Culture and State in Chinese History: Conventions, Accommodations, and Critiques 9780804765060

Citation preview

CULTURE AND STATE IN CHINESE HISTORY

+

Irvine Studies in the Humanities Robert Folkenflik, General Editor

fij~ [] Jtate IN CHINESE HISTORY

Conventions, Accommodations, and Critiques

+ Edited by Theodore Huters, R. Bin Wong, and Pauline Yu

STANFORD UNIVERSITY PRESS, STANFORD, CALIFORNIA 1997

Stanford University Press Stanford, California © 1997 by the Board of Trustees of the Leland Stanford Junior University Printed in the United States of America CIP data appear at the end of the book

A Note on This Series

This is the ninth in a series of volumes on topics in the humanities and the fourth in the new series published by Stanford University Press. This volume originated in a conference, Culture and State in Late Imperial China: The Political and Cultural Construction of Norms, held in Laguna Beach, California, in June 1992. For help with a broad range of problems, I am indebted to the Editorial Board of the Irvine Studies in the Humanities: Ellen Burt, Rey Chow, Murray Krieger, J. Hillis Miller, Kenneth Pomeranz, Leslie Rabine, David Smith, John Smith, and Linda Williams. Former Dean of Humanities Spencer Olin provided support, including the secretarial help of Joann McLean and Nancy Tablyn. Robert Folkenflik, General Editor

Acknowledgments

The chapters in this volume were first presented at the conference Culture and State in Late Imperial China: The Political and Cultural Construction of Norms, which took place in Laguna Beach, California, in June 1992. Funding for the conference was provided by grants from the Pacific Rim Research Institute of the University of California, the Committee on Research of the University of California, Irvine, and the Humanities Research Institute of the University of California. For their generous support, we are profoundly gratefuL All of us who presented papers at the Laguna conference benefited enormously from the commentators on the proceedings, Professors Bruce Lincoln of the University of Chicago and Hermann Ooms of the University of California, Los Angeles. The insights, critiques, and conclusions they offered helped to give conceptual shape to a rather disparate assortment of papers, and the comparisons they drew to examples outside the sphere of China provided an important perspective within which to situate our discussions. We have aimed to incorporate their observations and suggestions into our revisions and are extremely grateful for their contributions. We also thank the anonymous reader who reviewed this volume for Stanford University Press; this scholar's detailed and cogent comments on the text were similarly helpful to all the contributors. James Benn, Eileen Chang, and Meng Yue, graduate students at the University of California, Los Angeles, supplied helpful editorial assistance in the tedious task of consolidating the character list and works cited, which we gratefully acknowledge. To all the participants, the editors extend their appreciation for their good spirits and collegiality over the course of the event and their extraordinary patience in awaiting the appearance of this volume. The conference was, regrettably, Marston Anderson's last public event before his untimely death in August 1992; we were blessed to be able to benefit from his presence. T.H. R.B.W. P.Y.

Contents

Contributors

XI

Introduction: Shifting Paradigms of Political and Social Order R. Bin Wong, Theodore Huters, and Pauline Yu

I

I. ELITE EDUCATION AND CULTURAL CONVENTIONS

r. Examinations and Orthodoxies: Io7o and I 3 I 3 Compared

29

Peter K. Bol 2.

The Formation of "Dao Learning" as Imperial Ideology During the Early Ming Dynasty Benjamin A. Elman

3· Canon Formation in Late Imperial China Pauline Yu 4· Salvaging Poetry: The "Poetic" in the Qing Stephen Owen

58 83 I05

II. THE POWER OF FAITH

5. A ]iao Is a ]iao Is a ? Thoughts on the Meaning of a Ritual Robert Hymes

I 29

6. At the Margin of Public Authority: The Ming State and Buddhism Timothy Brook

I6I

7· Power, Gender, and Pluralism in the Cult of the Goddess of Taishan Kenneth Pomeranz

I82

CONTENTS

X

III. ACCOMMODATIONS AND CRITIQUES

8. Style and Suffering in Two Stories by "Langxian" Katherine Carlitz

207

9· Ming-Qing Women Poets and the Notions of "Talent" and "Morality" Kang-i Sun Chang

236

The Scorpion in the Scholar's Cap: Ritual, Memory, and Desire in Rulin waishi Marston Anderson

259

The Shattered Mirror: Wu Jianren and the Reflection of Strange Events Theodore Huters

277

IO.

II.

IV. VISIONS OF COMMUNITY AND SOCIAL ORDER

12. Confucian Agendas for Material and Ideological Control in Modern China R. Bin Wong

303

13. Community, Society, and History in Sun Yat-sen's Sanmin zhuyi David Strand

326

14. Constructing the Civilized Community Ann Anagnost

346

Notes Works Cited Character List Index

369 441 475 487

CONTRIBUTORS

+ Ann Anagnost Marston Anderson Peter K. Bol Timothy Brook Katherine Carlitz Kang-i Sun Chang Benjamin A. Elman Theodore Huters Robert Hymes Stephen Owen Kenneth Pomeranz David Strand R. Bin Wong Pauline Yu

CULTURE AND STATE IN CHINESE HISTORY

Introduction: Shifting Paradigms of Political and Social Order R. Bin Wong~ Theodore Huters~ and Pauline Yu

State Power and Cultural Practice Whatever their opinions on other aspects of Chinese behavior, nineteenthcentury Western missionaries in China generally were impressed with what they saw as the law-abiding qualities of the populace. Even Arthur H. Smith, who was assiduous in pointing out the depravity he found in the country, could write: "One of the admirable qualities of the Chinese is their innate respect for the law. Whether this element in their character is the effect of their institutions, or the cause of them, we do not know. But what we do know is that the Chinese are by nature and by education a law-abiding people. " 1 Smith was vague in his admiration, devoting little effort to speculating on the factors behind this singular virtue. Writing at approximately the same time, however, the British journalist Alexander Michie drew the institutional implicatio~s of Chinese social morality: China occupies the unique position of a State resting on moral force, a conception almost as alien to the Western mind as material progress is to the Eastern, hence the proposition is apt to be received with amused contempt. Yet a state administered without police and ruled without an army, is a something which cannot be explained away. Government by prestige is, other things being equal, surely the most economical as well as the most humane of all species of government; but an obvious conseque~ce is that in emergencies the Government is beholden to volunteers, and is often driven to enlist the services of banditti and other forces proscribed by the law. Imperial prestige, which embraces the relations of the surrounding tributaries, is but an expansion of the authority of the head of the family and of the elders of the village, which rests on moral sanction only. 2

The particular historical context of these observations is important. These missionaries were writing in a period when the weakness of the Chinese

2

INTRODUCTION

state in relation to the Western states made that weakness appear remarkable to them. The Italian missionary Matteo Ricci, for instance, never seems to have noticed the similar shakiness of the state apparatus in early seventeenth-century China, perhaps because his presence at court prevented him from observing the kinds of conditions that caught the attention of nineteenth-century witnesses or perhaps because the absence of state control reaching far beyond the capitals of Europe made the Chinese case seem unexceptional to him. The different perceptions of Western observers aside, what demands our attention is the continuing strength of such social morality even in times of state weakness; contemporary sinology has yet to address this strength and its relationship to state power systematically. Over the past forty years, scholars have tended to be skeptical of the imperial state's capacities, despite their occasional assertions of state intent to dominate society. Hsiao Kung-chuan, for instance, began his epic Rural China: Imperial Control in the Nineteenth Century (r96o) with a classic statement about the nature of the central state: "Autocracies such as imperial China presuppose a sharp distinction between rulers and subjects and a consequent divergence of interests between them." Hsiao then presented a short list of coercive means by which the state enforces its will over its subjects, including ideological indoctrination, and characterized the discursive side of the resulting administrative apparatus as a process of "inculcating in their [subjects'] minds carefully chosen precepts-mostly from doctrines of the Confucian tradition-that tended to make them accept or acquiesce in the existing order." 3 But he went on to identify "grave shortcomings" in the Chinese administrative system that contributed to a decline from "an incomplete hold on rural China" to a nineteenthcentury system of rural control that "proved virtually useless. " 4 In a related vein, Arthur Wolf later observed, "Historians and political scientists often emphasize the failure of most Chinese governments to effectively extend their authority to the local level. Certainly many governments had difficulty collecting taxes, and some allowed this function and others to fall into the hands of opportunistic local leaders. Judged in terms of its administrative apparatus, the Chinese imperial government looks impotent." So far, this affirmation of Hsiao's picture of the real weakness of autocratic totalism reads remarkably like Michie's account of the government's lack of coercive tools. Wolf did not, however, end his evaluation with this judgment, for he continued: "Assessed in terms of its long-range impact on the people, it [the Chinese state] appears to have been one of the most potent governments ever known, for it created a

Wong, Huters, and Yu

3

religion in its own image. " 5 While admitting the government's striking deficiencies in material capabilities, Wolf granted it vast agency in the .discursive realm, a conclusion that Michie resisted and Hsiao found not very important because he was principally interested in coercive control. For his part, Smith argued for considerable state power; he contrasted the ability of the state to order people to leave the coastal area being attacked by Coxinga (Zheng Chenggong) in the late seventeenth century with its incapacity to build an effective navy two hundred years later (p. 155). If the contrast that Smith pointed out can be explained by a theory of the state's inability to innovate, the contrast that lies at the heart of Wolf's formulation is mare difficult to interpret. Isn't his model of the state's great ideological power at odds with his sense of the state's limited material capacity? One way to reduce this apparent gap is to grant greater importance to non-state actors who pursue their own interests, in particular to elites promoting their own power and status. Both American and Japanese scholars in the 1970s did just this. In the United States conflict between local officials and an educated, often degree-holding, elite commonly referred to in English as "gentry" was stressed in several areas of interactions between these two groups, especially taxation. 6 Japanese scholars have, in contrast, often stressed the complementary interests of the state and gentry elites; in an influential study on "gentry rule" (kyoshin shihai), the late Shigeta Atsushi spoke of the dominance of gentry in local society as a key feature of late imperial Chinese history. 7 American and Japanese scholars have continued to look at elites in late imperial China, but often with less attention to the state than had characterized the scholarship of the 1970s. 8 Recently, Prasenjit Duara has taken us even deeper into Chinese society, to the village level where gentry are not usually found, particularly in the twentieth-century North China he examines. He finds a "cultural nexus of power" that locates both state and elites in a common framework that highlights "culture." But this "cultural nexus of power" appears to have no particular physical or institutional location. The absence of material loci makes it difficult to track relationships between state and elites, especially those of a cultural nature. Duara's stimulating formulation challenges us to bring the state back into discussions of culture and power and encourages us to search for the institutional and intellectual sites of state and elite competition and accommodation over cultural issues. As we do so, we can recall Michie's observation about the moral force of the state, a claim that urges us to re-examine our assumptions as to how state power functions at various times and places and to

