: “The two Zhengs were leading Confucian scholars of the same ancestral line” (二鄭者同宗之大儒). 18 See Yuhan shanfang jiyi 玉函山房輯佚 644–648. 19 See 1.3b (Tianguan: Zhongzai 冢宰): “The phrase ‘Zheng said’ (Zheng wei 鄭謂) is used for the most part through the entire text. Whenever Zheng Xuan adds a com-

ZHENG XUAN’S COMMENTARY

163

ditional material that he attributes to Zheng Xuan, he regularly indicates this with the formula “Zheng yun” 鄭云 (see above). A similar uncertainty appears when one looks beyond Zheng Xuan’s time to assess his place in the subsequent chain of exegetical filiation, that is, when one attempts to separate the strands of cited material incorporated into many of Jia Gongyan’s shu expansions. As mentioned above, I have observed among those many passages explicitly labeled by Jia as Zheng’s own words a significant number of passages not found, as far as I have been able to trace, within the lines of zhu commentary reproduced by Ruan Yuan. The mode of presentation, and the terms used to cite the “Zheng Xuan” comments, in the Wujing zhengyi recension, I might add, are fairly consistent throughout the entire length of the Zhouli and do not seem to reflect any significantly different treatment in the suspect “Artificer’s Record” (“Kaogong ji” 考工記) chapter that might have a bearing on the vexed issue of this section’s authenticity. This gives rise to the question of what sort of manuscript or other sources Jia Gongyan may have had to work with: did he have at his disposal a fuller version of the Zheng Xuan commentary that is no longer extant, or did he perhaps exercise a measure of poetic license, or creative imagination, in adducing the teachings of his long-dead master? By the same token, there are quite a few spots in the space reserved for Zheng Xuan’s zhu glosses in the Shisanjing zhushu where the editors (or whoever composed the woodblocks) simply skip anachronistically to Lu Deming 陸德明 (556–627) in the textual location ostensibly reserved for the Zheng zhu (as is also observed in the Classic of Poetry volume of the Shisanjing compendium and elsewhere). In one very glaring example, a comment attributed to Zheng Xuan (inside a shu passage prefaced by “Zheng yun”) contains a quotation allegedly cited from Gan Bao 干寶 (fl. 317–322), more than a hundred years after Zheng’s lifetime.20 Yet another context in which we suspect that materials collated in the zhu and shu sections of the Shisanjing zhushu have been allowed to merge arises in connection with the sound glosses that occupy a major portion of the “Zheng Xuan” commentary. Despite the sweepment to those of Zheng Sinong or other scholars, it is indicated by ‘Xuan comments,’ in which case, it may be noted, it is never presented with the phrase ‘Xuan said.’ When a given comment comes after those of other scholars, it is then labeled ‘Xuan said,’ to distinguish it from those of the others” (鄭謂者大略一部之內。鄭玄若在司 農諸家上注者是玄注。可知悉不言玄謂。在諸家下注者，即稱玄謂以別諸家。). 20 See 1.4a (Tianguan: Zhongzai 冢宰).

164

CHAPTER FIVE

ing generalization of W. South Coblin to the effect that no sound glosses of the fanqie type were used prior to the Six Dynasties period, the Zheng commentary printed in the standard recension of the Zhouli does seem to contradict this, since it routinely provides fan spellings for most key terms—unless, of course, we dismiss all of these as later interpolations.21 Setting aside for the time being all of these uncertainties and caveats regarding the precise identification of the exegetical materials associated with Zheng Xuan, let us now take a closer look at some of the contents of what is taken as the body of his extant commentary on the Zhouli. As I have said, the vast majority of the Zheng Xuan zhu comments that have come down to us through the sources that served as the basis of the Shisanjing recension are devoted to straightforward philological explication, often limited to the definition of the basic terms of a given Zhouli section. As examples of this sort of word gloss, I would like to cite a few instances that I find particularly apt, even if they are not necessarily profound or revolutionary: 1.4b (Tianguan: Zhongzai 冢宰) [on the phrase “to serve as the main beam for the people” 以爲民極]: The term “main beam” signifies a state of centered balance, whereby one brings about a situation in which every person in the world attains his correct balance, never deviating from his proper place. 極，中也。令天下之人各得其中，不失其所。 2.7b (Tianguag: Taizai 大宰) [on the expression “the eight unifying principles” 八統]: Unifying principles are the means whereby things are brought together in graded alignment. （釋曰鄭云）統，所以合牽以等物也。 10.22b (Diguan: Dasitu 大司徒) [on the phrase “assigning them to protect…to receive” 使之相保…使之相受]: This is the means whereby one provides incentives to the populace. All of these instances of “assigning one” refer to appointing elders of the community and instructing them to carry out their functions… The phrase “assigning them to protect [the people]” is thus equivalent to saying “give them an official position.”… For this reason, one should write the word “receive” (shou 受) as “impart” (shou 授). Du Zichun says it should be read as “receiving,” in the sense that wherever the people may arrive in their movements from place to place they receive these directives…