4

INTRODUCTION

recognize as peculiarly modern the notion that the state's coercive power is likely to have greater suasive force than its prestige. There is no doubt that few institutions can claim the historical durability of the Chinese imperial government. However, as the comments quoted above suggest, conventional measures of state strength and weakness based on analysis of European examples may be inadequate to gauge the nature of the imperial Chinese government and its strategies for rule. The European state-making process between 1500 and 1900 was driven, on the one hand, by military actions on the part of centralizing governments that expanded their armies and bureaucracies, and, on the other, by the development of principles of political representation leading to the ideals-if not always the realities-of liberal democracy. In the Chinese case, the use of military force was one of a broader range of tactics of bribery, persuasion, and coercion of groups along the frontier; the European complex of war-making supported by increased central government extraction coupled to a political ideology inventing strong claims to representation was absent. Nonetheless, a considerable consolidation of Chinese central authority did take place during the late imperial period (roughly 1500 to 1900)-a consolidation that consequently resists analysis in terms of concepts borrowed from an early modern and modern European story. In the passage quoted above, Arthur Wolf reminds us that the empire was always concerned with more than garnering silver to pay its bills and raising troops to defend its borders, and in some of these other areaswhat we loosely call culture and the construction of norms-it appears to have been surprisingly influential. He draws attention to one of the best-known means by which local gods were conceptualized in Chinese culture-in terms of a celestial bureaucracy clearly modeled on the organization of the temporal government. This was by no means the only model, and the specific agency he attributes to the state in developing it is arguable. To what extent officials were the sole agents behind such cultural constructions and to what extent they were actually effective are questions the chapters in this volume seek to explore. Drawing upon a broad vision of politics found in classical texts, late imperial Chinese officials and intellectuals judged governments in terms of their success in improving both their own moral status and that of the populace. Culture, in the more specific sense of creative arts and religious faith, presented itself both as an index of the state's achievements in this area and as a repertoire of tools available to officials and elites to create desired moral conditions among the people. Because the state understood the art of ruling to include shaping people's moral behavior, it was inclined to be meddlesome, authoritarian, and censorious. But at the same time,

Wong, Huters, and Yu

5

because only people who freely choose to do the right thing can be said to have high moral standards, the state accepted and even promoted initiatives among its subjects that it saw as likely to make people take their moral and social agency seriously, including projects that a narrowly instrumental and materialist state would probably have suppressed as threatening to its rule. An early modern European state would not, for instance, have been able to accept, let alone advocate, elites mobilizing men and resources to form their own military forces; indeed European state makers sought to assert centralized control over military bureaucracies. Yet the late imperial Chinese state did just this, exhorting elites to mobilize militia to defend the empire against rebel groups. For their part, elites and commoners had ideas about appropriate beliefs and behavior that sometimes competed with the policies and preferences expressed by officials; at other times these represented complementary efforts by non-state actors who shared the state's agenda and seemed, unlike roughly comparable groups in European or twentieth-century Chinese settings, to have been far more concerned with that shared agenda than with "turf" battles with the state. Indeed, there is little evidence to suggest that they framed the question of agency in a way to separate themselves clearly from the state. Because these groups acted as allies on some occasions and antagonists on others, our late twentieth-century assumptions about modern historical change are an inadequate guide to thinking about political and social order in China. Western theories centered on the formation of a sharp divide between state and civil society and between state power and personal autonomy develop nicely out of various readings of European history. When confidently promoted as universal or general processes, however, they limit our capacity to conceptualize the political and cultural terrain upon which relations between state and various social groups were constructed in late imperial China. The chapters in this volume suggest that the state was only one set of actors-albeit often the most powerful set-in a culture that elites and commoners could shape on their own, separately or together, in cooperation or competition with the state. Furthermore, if we ascribe to the Chinese state an intent to influence cultural practices only in order to promote social order and stability, we obscure its own placement within the culture. In other words, the government could not embark on projects and efforts from a privileged position outside Chinese culture. It follows that analyses of Chinese government activities considered only in terms of categories derived from European experience may prevent us from recognizing the cultural specificity of Chinese state practices. For instance, a county magistrate's response to a severe drought could include making

6

INTRODUCTION

plans to reduce the price of grain by selling grain from the government granary, calling on local elites to sell or lend grain on favorable terms, or arranging for imports of grain through either official channels or private ones; the same official would very likely go to a nearby temple and pray to a deity responsible for bringing rain. It is difficult for a Western perspective to embrace price reduction efforts and prayers for rain equally, but given the Chinese vision of correlative associations between events and processes on earth and in the heavens, a late imperial official might place equal faith in both types of efforts. Once we understand the embeddedness of political practices within the culture as well as the cultural commitments of the state, the moral elements in Chinese ideas of what it meant to rule take on much more complicated shadings. So, too, does the more ambiguous separation of "state" and "society," compared to developments in Europe. What a Europeanist might see as signs of a failure of state power or a successful assertion of "society's" power against the state may be better conceived as an expression of the imperial system's permeation of state and society. We think of Chinese strategies for ruling as having been chosen from a repertoire of Confucian precepts and policies, but this has tended to obscure two important features of China's Confucian ways. First, as is well known, much of the culture lay outside what people of the time recognized as Confucian: Buddhism and Daoism in particular were powerful elements in state, elite, and popular life. Second, Confucian culture in China was not a narrow, rigid, and stagnant set of pieties reproduced mindlessly through the centuries; it was far more plastic than we often assume, alternatively open to new possibilities and doggedly defensive about nonconformist practices. Worse yet, we are ill-prepared to consider how Chinese cultural categories facilitate certain kinds of changes rather than others. Since Confucian categories of thought rarely yield the particular kinds of changes we are most accustomed to finding in Western historical transformations, the range of possibilities they do enable becomes difficult to apprehend from contemporary historiographical and theoretical perspectives. Indeed, because this Chinese range does not seem to have included the Weberian types of modernity that have framed the analysis of European history, many scholars have considered this very flexibility unimportant in understanding developments in twentieth-century China. Instead, scholars of post-imperial China have more often stressed the new possibilities introduced by the import of Western models, an inclination reinforced by the Chinese architects of a post-imperial state who themselves self-consciously

Wong, Huters, and Yu

7

emulated Western models in order to compete internationally and gain recognition as modern. Many in power made the unprecedented assumption that traditional culture was irrelevant, an assumption that would have dire consequences in the ideological realm and ultimately in practice as well. Would-be modernizers, who thought their immediate predecessors had been imprisoned by their imperial heritage, aimed to shed these traditional shackles or, on some occasions, simply to employ older ideas instrumentally and cynically. In imagining positive scenarios of social change, Chinese leaders usually reached outside their own culture to capture strategies and sensibilities deemed more advanced, even as they continued to face challenges rooted in their own "backward" culture. They often thereby remained blind to-or perhaps deliberately concealedtheir very indebtedness to that past. Assessing the nature of relations among the state, elite thought, and popular culture throughout China's history will require much new research over a wide range of specific topics. The chapters in this volume seek to contribute to this project by refining our understanding of political and cultural initiatives and contentions leading to the formation of social order. Although we focus on the late imperial period, we examine both earlier and more recent contexts as well, since the relationships at work and the problems at stake come into sharper view through consideration of this longer stretch of Chinese history. The chapters of Part I, "Elite Education and Cultural Conventions," suggest new perspectives on philosophical, historiographical, and literary debates and their relationships to the late imperial Chinese state. Part II, "The Powers of Faith," focuses more particularly on religious behavior to reveal the multiple roles of officials, elites, religious specialists, and commoners in constructing norms of belief and practice. The authors of the chapters in Part III, "Accommodations and Critiques," present criticisms by late imperial intellectuals of state policies and explore the adaptability of cultural conventions to embrace social change. And finally, the chapters in Part IV, "Visions of Community and Social Order," discuss ways in which late imperial and post-imperial officials have imagined and at times forged links between the central government and local social order.

The State, Elite Education, and Self-Cultivation The Chinese state cared about and intervened in upper-class education in a manner unusual in world history. Individuals earned the opportunity to serve as officials by passing civil service examinations. Elite education

8

INTRODUCTION

was, not surprisingly, geared to a high degree toward mastery of the texts and themes deemed appropriate by those charged with setting the examination questions. This is not to say, however, that educational priorities were at all times uniformly shared: the curricula of academies and local schools, for example, were not always strictly tailored to those of the examination and in fact at times stressed interpretive stances directly opposed to the officially mandated ones. Furthermore, the relationship between government notions of proper learning and those of intellectuals more generally also changed significantly over time. As Peter Bol explains in "Examinations and Orthodoxies: 1070 and 1313 Compared," literati successfully opposed the efforts of the statesman Wang Anshi in the ro7os to narrow the focus of the examination questions as a means of establishing a uniformity of thought. Yet they accepted the imposition of a nco-Confucian curriculum in 1313 because they were able to see it as very different from prescribing acceptance of particular views. Bol presents evidence of variations and differences in the thought of early fourteenthcentury intellectuals and argues that contemporary literati were willing to accept a state-sanctioned orthodoxy because they believed that it left open the possibility of individual reflection, participation in an ongoing interpretive tradition, and implementation of moral action-all in ways that Wang Anshi's program had not. Similarly, a more anxious and narrow-minded concern for orthodoxy would probably not have enabled the very different interpretations in the early and late Ming ( r3 68-r643) dynasty of the Yongle emperor's succession, discussed by Benjamin Elman in "The Formation of 'Dao Learning' as Imperial Ideology During the Early Ming Dynasty." After describing the Yongle emperor's bloody usurpation and his creation of a narrative of legitimate succession, Elman discusses late Ming scholars' critique of the usurpation, a critique appealing to the same reservoir of principles that had once been used to sanction and consolidate Yongle's accession to the throne. Confucian ideology thus served as a double-edged sword, sustaining first a ruthless grab of imperial power and then idealistic protests from the literati two centuries later. The late Ming court itself actively supported a compelling counter-narrative that portrayed the Yongle emperor-from whom the reigning emperor directly descended-as the perpetrator of the most profound injustice, and it restored passages of the Mencius that he had excised from the curriculum because they offered critical leverage to oppositional elements. While we might be tempted to interpret this as simply reflecting the relative relaxation of an authoritarian regime-a familiar scenario in any number of cultural and historical

Wong, Huters, and Yu

9

contexts-the harsh manner in which the late Ming state dealt with the Donglin movement of dissenters from state policies and practices makes clear that no simple relaxation was taking place. Even if a "relaxation" thesis made more empirical sense, it would remain difficult to explain the court's promotion of what could be seen as a threatening text, one that in fact conceivably undermined its own legitimacy. To make sense of this situation, we must recognize the remarkable fluidity of the ideological canon and the extent to which-from a statecraft standpoint-it allowed for a variety of constructions that sufficed to convince active members of the polity to stay in a game marked by a generally accepted set of rules. Chinese officials were expected to be men of moral vision and upright virtue. They were not construed as paper-pushing functionaries operating according to clearly articulated rules; rather, they (especially those at high levels in provincial and central government positions} were seen as individuals who could exercise educated judgment on difficult issues. Moral sincerity and personal character, therefore, prepared officials for bureaucratic success. Problems were conventionally seen to stem not from organizational failures, though reforms could be called for, but from ethical flaws of individual officials. Officials devoted to major organizational reform, like Wang Anshi, were often criticized vigorously, whereas those, like the Ming dynasty figure Hai Rui, who leveled criticisms at the emperor that would strike many a Western bureaucrat as both eccentric and ineffective, were taken as models worthy of emulation. Thus Chinese officialdom could never be, even in ideal terms, the sort of rule-governed rational decision-making body envisioned by Max Weber as the modern bureaucracy. Rules there certainly were, but the opportunities for choice among them to fit a particular situation and the many possible applications for any one cluster of regulations meant that officials could not be expected to deliver a single rational decision on any case. Philip Kuhn's depiction of a 1768 sorcery scare captures the particular uncertainty and anxiety of officials aiming to respond to the emperor's fears of evil spirits. 9 Nor were Chinese officials of the Ming and Qing (r644-191I) dynasties-either real or ideal ones-likely to have been individuals that Queen Elizabeth I, Louis XIV, or Frederick the Great would have embraced for service in their expanding bureaucracies. European centralizers recruited officials whose loyalty they could depend upon; these officials came from the small stratum of educated elites-clerics, lawyers, and nobility. Technical skills, especially abilities to handle fiscal and military affairs, were particularly valued in the English, French, and Prussian cases. 1°Chinese bureaucrats, in contrast to both Weberian ideals and the realities of early modern

IO

INTRODUCTION

European state-making, were charged with the responsibility to decide for themselves on proper courses of action, a demand they did not always meet aggressively-at times they chose inaction as the safest way to avoid making mistakes. But even when failing to act, they could not seek to behave as narrowly focused functionaries conditioned by an efficient and effective calculus of rationality to make logical and correct decisions. They were in fact as well as in theory more like learned scholars versed in broad philosophical principles as well as more focused empirical fields of study. Poetry was one arena in which both intellectuals and the state legitimated themselves by affirming shared cultural ideals. The writing of poetry formed part of moral instruction because it could, ideally, enable human passions to find appropriate expression; good poetry could both serve as an index of the achievement of a larger moral order and inspire good conduct. The Ming and Qing governments' projects to collect and select poetry of earlier dynasties were elements in a larger process discussed by Pauline Yu, who focuses in "Canon Formation in Late Imperial China" on how Tang (6r8-9o6) poetry achieved a privileged position within this cultural discourse. Poetry in the Tang, especially during the eighth century, had been central both to literati self-definition and to the state as a skill to be demonstrated on the civil service examinations, but for much of the succeeding millennium it was not a central element in the examinations. Until its reinstitution as an examination requirement in 1757, poetry could simultaneously take for granted its political and cultural importance, the legacy of the Tang, and take refuge in a certain aesthetic distance and incorruptibility. Yu explains the return to a testing of poetic composition as both part of the construction of a literati self-image seeking to recapture past ideals and an acknowledgment that identifying a set of canonical texts and methods to teach them made poetry writing a skill that could be taught and tested in systematic ways. Tang poetry made good sense pedagogically because of its regulated themes and forms, to which critical and analytical terms developed in connection with the formalistic "eightlegged" essay (bagu wen) that had dominated the examinations for centuries could conveniently be applied, and ideologically as the expression of a cultural moment during which poetry clearly had mattered. In "Salvaging Poetry: The 'Poetic' in the Qing," Stephen Owen discusses the ways in which Tang poetry also ultimately represented an ideal that late imperial literati could no longer hope to achieve. They were forced to confront the increasingly unavoidable tension, one that had existed within formulations of poetry from the earliest theoretical articulations,

Wong, Huters, and Yu

II

between poetry as the genuine, spontaneous expression of feeling and poetry as part of a civilizing program bearing moral, regulatory force. Far more poetry was written in the Ming and Qing dynasties than in earlier times, but this very proliferation, perhaps, made it difficult to escape the potential for irony introduced by vernacular literature, which-in its focus on human failings and the hypocrisy of much social conventionincreasingly represented "the poetic" as something contextualized and the result of deliberate effort, rather than something authentic and natural, and thus not really an expression of human passions that could be an alternative to less desirable expressions. Since late imperial poetry could not occupy an unmarked aesthetic space like that Tang poetry was felt to have enjoyed, it could never claim the same creative successes. The inevitable impingement of the social thus threatened to condemn Ming and Qing literati to a world of artistic emulation, understood on one level to fail for reasons beyond the control of individual literati. At the same time, however, as Owen demonstrates in his reading of Hong Sheng's play "The Palace of Lasting Life," this account of the legendary love between_,-a Tang emperor and his consort thematizes the crossing of boundaries and the conflict of motives and creates thereby a literally separate space for genuine passion, allegorically recapturing an isolated, transcendent lyricism in heaven. Significantly, however, this redemption of the poetic and the absorption of pure feeling can take place only by literally abandoning the real historical world with its multiple versions of truth and lie. The social dimensions of the "poetic" set boundaries beyond which a stable and separate aesthetic space could not exist.

Religious Experiences: Contending Concerns The effort to salvage a poetic space discussed by Stephen Owen takes place independently of concerns regarding strategies for ruling and explores the impulse to establish an autonomous aesthetic realm, however politically and socially contextualized. Robert Hymes's "A]iao Is a ]iao Is a ?: Thoughts on the Meaning of a Ritual" also focuses on practices ostensibly remote from central political activity, and his analysis offers as well a particularly illuminating example of the way in which practices or policies could be simultaneously understood in mutually contradictory ways by those involved without threatening the performance of rituals. Hymes argues that participants in Daoist jiao, the central sacrificial ritual of the Daoist liturgy, manipulate symbolic power in various ways that affirm their competing understandings of the ceremony without directly forcing

12

INTRODUCTION

the other parties involved to acknowledge their versions and the conflicting social hierarchies thereby implied. This analysis undermines simple binaries of orthodox and heterodox that often structure discussions of such rituals. It also puts into question the more nuanced notion of a continuous spectrum ranging from practices promoted by the state to those preferred, those deemed acceptable, those tolerated, those opposed in principle, and those opposed in actual practice. We are confronted here with a phenomenon that cannot be reduced to a single set of meanings agreed upon by the participants, who are not, moreover, by any means indifferent to the discrepancies among interpretations; in fact, their projections make possible a situation in which each group can think that the ritual affirms its superiority to and power over the others present. The Daoist jiao is, significantly, a ritual without an explicit Confucian voice. Indeed, the absence of either an official or elite interest in establishing a set of Confucian meanings for this ritual may have had something to do with the openness, indeterminacy, and contradictions of its interpretations. This does not mean that the elites engaged in Daoist jiao were not also Confucian; rather, they did not bring to this ritual any specifically Confucian priorities that could conflict with those of the jiao. Of course, much of Chinese thought and ritual practice lay outside any narrowly defined notion of Confucian ideology; Buddhist and Daoist beliefs and practices did have a place in a Chinese cultural order in which emperors recognized sacred Daoist sites and the Qianlong emperor had a Daoist portrait of himself painted. At the same time Confucian precepts often inspired officials to monitor, and sometimes to control, religious performances. The Daoist jiao examined by Hymes offer an instance in which such impulses appear absent; indeed officials themselves could even commission these performances. Timothy Brook's "At the Margin of Public Authority: The Ming State and Buddhism" examines a situation in which state concerns about religion were, in contrast to Hymes's example, important, and in which elite perceptions of the state's proper relationship to Buddhist monasteries changed significantly over time. After pointing out that European patterns of relations between church and state provide no useful guidance for understanding Ming Buddhism's relationship to the state, Brook notes a dramatic shift from imperial support for Buddhism in the first decades of the Ming to policies designed to limit Buddhist influence in later years. Ming and Qing emperors did not, however, uniformly oppose Buddhist rituals or Daoist practices, and, as Brook argues, although they sometimes reversed themselves in the Ming from open promotion to explicit regulation, such

Wong, Huters, and Yu

!3

efforts were seldom effectual. Perhaps this reflects the real limitations on Ming bureaucratic capacities; perhaps it signals to us that the state simply did not care enough to intervene more rigorously. Whatever the reasons for the modesty of state efforts at control, we have known for some time that Buddhism as a set of doctrines enjoyed widespread credence during the late Ming, and Brook contributes to this understanding by demonstrating the gentry's decision to patronize Buddhist monasteries as a tactic to attain a more elevated local status in many parts of China. In this way, gentry elites carved out a sphere of activity separate from official Confucian priorities or concerns for local order, but at the same time they continued to fulfill such obligations in other ways, such as establishing lineage organizations, schools, and granaries. The subsequent decline in patronage of Buddhist monasteries by the eighteenth century seems to have been the result more of a shift in gentry attitudes than of any direct state effort to censure such activities. But while changes in patronage may not have been directly caused by government directives, Qing officials clearly cared about promoting proper Confucian practices and therefore opposed certain types of religious behavior of which they were wary. These priorities did not conflict with the great faith of the eighteenth-century emperors in Tibetan Buddhism, for themselves personally and as a means of ruling other Inner Asian peoples, especially Mongolians, who were followers of this form of Buddhism. Within China proper, the Qing state promoted a Confucian agenda of rule. Kenneth Pomeranz's "Power, Gender, and Pluralism in the Cult of the Goddess of Taishan" presents an example of efforts by officials and elites in the eighteenth century to control worship of the goddess Bixia Yuanjun. Proper Confucians worried about this deity, in part because women in great numbers went on pilgrimages to her shrine at Mount Tai. Among those especially likely to make this journey were midwives and healers, women whose occupations could not be easily contained within Chinese visions of a stable Confucian moral and social order. Moreover, however orthodox the ends for which Bixia Yuanjun was usually invoked, her powers relating to female sexuality in particular were clearly dangerous and made elites nervous in a way that the worship of other deities popular among women never did. Efforts by her followers to incorporate Bixia Yuanjun into the orthodox pantheon by sanitizing her disquieting elements foundered, for they would no doubt have destroyed or radically altered the cult's basic appeal. The elite "guardians of orthodoxy," as Pomeranz calls them, acted independently of official dicta and yet served a vision of Confucian cultural