21

See Coblin 1983, 5. Coblin seems to waver on this point and later allows that perhaps some of Zheng Xuan’s original sound glosses have been lost.

ZHENG XUAN’S COMMENTARY

165

此所以勸民者也。“使之” 者皆謂立其長而教令。“使之保”猶 任也。 故書受為授。杜子春云當為受，謂民移徙所到則受之。 14.5a (Diguan: Shishi 師氏) [on the word “moral training” 教]: The point of “moral training” is to cause an individual to gain an understanding of the meaning of past events. 教之者，使識舊事也。 39.5a (Dongguan: Kaogongji 考工記) [on the phrase “he who understands is the creator of things” 知者, 創物]: This means the fashioning of a material object de novo, just as one speaks of the advent of a new era. 謂始闓端造器物，若世本作者是也。

In light of the prevailing reputation of Zheng Xuan as the foremost expert of his age in the field of ancient ritual lore, the student of the Zhouli can derive particularly valuable insights from those points at which he draws upon his proverbially encyclopedic learning to explicate specific aspects of the ritual system enshrined in the canonic work. As a rule, Zheng’s primary method consists in fleshing out the idealized picture of archaic ritual practice outlined in the bare bones of the Zhouli text with specific information gleaned from a wide variety of sources, many of which are unknown to us.22 But at a number of points one senses that the details he brings to bear in his elaborations may reflect the realia of the political, social, cultural, and especially religious life of his own latter-day world in late Han times. The range of topics to which Zheng Xuan devotes at least a few lines, and occasionally more detailed discussions, is quite broad; for convenience I will arbitrarily divide these subjects into those concerning the idealized institutional forms of archaic statecraft and those that reflect ritual concepts and practices of a more mundane or private nature. At the same time, we may also distinguish between instances in which the exegete prescriptively outlines the framework of the ritual system and those passages in which he appears to be more interested in showing off his encyclopedic knowledge of Han ritual lore. 22

These include both received canonical texts and extracanonical writings, as well as unidentified bodies of lore. For example, Falkenhausen (2008) notes that Zheng Xuan’s elucidation of the term yuexuan 楽懸 (suspended musical instruments) in the Xiaoxu 小胥 section of the Chunguan 春官 division of the Zhouli (Dahua ed., juan 23, 1715–1716) draws upon information that had long before disappeared from the received ritual corpus and has only recently been confirmed by archaeological finds. An exhaustive examination of the range of Zheng Xuan’s sources must, of course, await a full-length study of his overall exegetical method.