14

INTRODUCTION

order that the government could comfortably accept. For their part, officials had to tread more carefully than righteous elites because Bixia Yuanjun appeared, after all, to be an efficacious goddess. Devotees, for their part, simply sought state sanction for their worship; they seemed willing to accept a subordinated space within the state-ordered spiritual hierarchy as long as it involved some recognition of the social and cosmic powers their deity represented, but even this was difficult to obtain. Indeed, the very efficaciousness of Bixia Yuanjun, as measured by widespread belief in her powers and the location of her principal shrine at such a prominent spot, made Qing officials and elites unable to ignore her disruptive potential. Officials shared an anxiety acted upon more directly by those Qing elites who found granting the goddess even a subordinate niche as a supporter of orthodox order too great a concession. The government and elites, therefore, cared in different ways about religious beliefs and practices, which were in any event far too diverse to incorporate within a narrow definition of orthodoxy. Confucian concerns varied from context to context, shaped by elite priorities and official recognition of limited bureaucratic capacities to control social practices. However hostile elite sponsors of the Daoist ritual discussed by Hymes may have been to its practitioners, each side was able to co-opt the other by developing contending understandings of who really wielded power. Brook's essay suggests ways in which Buddhist practices appear not to have truly mattered to the government in the late Ming; alternatively, perhaps the state was simply too weak to articulate its opposition to monastic patronage. Pomeranz's essay shows that religious meanings could matter to the government, and perhaps even more to elites, as increasingly rigid definitions of orthodoxy took shape. We can discern here a temporal arc from late Ming elite promotion of Buddhist activities to Qing preoccupations with Confucian social and moral order; outside of these seem to rest activities like the Daoist jiao, with its inherently incompatible symbolic representations of power. This arc defines a kind of Ming-Qing state strengthening that fails to follow European lines of development. As elites began to flex their muscles on matters of faith and ritual practice, their efforts conformed to what we can consider an agenda for cultural order shared by officials and elites alike. This cultural transition from Ming to Qing and the varying roles of the state and cultural elites under each dynasty also raise in a fresh manner questions about the dimensions of spiritual belief in late imperial China and the manner in which these changed for the state, elites, and the population at large.

Wong, Huters, and Yu

15

Acceptance and Critique of Social Change The four chapters in Part III consider some of the critiques of a changing social and political order between the sixteenth and early twentieth centuries, as well as methods for coping with the challenges of rule. Katherine Carlitz argues in "Style and Suffering in Two Stories by 'Langxian"' that some Ming dynasty literature reveals the displacement of state-supported norms of virtue to critique both a government and a society that are seen to fall far short of Confucian ideals. Critique became necessary because late Ming economic changes created desires for wealth, burdens of government service led to bankruptcy, and human feelings were thwarted by social convention and hierarchy. Through Carlitz's reading of late Ming fiction and drama, we see how Confucian categories could be used to inveigh against changes such as commercialization and the monetization of tax payments and also be deployed to criticize vestiges of a more rigid and simple social hierarchy from earlier times. Kang-i Sun Chang's study of women's literary production moves us from the late Ming examined by Carlitz into the Qing dynasty and from issues of critique to strategies of accommodation. In "Ming-Qing Women Poets and the Notions of 'Talent' and 'Morality,"' Chang begins by examining the literary position of courtesans in the late Ming who became recognized for Confucian virtues of uprightness and loyalty, thereby breaking new ground for women writers and attesting to a measure of flexibility in the application of Confucian norms to a changing social order. During the Qing, however, courtesans were displaced as writing women by married gentrywomen poets, who were deemed better suited to upholding a conservative, didactic vision. The Confucian accommodation of gentrywomen's poetry-despite strong oppositional voices like that of the eighteenthcentury scholar Zhang Xuecheng-significantly reconfigured the parameters of orthodoxy, expanding the range of acceptable social roles and creative activities for properly Confucian women and opening up a new cultural space. However, since much of the justification for the new role of elite women as writers was based on the claim that their work embodied a superior morality, an older tradition of courtesan poetry had to be read out of the transformed canon of women's poetry; thus a new boundary was erected even as an old one was erased. Amid the social changes spanning the sixteenth through the eighteenth centuries, Confucian critiques of state and society continued. In "The Scorpion in the Scholar's Cap: Ritual, Memory, and Desire in Rutin waishi,"

r6

INTRODUCTION

the late Marston Anderson explains how twentieth-century readers of Wu Jingzi's novel are often disappointed by the evenhandedness of its satire and by critiques that do not represent a consistent perspective or substantiate a unified, coherent analysis of social and political troubles. This dissatisfaction suggests, however, that modern scholars have ignored the context in which the novel was produced, owing to their desire to use this historical text to pursue their own agendas of sociopolitical change. As Anderson points out, in appealing to classical Chinese notions of ritual (li) to order society through the mediation of human desires and meaning, Wu Jingzi came to confront and explore the seemingly unbridgeable gap between moral intentionality and a real historical world that dooms utopian visions; the constant re-creation of social rules and moral applications allowed for no fixed standpoint from which morality could be dictated or even inferred. However nostalgically the characters might hope for the efficacy of the Confucian ritual that constitutes the core of the novel, the narrative makes clear that such ideals cannot be reified and assumed to provide an infallible guide to conduct. The incapacity of the system to present a persuasive vision of human possibility is matched by the failure of others to mount a consistent critique. Together these phenomena suggest the uncertainty and instability that bound the successful adaptation of Confucian categories to the changing world of China between the sixteenth and eighteenth centuries. A similar inability to cope completely with social change, this time in the early twentieth century, is addressed in Theodore Huters's "The Shattered Mirror: Wu Jianren and the Reflection of Strange Events." This essay helps explain how difficult it was in the first decade of the twentieth century to imagine new paths of political and social transformation. The novel Huters analyzes, Ershinian mudu zhi guai xianzhuang (Strange events witnessed over twenty years), was written during the last decade of the Qing dynasty, as a work of social critique. It builds on a number of perspectives new to Chinese narrative in its plot development and point of view, but in the end the text fails, as Huters explains, to synthesize an internally consistent and cogent critique. We might have expected new critical positions associated with Western ideas to offer some interpretive leverage over social change, but just as Wu Jingzi suggests the futility of social action in Rulin waishi (The Scholars), so Wu Jianren's twentiethcentury narrative calls for a more effective sociopolitical order without being able to formulate the consistent ideology necessary to achieve it. Wu Jianren's attempt to stand outside Chinese culture and gain a critical purchase on it ultimately fails. Even if he is wrong (and if he is not, then

Wong, Huters, and Yu

I7

almost all analyses of China have been misguided), the challenge of finding a critical vantage point from which to evaluate the Confucian order remains a difficult one. The very evenhandedness of the tradition represented by The Scholars ill-prepares Chinese writers to gain rhetorical control over a situation that has now become critical. By the late imperial period, literati had created an intellectual culture far broader than that directly tapped by the state, even if the mid-eighteenthcentury state appropriated poetry for the civil service examinations. This culture, as the chapters in Part II suggest, could include on occasion practices like Daoist rituals and Buddhist monastic patronage that were well outside Confucian conventions. In Part III, we learn that literati culture could domesticate changes in gender roles accompanying social and economic transformations between the sixteenth and eighteenth centuries at the same time as the capacity for political and social critique continued to grow. The development of critical perspectives was supported by the growing numbers of literati who never served in the bureaucracy and found alternative careers outside government, a trend created by the limited number of offices and growing number of degree holders. But distance alone did not create an autonomous space within which literati could view government. The late Ming Donglin movement, a group of officials and literati who opposed government corruption and factional favoritism, did not serve as the basis for an opposition party. Nor did literati make claims for an institutionalized position independent of the government from which they could level criticisms buttressed by their own independent force and legitimacy. Individual disaffection and retreat from the world of government service never posed a serious threat to the government.

Community, State, and Social Order The disjunction between certain kinds of Confucian moral ideals and political realities that Anderson's analysis of The Scholars draws out persisted through the eighteenth century. But this situation did not preclude major efforts to create and implement a distinctively Confucian agenda of local rule in this period. In "Confucian Agendas for Material and Ideological Control," R. Bin Wong examines eighteenth-century programs for creating institutions of local order, focusing in particular on granaries and schools. He argues that the mix of official and elite activism that developed and sustained these local institutions varied across the empire, with officials playing larger roles in peripheral areas where replicating the conditions necessary to sustain a stable agrarian order was considered a sign of

r8

INTRODUCTION

successful moral rule. Both officials and elites participated in the formation of community-level institutions that organized important features of social life. The construction of community in agrarian China should not be imagined, as it oftenis in European history, as having resulted entirely from private actors and non-state activities and as having existed as some earlier form of social life on top of which the state was constructed. The joint participation of eighteenth-century officials and elites in a manner that admits of no easy separation between them for the purposes of creating local institutions for social order suggests to Wong that the notion of a "public sphere," a realm of public opinion and political activity that emerged in post-feudal societies, which was first debated in studies of European history and then promoted as a vehicle to explain Chinese social change, is not in fact very helpful for understanding China. What happened in China in the nineteenth century, according to Wong, was neither the collapse of state efforts at the local level nor an aggressive seizure of the state's turf by assertive elites, but a breakdown of vertical integration in the bureaucratic oversight of institutions charged with promoting local order, an integration that is often difficult to see in the discussions of local institutions themselves. The breakdown mattered, even if the local institutions continued to work without oversight, for it led to a decline as China moved into the twentieth century and the uncertainties of local order in the period following the collapse of the Qing dynasty in 19rr. The early decades of the twentieth century were filled with political uncertainty and, as it turned out, profound difficulties in forging a stable post-imperial political and social order. Yet these were also times of tremendous possibility in which we might have imagined-from the vantage point of modern state-making-a Chinese state marching forward to constrict, or even eliminate, some of the cultural spaces that the late imperial government appears to have tolerated or even promoted. The expanding presence of Western ideologies, organizational models, and bureaucratic strategies could have been deployed to further a broad political agenda to control culture. We might also easily have imagined-from the vantage point of modern liberal social thought-the creation of a more fully autonomous social and intellectual elite committed to forging perspectives independent of the state from which to view politics and culture. But neither of these processes unfolded in the early post-imperial years: China neither adapted many Western political institutions and practices in the manner of Japan nor created a "public sphere" like that found in Western Europe and North America. And indeed, given the patterns of political and cultural practices examined in the chapters discussed thus far, there is no compelling reason to have anticipated developments in

Wong, Huters, and Yu

19

such directions. Despite Chinese elites' awareness of the West in the twentieth century and the aggressiveness with which powerful foreigners asserted the superiority of their institutions, assumptions about statesociety relations developed from European experiences do not consistently prove helpful for understanding post-imperial Chinese politics. A crucial issue in the twentieth century remains how to create local social order and use that to build a larger political system. To complicate an already difficult challenge, this task is now conceived within an expanded frame of reference in which China's construction as a "nation" among an international system of nations becomes a pressing task. In "Community, Society, and History in Sun Yat-sen's Sanmin zhuyi," David Strand considers how one major leader, Sun Yat-sen, grappled with defining China's political future as a nation. Sun is conventionally regarded as a loose thinker who inconsistently mixed various Western ideas, but Strand aims to treat him more seriously, examining in particular how Sun articulated a social basis for constructing the Chinese "nation." He finds that Sun used family, lineage, and other indigenous social forms, locating in kinship ties a primal source of the societal unity that impressed him about the West and linking this communalism in turn with nationalism. Sun set himself the challenge of creating a new sensibility, that of "nationalism," out of older sentiments rooted in local communities. His effort represents a key example of a Chinese political leader attempting to construct new visions, in this case of the "nation," upon categories that ordered China as an agrarian empire. Sun's rhetoric proves important not merely because of his role as a political leader but also because, as Strand points out, of the resemblance between this strategy and those embraced by other important political leaders including Mao Zedong. How problematic Lawrence Stone's statement about the modern English state and kinship would have sounded to either Sun or Mao: "The modern state is the natural enemy to the values of the clan, of kinship, and of good lordship and clientage links among the upper classes, for at this social and political level they are a direct threat to the state's own claim to prior loyalty." 11 Here we find a clear declaration of the modern European separation of state and society basic to our notions of what modern European state-making has been about. The construction of new ideologies in China, however, appeals to understandings of Chinese politics and society based as much on earlier domestic experiences as on new knowledge learned and imported from the world at large. The Communists broke with many of the older institutional forms and ideological views, as much scholarship has skillfully demonstrated, but the revolution was not as sharp and complete a rupture as the Communist

20

INTRODUCTION

narrative would have us believe. When we turn to political institutions, we find that post-imperial governments had a difficult time defining the relationship of their policies to those of earlier times, for the general rejection of Confucian principles of rule made it difficult to recognize connections to past practices. Ann Anagnost offers us an excellent example of these problems in the period after 1949 in "Constructing the Civilized Community." The officials she discusses demonstrate an ambivalent recognition that the village compacts (xiangyue) they are promoting as a local institution to mediate disputes and promote harmony and order have a pre-1949 history. They are also uncertain about how large a role they should play and how much initiative they should assign to the local community in order to build these compacts, a problem with which their predecessors in earlier centuries had grappled as well. The compacts, therefore, carry a number of mutually contradictory meanings reminiscent of the multiplicity of perspectives we have seen to be characteristic of other situations in late imperial China: designed by the central authority as a means of extending discipline and control, they are also simultaneously represented as evidence of a more participatory local politics, an expression of the spontaneous will of the people, and an invention of the peopleeven as their well-known Qing dynasty antecedents are freely admitted to constitute a problem. The term xiangyue was used both for the late Ming village compact, an organization that mobilized people horizontally throughout the village, and the Qing village lecture system, a vertically structured system of village elders and local officials delivering didactic lectures to the people; the tension between an institution that mobilizes horizontally within the community on the one hand and a vertically structured and state-centered organization on the other hand is the same one contemporary Chinese officials face. The multiple uses of the term yue in late imperial times embraced the notion of "contract" among relative equals, an "agreement" between superiors and inferiors, and, no doubt most exasperating for officials, a covenant among people engaging in protest over issues like rent payments. 12 The possibilities and pitfalls of creating village compacts in contemporary China show us how issues of power and control in the Chinese countryside may not fit neatly into Western categories of analysis. They remind us rather of the plasticity of earlier Chinese practices and the competing meanings created in situations like the Daoist jiao studied by Hymes. This discovery suggests that the kinds of structures and organizations imagined for a particular situation emerge from a culturally specific repertoire of possibilities. Certainly the options have shifted over

Wong, Huters, and Yu

21

time, but the institutional choices remain structured by culturally defined criteria; for new alternatives to become persuasive, they must either address the structural issues defined by older practices or successfully displace them. Constructing a viable relationship between local agrarian communities and a larger state structure is simultaneously an issue rooted in the Chinese political tradition and a problem of "nation-building" in the modern world. The double meaning of the challenge affirms both the historical and cultural specificity of Chinese constructions of relationships between community and state power and the larger set of systemic problems created by the global expansion of Western political and economic power in the nineteenth and twentieth centuries.

Implications and New Directions The chapters in this volume demonstrate the variety of ways in which the power of the Chinese state over spaces both physical and conceptual was contingent and incomplete. Addressed in the Wong and Anagnost chapters as a challenge to be faced by the state, the spaces become arenas for competing or complementary agendas among a range of actors in the three chapters on religion. Contending interests can be mutually accommodated in alternative interpretations of the Daoist ritual examined by Hymes, and, as Brook's chapter shows, imperial pronouncements on Buddhism can be made without challenging local gentry practices. These chapters also point out the difficulties of separating out the specific agency of officials and elites in local initiatives to exert power and control. Brook's late Ming gentry assert local cultural hegemony by monastic patronage not supported, let alone encouraged, by officials, and the Qing elites in Pomeranz's discussion more actively promote Confucian orthodoxy than officials do against people who probably saw their efforts as on the whole complementing the agenda of both state and elites. The Qing elites and officials in Wong's account play basically complementary roles, while in recent years, as Anagnost makes clear, the post-1949 state can bureaucratize local elites in a manner unthinkable in late imperial times. A better understanding of the reach of the Chinese state, its embeddedness in cultural matrices, and the peculiarly conglomerate nature of its composition will come from considering more instances of state activity at the local level. At present we can only speculate on what future research may suggest about temporal changes in state strategies to co-opt and control local social elites. We do know that the granaries and schools that Wong finds

22

INTRODUCTION

to be jointly founded by officials and elites in the eighteenth century were intended to be local institutions outside formal government administration when proposed in the twelfth century. This move to set up local institutions outside government control took place in the same period that Hymes's local elites promoted direct appeals to the Daoist Immortals who sat outside the bureaucratic organization of deities. Yet within Hymes's Daoist case, as he notes, there already exists the paradoxical agreement of the petitioner to perform an elaborate sacrificial ritual steeped in symbols of mediated authority. In institutions like granaries and schools, we see a contrast between the ideal of local institutions being created and sustained outside formal bureaucratic authority and the periodic promotion by officials in both the Ming and Qing dynasties of these institutions with varying degrees of official involvement. The fluid and uncertain division of labor between officials and elites in constructing local institutions makes any notion of an autonomous social sphere conditional, not upon rules that the state might change, but rather upon the absence of any clearly stated rules that distinguish between the realm of the state and what lies outside its purview. These chapters also track a much larger political universe. China's imperial ideology made claims on cultural practices, like the poetic canon formation discussed by Yu, not only as a means of enhancing the legitimacy of the state but also as part of the literati's own self-definition. At the same time we find, in the contributions of Bol, Elman, and Anderson, an ideology that accords the scholar-elite the resources to mount critiques of imperial actions. The possibilities of ideology in China overflow the limits of usage desired by the state. Social critiques couched in Confucian terms, like those discussed by Carlitz, can reproach the state as well; Confucian categories can also be expanded and reshaped, as Chang shows, to accommodate new cultural practices for elite women. Neither the state nor any other set of actors can gain complete and lasting instrumental control over the culture that they all seek to shape. "Orthodoxy" becomes therefore a category subject to constant revision, stretching, and twisting; to debate and disagreement; and even to disregard. Amid this flux are intellectual and aesthetic losses, such as the collapse of an unmarked space for the poetic in late imperial times discussed by Owen. The exhaustion of native categories does not mean that they can be smoothly displaced and transcended by those from the West, as Huters clearly demonstrates. Indeed, the persistence of indigenous categories, reshaped and redeployed by twentieth-century experiences, as shown by both Strand and Anagnost, reveals the extent to which post-imperial

Wong, Huters, and Yu

23

experiences belong to Chinese history. Collectively, the chapters in this volume suggest how we might rethink the constitution of relations between state and society in recent Chinese history and the particular dimensions and limitations of strategies for rule in both imperial and post-imperial times. By calling our attention to situations in which power is diffuse rather than concentrated, orthodoxy is less than unambiguous and allencompassing, and contemporary practices draw heavily on legacies of the Chinese past, they suggest the importance of conceptualizing political and cultural authority in ways that are not modeled exclusively on the modern Western experience. We hope that future research will complement the research reported in this volume and explore more closely some of the directions we are only able to suggest. "Orthodoxy" and "heterodoxy," for example, are complex and difficult concepts. A simple dichotomy between them fails to characterize the graded space that fills in a continuum between what most everyone would see as correct and what most everyone would consider wrong. There is conscious and continuous disagreement about the boundaries of what is "orthodox," which also shift over time. What state officials deem "orthodox" learning in the context of the examination system may well be different from what scholars accept as the traditions within which they work. How Buddhist practices and belief fit within convention, how Daoist rituals resist a unitary interpretation, and how the margins of Confucian orthodoxy shrink and expand in the late imperial period make problematic any simple notion of "orthodoxy." Ongoing research regarding the Manchu emperors' commitment to Tibetan Buddhism will further illuminate the multiple dimensions of the Qing state's ideological universe. These emperors did not, we think, become any less committed to Confucian precepts and strategies of rule in China proper, but they became larger and more complex personalities with identities that cannot be neatly boxed into any particular belief system. Amid the social changes of the late imperial period, we observe the plasticity of Chinese cultural categories to embrace and valorize new phenomena. This openness makes it impossible to conceive any overarching unity in the cultural system from within the Chinese intellectual universe. Neither critiques based on Confucianism nor ones built upon foreign perspectives can achieve a unitary coherence. And yet the imperial state perdured for centuries, and when it fell, the problems confronted in post-imperial times remain connected to late imperial difficulties. More generally, Chinese aspirations for all-embracing coherence in a social and political order dominated by harmony are doubly doomed to