166

CHAPTER FIVE

One area of institutional ritual upon which Zheng Xuan elaborates in more than one passage has to do with the paraphernalia and the protocols of court assemblies. For example, at 37.10a (Qiuguan: Da xingren 大行人) and 38.2a (Qiuguan: Siyi 司儀) he goes into some detail regarding the prescribed forms to be observed by the “lords of the states” (zhuhou 諸侯) in attending seasonal musters and other imperial levees, including such matters as dress, insignia, and even the proper procedures of lining up.23 In one such context, he provides a piece of interesting information, repeated in several other places, about the use of a wood or metal stick to beat a sort of assembly gong (muduo 木鐸) sounded to summon people to attend a court levee or the reading of an imperial decree.24 Another important area of institutionalized ritual that is ascribed to the Zhou court and that elicits a number of more extended comments from Zheng Xuan relates to certain aspects of what we may call the “legal system” (fa 灋 [法]). Needless to say, the use of the word fa in the Zhouli generally signifies normative practices of the Zhou ritual order and not the “law” per se, but on at least one topic, the administration of justice, Zheng’s treatment is of substantive legal interest. This is his explication of some of the concrete details of the application of canonic punishments, as prescribed in the rudimentary penal code visible in certain passages of the early Confucian corpus. We have, for example, the following: [In face-branding,] first they make incisions in the face [of the convict], which they then fill with dye. “Nasal amputation” means literally the severing of the nose, in the manner of some of the border peoples in the east and west in our own day, who sometimes use face-branding and nasal amputation as their common practice… “Sterilization” in the case of men refers to severing the male organ, and for women, sealing the womb, as one does in our time for initiating men and women into official servitude.

23

See also 35.20a (Qiuguan: Chaoshi 朝士), 37.14a (Qiuguan: Da xingren 大行人), 38.1a (Qiuguan: Siyi 司儀). 24 See 3.10b (Tianguan: Xiaozai 小宰): “In ancient times, whenever a new decree was to be issued, a wooden gong would always be struck to alert [the populace]. The term ‘wooden clapper’ refers to its wooden tongue. For civil matters a gong of wood was struck; for military matters it was a gong of metal” (古者將有新令必奮木鐸以 警。木鐸，木舌也。文事奮木鐸，武事奮金鐸。). This should be distinguished from the well-known use of the same term to indicate a kind of wooden clapper traditionally believed to have been sounded to summon villagers for the purpose of gathering folk songs and other expressions of popular sentiments in archaic times.

ZHENG XUAN’S COMMENTARY

167

先刻其面以墨窒之。劓，截其鼻也，今東西夷或以墨劓為俗 者，丈夫則割其勢，女子閉於宮中，若今官男女也。25

宮

Other areas of archaic statecraft outlined in the Zhouli about which Zheng Xuan has a significant amount of information to impart include topics of centralized economic control, with respect to the regulation of markets (15.6a [Diguan: Quanfu 泉府]), surveying of land divisions (10.2b [Diguan: Dasitu 大司徒], 42.1b [Dongguan: Jiangren 匠 人]), variable taxation of lands in accordance with different types of productivity (13.15b [Diguan: Zaishi 載 師 ]), and disaster relief (10.18a [Diguan: Dasitu 大司徒]). As stated above, some of the more interesting among Zheng Xuan’s relatively extended comments deal with ritual modes of behavior that go beyond the sphere of central administrative control. These include such topics as food (e.g., details of the specific contents of certain ritual offerings; 5.24a [Tianguan: Bianren 籩人]), clothing (e.g., court dress, insignia, and decorations, sumptuary regulations; 8.15b [Tianguan: Zhuishi 追師], 10.21a [Diguan: Dasitu 大司徒], 37.14a–b [Qiuguan: Da xingren 大行人]), hygiene (e.g., bathing and purification; 6.7a [Tianguan: Gongren 宮人]), medicine (e.g., use of medicinal and therapeutic treatments; 5.9a [Tianguan: Shouyi 獸醫]), music (e.g., harmony of standard tones; 23.10b [Chunguan: Taishi 大 師]), and many more. Among the various aspects of personal ritual behavior that are treated in greater detail in the Zheng Xuan commentary, special attention is focused on the theme of marriage, with respect to both specific practices and theoretical underpinnings. In a number of these passages, as one would expect, Zheng cleaves fairly closely to the prevailing Han Confucian views on the proper regulatory function of marriage in ensuring both social and cosmic harmony. In one discussion, for example, he places a formulaic dictum on proper marriage in the context of a distinction drawn in the text between what seems to mean here “external rituals” (yangli 陽禮; i.e., those exposed to the light of day) and “internal rituals” (yinli 陰禮; i.e., those kept in the shadows for the sake of modesty): The expression “external rituals” refers to such ritual occasions as archery fests and wine drinking, while “internal rituals” means ritual propriety between men and women. As long as marriages are conducted at the 25

36.1a (Qiuguan: Sixing 司刑). See also 10.26a (Diguan: Dasitu 大司徒), 35.2b (Qiuguan: Xiaositu 小司徒).