24

INTRODUCTION

failure. First, in all settings, there are always multiple perspectives and interpretations of a given subject; uniform and consistent belief is an impossibility. Chinese culture is no exception.U Differences of interpretation can persist, as we have seen in the case of a Daoist jiao. 14 Second, a state like that of late imperial China finds it particularly difficult to cope with disagreements predicated upon alternative understandings. This was a government that claimed to rule by "educating" people to enact proper behavior, which was seen as flowing from correct beliefs. Thus, whenever beliefs that contradicted state interpretations of orthodoxy were brought to the attention of officials, the reasonable resolution involved the assertion of ideological control and coercion. This logic has applied in both late imperial and post-imperial times. Aspirations to cultural control on a Chinese scale were never an option in Europe, and indeed were unimaginable in Europe's late medieval and early modern periods. Once articulated, the contrasts seem all too obvious, but a brief review is necessary in order to consider the less obvious implications of these differences. No state in Europe could aspire to the spatial scale of control conventional in Chinese history. With respect to issues of faith and religious belief, the Catholic church represented in late medieval times the primary institution responsible for the moral wellbeing of the populace. But as successive schisms occurred in Christianity, no one view of the moral universe could dominate. Moreover, the institutional separation and functional equivalence of these religious organizations in different regions compelled an important recognition of difference. The succeeding separation of secular and sacred assigned churches the responsibility to save people's souls, while states concentrated on less lofty concerns. In the nineteenth and twentieth centuries, when European governments aimed to create "nations" of "citizens," the dimensions of belief that states aimed to shape left many other spheres of thought and cultural expression outside their concerns in circumscribed realms known under such rubrics as "private" and "religious." What belongs in each of these realms and their relationships to what is "public" have been contested in the nineteenth and twentieth centuries, but the existence of such realms has never been doubted, so strong has been their cultural construction. The Chinese state did not in fact achieve a thoroughgoing dominance over cultural expression at any point in its history. To stress its desire to assert itself over all manner of cultural matters does not mean, as the chapters in this volume attest, that there was not a considerable cultural space outside routine bureaucratic control. But the state could be called upon to intervene in extraordinary situations, whether the largely fabricated

Wong, Huters, and Yu

hysteria over queue-cutting spirits ( r 7 68) that Philip Kuhn has artfully reconstructed or the more serious issues of sectarian belief in both the White Lotus Rebellion (r796-r8o4) and the Boxer Rebellion (r9o0-I9oi). Surely there were limits to the organized bureaucratic capacity of the state to intervene in issues of belief and cultural expression, but it is too easy to measure these limitations by late twentieth-century standards. When Chinese bureaucratic capacities are compared with those of contemporary Western governments, there is no gap in the West's favor until the late nineteenth and early twentieth centuries. Arthur Wolf's contrast between modest material capacities and far greater ideological ones in China can be muted from both directions-the late imperial Chinese state had far greater bureaucratic capacity to extract revenues than did European states (other than England) before the nineteenth century; at the same time its capacity to impose order on religious practice depended greatly on elite participation and its successes never embraced all Buddhist and Daoist practices. Perhaps more basic to the Chinese situation than the extent of the state's routine material or ideological capacity was its ability to permeate society and color almost any social practice it chose to affect. Institutionalized barriers to such penetration did not exist, and no sharp boundaries were drawn between what the state and different social groups did and cared about. Instead there was a constant expansion and contraction of the pressures exerted by the government. For their parts, elites and the people more generally could accept, openly resist, or tacitly ignore state priorities and concerns. The large measure of overlap and shared commitments to a Confucian agenda of rule meant that officials and elites recognized common methods for securing social order. Bureaucratic capacities were too limited for officials themselves to implement strategies for local order without the participation of elites. When elites did not accept official priorities, they deflected official concerns or hid their own activities. They had no "right" to pursue their activities; there was no expectation, as there was in much of Europe, that contested priorities between government and elites should be resolved by explicitly setting limits to the state's actions by defining arenas of activity free of government interference. The absence of such boundaries between state and society explains how the Chinese state could simultaneously appear stronger than European states and yet remain weak. In different ways and with variable results to be sure, European societies and their states established distinct spheres of autonomous cultural expression, allowing for a plurality of beliefs and sensibilities. Although the

INTRODUCTION

significance of such difference and diversity can be questioned in Western traditions, the contrast with China remains stark. In late imperial China, there was no autonomous realm for the poetic; religious practices were never free in principle from potential state interference; political critique never marked off a vantage point from which forceful, coherent, and effective criticisms could be leveled; and officials and elites were jointly tied to the creation of community and order without clearly demarcated separate roles. Certainly there has been significant change in Chinese history, and aspects of this are captured in this volume-the changing place of poetry and the poetic, the dynamics of preference among Buddhist and Confucian cultural practices, alternative social roles for women, the shifting relations between community and center-but these changes do not fit well within Western paradigms of historical development. Future scholarship that seeks to be global or comparative, be it literary, cultural, and historical studies or research in the social sciences, must first develop strategies for analysis and interpretation that allow narratives of historical development different from Western ones an equal opportunity to explain the past and understand links to the present. By highlighting features of the relationships between culture and state in Chinese history, we can hope to ground more fully our understanding of historical change and possibility in China. The chapters in this book remind us that the late twentieth-century predicament of the Chinese state and its subjects is not simply the product of some abstract socialist law and morality, but is at least in part the descendant of practices and possibilities created in late imperial times. Whereas states in the Western tradition have, perhaps of necessity, recognized their inability to create cultural coherence rooted in a unitary vision, the imperial Chinese state was never forced to accept this situation as the norm or ideal, even when it was at least implicitly recognized in practice. The political challenges faced in post-imperial China include problems previously solved in late imperial times as well as ones introduced by China's repositioning within an international world. How these two axes intersect creates the space within which culture and state in Chinese history continue to meet. If we are to improve our grasp of the Chinese past and its unrelenting connections to the present, we must learn to map this space more fully and traverse it with greater care.

CHAPTER I

Examinations and Orthodoxies: 1070

and

1313

Compared

Peter K. Bol The men with whom the Son of Heaven shares the world under heaven all come from the literati [shi]. The dao with which the literati serve the ruler and do things for the populace all come from what they learn. Thus the ruler's selection of literati is a serious matter, and because it is serious, there are rules [fa] for it; what the literati learn is a serious matter, and because it is serious, there are also rules for it. -Xu Ji ( ro28-r ro3 ), from an examination question

J

r7' find the way many of us think about Nco-Confucianism puzzling,

even troubling. We are all too willing to dismiss as mere self-serving ideology an extraordinarily successful intellectual movement that took form during the course of four centuries of political, social, and economic change, from the mid-Tang to the Southern Song. At times Nco-Confucianism is presented as a narrowly conceived orthodoxy created and maintained by the literati elite to defend its privileged place in local society and national politics. At other times it is treated as a tool deployed by autocratic rulers for camouflaging despotism and justifying demands for conformity. I think these views are largely incorrect and am more sympathetic to those who suppose that for most serious Neo-Confucians the core of the movement was the cultivation of self-reflection and self-consciousness as the foundation for personal morality. Still, the Neo-Confucians have made it easy for us to see Nco-Confucianism as an orthodoxy promoted by the literati and the court. First, Neo-Confucians were undeniably concerned both with how people ought to think and with excluding well-established alternatives-especially but not only Buddhism and Daoism-from intellectual culture. Second, both literati and court were

30

EXAMINATIONS AND ORTHODOXIES

wont to defend the idea of "unity," whether intellectual or political, as essential to social harmony and dynastic stability. Claims to knowledge of the "one way" for all became commonplace in intellectual debate, although in fact this meant competing claims that one's views represented orthodoxy. Finally, there was an institution that appeared to realize simultaneously the interests of literati and court in establishing an orthodoxy: the examination system. 1 For many, the civil service recruitment examination, as the pre-eminent cultural institution tying local society to capital politics, was a likely vehicle for attempts to disseminate values and inculcate "orthodox" attitudes. The literati were directly affected by what the examinations tested. The court could envision using the examination system to select literati favorable to its leadership; literati reformers could propose revising the system to secure men committed to their own values. In fact, the two most important "reforms" of the examination system, out of which came the system in place during the Ming and Qing dynasties, were intended to transform literati values. The first took place in 1070 and was the most radical change in the contents of the examination system since the eighth century. Under Wang Anshi's (ro2I-86) leadership, the court created a new curriculum, instituted a national school system, and made school attendance a prerequisite for sitting for the examinations. The second was the Yuan government's restoration of the examination system in I 3 r 3, which simultaneously mandated a Nco-Confucian or Daoxue curriculum centered on the learning of Zhu Xi (II30-I2oo). The Ming adopted the Yuan system and the Qing the Ming system (see Chapter 2, by Benjamin Elman, for the significance of the Ming-Qing system). The history of the examinations is a useful vehicle for looking more closely at the interplay between the state and culture, or between those leading the institutions of government at a given moment and those contributing to intellectual culture. And, because it is so often taken as a sign of the stagnation of literati culture and the politically expedient imposition of an orthodoxy on the literati, the examination system is a promising terrain for making an argument about the nature of the intellectual transformation represented by Nco-Confucianism. I hope the argument I develop here will have some bearing on how we think about the intellectual culture of the later imperial literati-those who called themselves shi or shidafu-and that it will give us reason to doubt the ability of either the court or literati reformers to impose intellectual orthodoxies. Perhaps, too, it will contribute in a small way to the large and difficult question of how we should account for the spread of Nco-Confucianism.

Peter K. Bol

I make my arguments with reference to the following problem: Why did leading intellectuals oppose the imposition of Wang's new examination system in the royos but not the imposition of the Neo-Confucian curriculum in the 13 ros? Wang's curriculum did not survive-two of the three commentaries on the Classics prepared at his behest as well as Wang's glossary of morals are lost-even though regimes favoring Wang Anshi's New Policies controlled the court for the last fifty years of Northern Song (960-1 126). In contrast, the Neo-Confucian curriculum has survived intact. I believe that an answer to this question will tell us something about the values that defined literati culture in the later imperial period, a matter that is of considerable importance for how we think about the cultural power of the later imperial state in which literati were the dominant social stratum. There is one obvious explanation for the opposition to Wang's New Learning and the acceptance of Zhu Xi's Dao Learning: in the royos literati resisted the imposition of a curriculum that served the partisan political ends of one literati faction, but in the I 3 ros the government was simply acknowledging the pre-eminent literati intellectual movement of the day, a decision the literati applauded. 2 There is considerable truth to this and to the conclusion we might draw from it: that the emergence of such a powerful intellectual movement-one the government found it politically important to acknowledge as orthodox-outside the examinations and state sponsorship shows that although it was possible to change the exams to fit court or popular opinion, the institution did not control, or allow the state to control, the development of intellectual culture. However, this answer begs the question of why literati accepted NeoConfucian intellectual claims in the first place. I cannot answer this larger question adequately. I shall approach it by asking a more limited question, but one that allows for a comparison with the royos: Why did the kind of literati who defended eclecticism and pluralism in intellectual culture-who opposed Wang Anshi in the royos on these grounds and were regarded by the most staunch NeoConfucians with grave suspicion precisely because they defended intellectual diversity-accept the imposition of the Neo-Confucian curriculum in the 13 10s? This approach allows me to focus on those who saw themselves as pragmatic literati, usually men associated with the "ancient style" in literature, who viewed intellectual values as mediated by history and culture. In the royos among the critics of Wang Anshi, Su Shi (103 yI 101) and his circle best represent this view. In the fourteenth century we can see a similar group around Huang Jin (1277-1357) and Song Lian

EXAMINATIONS AND ORTHODOXIES

(r3 ro-8 r) in Jinhua prefecture in Zhejiang. These two groups shared the idea that literature in the traditional sense of wen both mediated thinking about values and provided one of the primary vehicles for participation in intellectual life. Wen, in fact, is where this paper must begin, for at the heart of ideas about examination reform were ideas about literature and culture.

The Reform of the IO?OS"Making Values the Same and Customs Uniform" Wang Anshi's examination reforms replaced the traditional "literary" part of the examinations with essays on the Classics. This needs to be understood in the context of a debate over the use of literary composition as a test of talent (both intellectual and moral) that went back several centuries. These debates were complicated by the spread of Han Yu's (768-824) "ancient style" (guwen) movement in the tenth and eleventh centuries, which criticized traditional literary practice while placing its hopes for moral renewal and political reform in the spread of a new but "ancient" style of literary learning. Because it was one aspect of the larger category of wen (the culture of textual traditions, cultural forms, and moral governance that made China civilized), what we can rightly call "the literary" (in a minimal sense because it was concerned with how men wrote) contained a fundamental tension. On the one hand, its larger justification lay in the fact that it was part of wen, the repository of moral values and public models, and thus was supposed to further political and social order. This in turn justified treating men good at wen as leaders of learning and culture. In the early Tang, "Confucian teaching" (rujiao) and "wen teaching" (wenjiao) were synonymous, for the Classics and civilizing institutions were the original human wen, the product of sage-kings who translated the orderly patterns of the heaven-and-earth into human guides. On the other hand, the individual could use literary craft both to express personal, often disturbing, emotional responses to things and to impress others with his cleverness and skill. Literary composition had long since become a demanding task, requiring hard-won mastery of generic styles and intricate literary forms. Some thought the individuality and creativity encouraged by literary craft undermined social order. Others held that the more dangerous problem was the inherent autonomy of literary art-a consequence of the fact that one could objectively judge the quality of a composition by craft standards-which tended to separate the literary enterprise from social and political concerns. In the eyes of its critics, the literary

Peter K. Bol

33

tended to the superficial and empty. Perhaps this would not have mattered, but literary composition was wen. It was tied by traditions of writing and content to antiquity; its practitioners represented refined culture within the political elite. It was supposed to function responsibly.

Background-The Problem of Wen as Literary Artifice The Tang examinations were of two sorts. By the middle of the eighth century, the jinshi degree had become a "literary" test, requiring the composition of a regulated verse, a rhapsody, an essay, and three treatises on affairs of government. In addition to the jinshi, the "various fields" (zhuke) examinations tested rote memorization of Classical and ritual texts and commentaries. Fewer men entered the regular Tang bureaucracy through examinations than through either the yin (which allowed high officials to secure official appointments for sons) privilege or promotion from senior clerkships, but there were many candidates. In 7 3 7 there were 6o,ooo registered students. 3 It seems likely, given the difficulty of success in the examinations, that many of these 6o,ooo students made a career of the student life until driven to pursue less prestigious choices. Perhaps it was the popularity of the examinations that prompted those who attributed the dynasty's inability to prevent the An Lushan rebellion to a rhoral failing to attack the examination system in the 76os. These moralist critics restated an older objection to attaching greater value to literary learning and skill in composition (wenxue) than to ethical behavior (dexing). 4 An example is Yang Wan's critique of 763 that defined wen (the literary) along three polarities. The first was between behavior and speech: doing something is always preferable to talking about it, and behavior is a more reliable indicator of future action than is speech. Those who spent time making language wen, that is, elaborating and refining it, made it even more distant from action. The second polarity was between body and surface or substance and outward adornment. Wen as surface adornment could appeal with its beauty, but the surface was an "empty" disguise that easily became detached from the body; it tended to "float and flower" (fuhua). Finally, wen represented a cumulative tradition of occasional literary genres whose artfulness and occasional nature stood in contrast to the enduring models and consequential affairs in the Histories and Classics. Yang's objections are obvious: the jinshi examination encouraged literati to neglect proper behavior and the emulation of canonical models in the pursuit of empty adornment. Moreover, since candidates could nominate themselves, without being recommended by their betters, the system encouraged self-seeking. It is hard to imagine a much bleaker

34

EXAMINATIONS AND ORTHODOXIES

assessment. Yang's solution was a return to the Han dynasty model of recruitment through the local recommendation of men known for virtuous behavior. He supplemented this with two further tests: candidates would be asked ten questions on a single Classic (and its commentaries) of their choice and would write three treatises in response to questions on models of governance and current problems. 5 In service of his polemic, Yang Wan reduced the term wen to literary artifice. In response, Jia Zhi (71 8-72) reminded the court that wen was the quality that typified the ancient Zhou dynasty and the Classics were the textual legacy of civilization. By extension, it included later literary traditions that not only drew on the models of antiquity but also manifested the quality and aspirations of the individual, an individual who if properly taught would use his writing to establish models of civilized behavior. Restoring order in the aftermath of the rebellion required "civil" rather than military men, who would both extend the transformative influence of the court and alert it to its failings through their writing. The problem was not with wen, Jia concluded, but with the reliance on technical literary criteria to grade candidates; better to increase support for schools and choose examiners who understood the mission of wen. 6 Tang scholars, both critics and defenders of the literary examinations, shared the assumption that what one studied-in the sense of mastering a form or technique-affected character. The texts of the past, whether from among the Classics, Histories, philosophers, or literary collections, supplied normative models and qualities the reader would absorb. Thus one could object that those who spent their time learning to compose literary works were developing superficial qualities, because literary works were by nature concerned with surface and appearance, and at the same time call for the creation of new examination fields testing the memorization of various "substantial" texts (the three ritual canons, the three early Histories, etc.) to recruit men of corresponding substance. 7 Critics worried about a court led by men (chief ministers were expected to have degrees at this time) imbued with the superficiality and cleverness they associated with artful composition. These objections reappeared in the new social context of the Song dynasty. The Song court began expanding the examination system into the primary means of recruiting officials in the 970s. As a consequence, by the middle of the eleventh century, the political elite consisted of men who had demonstrated their ability to either memorize chunks of the textual tradition or, increasingly, to compose in good style. 8 The ever-growing pool of examination candidates increasingly made up the new political

Peter K. Bol

35

and cultural elite and defined them, in the words of an edict from 98 5, as men who as students cultivated ethical conduct and "made wenzhang [literary composition] their occupation. " 9 As scholars were quick to see, debates over what Song literati should study-and how they should study-had implications for practically the entire political elite. It is common to describe the change in the intellectual culture of the political elite from Tang to Song as a shift from Buddhism to Confucianism. However, for our purposes it is more useful to recognize that the leaders of Tang intellectual culture were literary figures, whereas in the Song they were becoming "thinkers." In the Tang scholars argued over where one should draw one's models from; in the Song they debated ideas inferred from the models they preferred. But what for the sake of contrast I am characterizing as "Song" went back to the great Tang literary intellectual Han Yu. Han combined seeking the "dao of the sages" (shengren zhi dao) through learning and writing in an "ancient style" into a single mission; I shall refer to this with the term Han invented for it: guwen or "ancient style." In Han Yu's time, it was still hard to draw clear lines between the literary and the intellectual, but even after the rise of NeoConfucianism made a sharper distinction possible, the line between literary and intellectual history was not always clear. One twelfth-century literatus commented, for example, that "in the Tang, wenzhang underwent three transformations [culminating in Han Yu]. Our dynasty's wenzhang has also seen three transformations: Wang Anshi with Classical scholarship, SuShi with opinion pieces, and Mr. Cheng [Yi; r033-II07] with innate morality. " 10 The Song figures all shared Han Yu's aim of turning literati away from the imitation of models and past forms and toward the search for those values (dao) that guided the sages in their composition of civilization as the set of political, social, economic, and cultural institutions and models that made it possible for the common good to be realized. For Han, the place to seek these values was in the Classics and early Confucian texts, since these were the historical remains of the sages and their followers, rather than in the workings of the cosmos or in regimes of spiritual cultivation. Han also insisted that literati had to understand the values of the sages for themselves, that they had to think on their own and not simply repeat what others said, and that turning back to the sages entailed being at odds with the times. For the idealistic literatus who found himself at odds with the age, turning to the textual remains of antiquity for the values that had guided the sages in creating civilization was a search for principles to follow and

EXAMINATIONS AND ORTHODOXIES

purposes to accomplish that had the highest order of cultural justification. It followed that publicizing his search through his writing and encouraging others to transform themselves was also a way of acting responsibly toward the present. Imitating the prevailing style could not produce good wen, good men, or good government. But resurrecting ancient forms would not do so either, for true continuity with antiquity was a matter of realizing the ideas and intentions of the sages, not imitating them. It seems to me that Han Yu and his Song admirers created a fundamental dilemma for themselves. We might think of this dilemma as an intellectual version of the literary problem of having to produce artful compositions while reaffirming the authority of classical models. For Han, the scholar was supposed to understand the dao of the sages in his own mind and produce works that had enduring value as guides to the common good. As he explained: "One reads books to learn and strings together words to write, not to boast of how much [he has read] or compete over embellishment [in writing]. For learning is about true values, and writing is about establishing norms [xue yi wei dao, wen yi wei li]." 11 This was a dilemma because the goals were in tension. The scholar who sought the values of the sages with his own mind risked becoming idiosyncratic, of having ideas that could not be shared. In a famous incident in the 1030S the guwen leader Ouyang Xiu (1007-72) attacked ShiJie (roo5-45) for being so anxious to abandon the conventions of calligraphy and to create something "ancient" that his handwriting had become illegible. On the other hand, those who translated the ideas of the sages into guides for their own time risked creating models for others to imitate, when a real social transformation required that all literati understand the sages for themselves. Han resolved this tension by focusing the inquiry into the dao on the Classics and early Confucian texts (especially the Analects and the Mencius) and by making an "ancient" literary style the public expression of one's search for values. The two were not entirely separable, for the process of seeking the dao was also a matter of textual immersion, and one could judge the degree of one's transformation by the changes in one's writing style. Both were necessary, as Zu Wuze ( roo6-8 5) explained in about 1040, "What is accumulated inside is called dao. What is expressed outside is called wen. When there is dao and there is wen, then one can be a superior man. " 12 To see why literati could suppose that the cultivation of a personal "ancient style" was a means of being socially responsible, we have to keep in mind that the literature of the period was largely occasional, that is, written in response to partings from friends, travels, affairs of office, death, and so on. It is also clear that Han Yu did not think