168

CHAPTER FIVE proper times, then men will not forsake [their wives] and women will have no cause for resentment. 陽禮謂郷射飲酒之禮也。陰禮謂男女之禮。婚姻以時，則男不曠， 女不怨。26

Curiously, in a longer comment in another place, Zheng first cites an extended quotation from Zheng Zhong that uses similar clichés of cosmic harmony through matrimony based on the same distinction between “external” and “internal” moral foundations, but then he goes on to turn this around and reinterpret the key phrases in the text in the sense of attaining harmonious balance through proper diet.27 Elsewhere, in a discussion of the putative archaic practice of arranging mass marriages at the height of spring,28 Jia Gongyan cites an entirely predictable comment by Zheng: “In midspring, marriage rituals were conducted, in accordance with the intersection of yin and yang, thus 26

See 10.6a (Diguan: Dasitu 大司徒). See 18.26a (Chunguan: Da zongbo 大宗伯): “Zheng Sinong says: ‘“the power of yin” states that the feelings [of attraction] between men and women are spontaneous, arising from their innate dispositions. This means that when the proper time is missed couples will just run off together. If it is before the proper time has been reached, then their lifeblood and breath will be in an unsettled state. On account of this the sages established that during this period [male] members of the population who were thirty years of age were to take wives, and women were to be married at the age of twenty, in order to avoid licentious and dissolute behavior and to ensure that there would be no moral lapses, to keep their natural inclinations suppressed and not allow them to be indecently exposed. That is why this is termed “the power of yin.”’ Xuan says: ‘“the power of yin” is the yin energy. In the human constitution this yin energy is weak. When it is kept pure and undiluted, the result is deficiency. That is why one eats animal products to develop it and make it vitally active, but if this is excessive, it is harmful to one’s natural disposition. Within the framework of this institution the rituals are conducted in order to regulate this. The “power of yang” is the yang energy. In the human constitution this yang energy is ascendant. When it is kept pure and undiluted, the result is overstimulation. That is why one eats vegetable products to develop it and make it quiescent, but if this is excessive, it is harmful to one’s natural disposition. Within the framework of this institution harmonious interaction and music are used to regulate this. As a consequence of this, people’s yin and yang energies are stabilized, their feelings and dispositions are harmonized, and they are thus enabled to reproduce their own kind’” (鄭司農云：陰德謂男女之情，天性 生而自然者，過時則奔，先時則血氣未定。聖人為制其中令民三十而娶女二十 而嫁，以防其淫泆令無失德，情性隱而不露。故謂之陰德 玄謂：陰德、陰氣 在人者陰氣虛, 純之則劣。故食動物作之使動，過則傷性，制中禮以節之。陽德、 陽氣在人者陽氣盈，純之則躁。故食植物作之，使靜，過則傷性，制和樂以節 之。如是然後陰陽平，情性和，而能育其類。). 28 14.15a (Diguan: Meishi 媒氏). The line “During the month of midspring they brought men and women into common assembly” (zhongchun zhi yueling hui nannü 中春之月令會男女) is rather fancifully interpreted by Granet and others on the basis of this and similar classical sources. See Granet 1919, 155ff. 27

ZHENG XUAN’S COMMENTARY

169

conforming to the temporal cycles of the cosmic order” (chongchun, yinyang jiao yi cheng hunli, shun tianshi ye 中春，陰陽交以成婚禮 ，順天時也。). Then Jia follows this up with a blunt counterargument from Wang Su 王肅 (195–256) taking strong issue with Zheng’s understanding of this passage.29 Under the heading of family or personal ritual, we may also mention here Zheng Xuan’s elaboration of certain telling details regarding funeral and burial practices, as well as such associated topics as divination and dreams.30 I find particularly appealing his definition of “dream” as “that which can serve as a basis for divination when one’s spirit returns to a waking state” (mengzhe ren jingshen suoyi wu ke zhan zhe 夢者人精神所寤可占者).31 These last items concerning the life of the spirit bring us to the crucial topic of sacrificial ritual, regarding which Zheng Xuan brings to bear the full breadth and depth of his learning in explicating certain aspects of the Zhouli ritual system. For example, he provides valuable information regarding the coordination between the four “suburban” altars and the “five sacrifices” (sijiao wusi 四郊五祀), picturesque details on the wall hangings and floor coverings required for the dalü 大旅 rite, bits of lore on the correlation of sacrifices with particular deities such as Gonggong 共工, Goulong 句龍, Zhurong 祝融, and the like, description of the enactment of the dazha 大蜡 expiatory ritual, and explications of other forms of ritual exorcism, including the less-well-known gui 禬 and yong 禜 rites and other responses to natural calamities.32 In one long exegetical excursus, he gives a comprehensive review of the entire sacrificial cycle directed toward various

29 See 14.15a (Diguan: Meishi 媒氏): “Explanation: in the discussion of Wang Su it says: ‘In my youth, while occupied with the study of Zheng Xuan’s exegesis, I searched for the meaning of these misleading words. I then understood that the ancients had the option [of doing this] in winter. From the time of Ma [Rong] on, this was understood according to the Offices of Zhou’” (釋曰王肅論云吾幼為鄭學之時 為謬言尋其義。乃知古可以於冬。自馬氏以來乃因周官。). 30 Regarding funerary ritual, see, e.g., 8.13a (Tianguan: Fengren 縫人): “At the point when the bereaved filial son opens the coffin and looks inside, it is as if he views the physical presence of his loved one” (孝子既啓見棺猶見親之身); and 19.11a: “One’s ancestors take on visible form in this venue, and one makes the sacrificial offering to their spirits to soothe them” (先祖形於此地，祭其神以安之。). On divination, see, e.g., 24.14a (Chunguan: Dabu 大卜). 31 24.13a (Chunguan: Xiaozongbo 小宗柏). 32 See 2.20a (Tianguan: Taizai 大宰), 6.10a (Tianguan: Zhangci 掌次), 12.9a (Diguan: Dangzheng 黨正), 25.6a (Chunguan: Taizhu 大祝), etc.

170

CHAPTER FIVE

“divinities” at different cultic locations. 33 From the perspective of comparative anthropology, I find particularly intriguing his treatment of the topic of blood sacrifice, in which he seems to enter into a running dialogue with his predecessors at various points regarding the efficacy of animal versus grain offerings.34 Of special interest is his recurring emphasis on the motif of “the smell of blood” favored in certain sacrificial contexts, which, together with the image of rising smoke invoked in one discussion, calls to mind the “sweet savor” of smoke columns rising toward Heaven in the sacrifices of the ancient Greeks and Hebrews.35 Finally, I would like to round out this survey of Zheng Xuan’s comments on the Zhouli by considering two areas in which his textual elucidations touch upon broader conceptual aspects of the Zhou ritual system. Zheng’s basic understanding of the concept of ritual, as reflected in a number of his more extended discussions, does not deviate, for the most part, from well-known formulas occurring in other parts of the canonic ritual corpus. For example, the following definition of li seems to echo any number of parallel sources, most obviously the “On Setting Bounds” (“Fangji” !") chapter of the Liji: [T]he five rituals serve to constrain deceptive behavior on the part of the people in all their multitudes and to instruct them in the principle of moral balance. Li is the means whereby one regulates and restrains the extravagant and deceptive behavior of the people, making it possible for them to put into practice the ideal of moral balance. #$%&'()*+,)-.%/#01()2*34567.36!

In keeping with a tendency observed in so much of the Confucian commentarial corpus, Zheng Xuan pays special attention to textual details on issues of calendrical import. Given the fact that even the orthodox commentators on the Gongyang 89, Guliang :;, and 33

See 18.5b–6a (Chunguan: Da zongbo ), 25.6b, 10a (Chunguan: Taizhu