Peter K. Bol

37

that institutional reform and policy-in contrast to the contemporary institutional historian Du You (73 5-812), Han thought the structure of government was adequate-was as important as the conduct of the people in government. From this perspective, the work of government was also "occasional." Policy was formulated in response to situations just as poems were. Learning about the dao of the sages was, Han saw, a matter of understanding the values that had guided the sages who had created the institutions, and thus would ensure that government once again would realize its original purpose of serving the common welfare. 13 The man who could bring the values of the sages to bear on the affairs of daily life in writing (an individual's literary style was in this sense his template for looking at the world and responding to it) accomplished two things at once. He showed others how a person with his dao responded to these events, and he showed how he would respond to the affairs of state were he given political responsibility. Han himself insisted that any attempt to define the dao of the sages in doctrinal terms was self-defeating, but clearly from the start others recognized that possibility. There is little doubt that guwen resulted in a new interest in the exposition of ideas in prose. 14 From the start, Song attempts at examination reform confronted the guwen dilemma. Zhang Zhibo's (?-1028) memorial in 1002 warned that a text might be "ancient" but still not worthy of transmission or emulation, for ancient texts with ideas contrary to the "great dao of the sages" harmed political unity. Zhang proposed to limit the themes assigned for the regulated verse, rhapsody, and essay to the Classics and Histories and those philosophers who agreed with the Classics. At the same time, he proposed making the first cut on the basis of the treatises written in response to questions about governmentY Zhang's views were echoed by other proponents of a narrower curriculum, but he also appealed to those who asserted that poetry was irrelevant to government and that valuing skill in composition led to "superficial ornamentation" and believed that the essay and treatises should weigh more heavily in awarding degrees. 16 The effort to focus the examination curriculum on ideas reappeared in 1044 when guwen-inspired reformers briefly held sway at court. Led by Fan Zhongyan (989-1052), who believed that moral values should be applied to politics, they proposed a more rigorous screening of the ethical quality of the candidates {thus to block those infected by literary superficiality), and to ensure that local officials would be able to assess personal quality, they proposed the creation of prefectural-level schools throughout the empire, tied admission to the qualifying examinations to school attendance, and legislated the appointment of official teachers. To favor literati

EXAMINATIONS AND ORTHODOXIES

who had ideas about government, they proposed to restructure the sequence of tests so that composition of the policy treatise and essay preceded the verse and rhapsody. Finally, they assured those who saw literary composition as evidence of individual talent that candidates would be given greater opportunity to display their abilities and relaxed the rules of composition for the verse and rhapsodyY Candidates for the various other examinations-traditionally memorization fields-were to be encouraged to think as well. They were free to limit themselves to the memorization of text and commentary, but they could choose instead to write essays on the "greater significance" (dayi) of passages from the Classics. What Fan Zhongyan's group did not do was to legislate a curriculum. The reforms were abandoned within the year, but the idea that literati ought to learn the dao of the sages and bring it to bear on government continued to spread, and the guwen style became commonplace. What was different, I think, was the growing split between those who wanted a public definition of the intellectual and political ideals that ought to constitute the dao of the sages and those willing to tolerate individual interpretations of that dao. The difference is evident in government, between those who saw a need for developing programs of institutional reform and those who thought government would be fine if the right men were in charge. It also appeared in attitudes toward writing and culture, as some began to contend that the literary was merely a vehicle for transmitting ideas in a fancy form and that language ought to be a transparent medium of communication. "Wen is that by which one conveys the Dao" was how the philosopher Zhou Dunyi (1017-73) put it. 18 His contemporary Chen Xiang (10I7-80) extended the point in asserting that even the subjects used in writing are mere devices for conveying moral ideas: "Wen is the boat for preserving dao, the presence of affairs in wen is like a boat conveying things; it must function to get [things] across." 19 There was a much more complex position that recognized the problematic relationship between cultural forms and values and that, as a consequence, justified the existence of creative, individual writing and inquiry. It had its origins in Han Yu, whose son-in-law introduced Han's writings with the assertion that "wen is a device for threading together the Way [guan dao ], " and in Liu Zongyuan's (773-819) understanding of his own claim that "wen is to illuminate the Way [ming dao]." Both Ouyang Xiu and SuShi, I think, were among its eleventh-century champions, and both held that the literary men and literary works that accomplished this had real value. 20 These men insisted that literary writing was necessary precisely because values could not be defined absolutely. What literati could share were ideas of

Peter K. Bol

39

greater and lesser persuasive power expressed by men on the basis of the subjects of their studies in a manner accessible to their times. In this view language, writing, and the individual writer mediated-sometimes obscur~ ing and sometimes clarifying-both the search for values in the past and their expression in the present. Culture could never be made totally transparent; it could not be merely a vehicle. These two different views of culture, the one seeing culture merely as a vehicle for right ideas and the other seeing culture as integral to the process of generating ideas, suggest different approaches to evaluating literati writing. Those who subscribed to the second view, such as Su Shi, tended to tell literati that they should cultivate the internal substance and the external wen would take care of itself. They tended to stress the integrity of the literary composition as evidence of internal integration. This could be taken further. Some contended that because internal integration was the key, any concern with external literary manifestations was misguidedY But those who saw culture as a mere vehicle tended to ask whether the content of the literary work agreed with right principles. Here, too, many held that being too interested in literary form was a sign that one was not interested in communicating right ideas. Thus we find some defending Han Yu, Ouyang Xiu, and others on the grounds that "their words agree with li [principle), their purpose with dao," and that what they write is useful to the age. In this case the writer meekly argued against rejecting literary composition entirely on the grounds that it "some~ times" can agree with dao. 22 If the goal of writing was merely to convey content that could be judged by non-literary standards, then, as some pointed out, there was no necessary connection between the literary style of the work and the moral quality of the author. 23 Guwen intellectuals had insisted that the literary enterprise had value because it was integral to the process of seeking and expressing larger values. But for those who saw it as a vehicle of communication, wen had, at best, value as a tool. Writing well was a craft that required intelligence and discipline, but it was not inherently connected to matters of greater significance. 24 However, some still insisted that those able to "make language wen" played a role in the definition of social and political mores even when their ideas departed from those of the sage. 25 There even were those who still insisted that the literary should continue to be a primary way for literati to define social values. 26 This was the context in which Wang Anshi replaced the literary compositions in the jinshi examination with essays interpreting the Classics and in which, for seven decades thereafter, literati debated the idea of unifying values from above.

EXAMINATIONS AND ORTHODOXIES

The Examination Reform of ID7DTeaching Literati How to Learn The new examination system announced in 1070 abolished the so-called various fields degrees hitherto granted for the memorization of Classical, ritual, and historical texts and commentaries. It maintained the jinshi degree but abolished the traditional regulated verse and rhapsody sections that made the jinshi a literary degree. Instead candidates chose one Classic for specialization-the Odes, Documents, Change, Rites of Zhou, or Record of Rites-on which they wrote five essays on the significance of assigned passages. This was followed in the second session by a set of ten questions on the "larger significance" of ten passages from the Analects and the Mencius. In the final sessions they continued earlier practice and wrote an essay and a treatise. The jinshi had become a test of the candidates' ideas written in fairly straightforward proseY The new program was a rejection of the "literary" as a criterion of selection in favor of "Classical studies" (jingshu). The examination reform was part of a larger program to unify literati values by establishing a national school system with a single curriculum. The SecretariatChancellery's minute stated the rationale: It was because antiquity's selection of literati was always based on schools that when morality was unified above customs were perfected below and all human talent was capable of actively accomplishing something in the age. Once the beneficence of the Former Kings was exhausted, the model of education and nurturing lost its foundation. Although a literatus might have excellent talent, he had not the schools, teachers, and friends for perfecting it. 2 ~

The later New Policies regimes, particularly during Emperor Huizong's reign ( r roo-r 12 5 ), expanded the program and made further institutional arrangements to ensure that students seeking to enter office through the examination system would have to pass through a graded school system. 29 By no8 the schools had 167,622 students, in 95,298 buildings, were expending over 2.5 million strings of cash and 30o,ooo piculs of rice, and were drawing income from over 105,990 qing (ca. 1.5 million acres) of endowed land. 30 Curricular development was the intellectual heart of the new program. Wang Anshi and his son took charge of the composition of the required commentaries on the Odes, Documents, and Rites of Zhou, appropriately titled the "new principles" (xinyi). The court later added Wang's Explanations of Characters (Zi shuo ), a dictionary and glossary of morals that discovered norms for things and affairs through an analysis of the struc-

Peter K. Bol

ture of the written characters used to name them. 31 Not until the r r2os were opponents able to force the abolition of the new examinations, curriculum, and centralized school system, leaving for Southern Song a much-expanded school system without an intellectual project. The New Policies intellectual program was part of a larger agenda for the transformation of the government's relation to the economy, society, and bureaucracy. No longer would it simply respond to events and cope with existing interests. Instead, it would create institutions that would ensure that private interests would serve the common welfare by making farmers, merchants, and literati directly responsive to the government's policy initiatives. This "activist" (youwei) agenda extended to the recruitment of talented men for government service. In Wang Anshi's view, the government had not solved the many problems of the day because there simply was not enough "talent" available in society. As he had announced in a long memorial to the throne well before gaining power, it was up to the government to create the talent it needed by "molding and casting it. " 32 For Wang a true school system would create the kind of talent necessary to serve the common good and make literati submit to it; there was no need for government to make itself appealing to the literati. Local schools would become both the source of officials and centers for the transformation of local society. 33 Although Wang's new order represented one sector of literati opinion, he intended to establish a single, unified system for all literati. And in his effort to "cast and mold" talent, he went beyond traditional efforts to legislate what literati would read and how they would write to address the question of how they should learn and think. The New Policies educational program was an attempt to impose an orthodoxy on literati from above, but it was justified from Wang Anshi's point of view because it was a total, integrated system inspired by the integrated social order Wang saw in the civilization created by the ancient sages. Literati who studied Wang's books could see this for themselves. Wang's enemies saw a man bent upon imposing his dogma on others. His admirers, however, were awed by his completeness. Shen Liao (103 285), for example, applauded the "coming of the New Learning" because he was convinced that Wang's Classical studies coherently integrated (yi guan) the essentials of human nature and the foundations of ethics while maintaining creative insight, the intent of literary writing, and the proper form for Classical scholarship. 34 The complexity of Wang's intellectual position stems, I think, from his desire to have a curriculum that would ensure that individuals could learn for themselves yet reach shared conclusions. In his youth he had aspired to sagehood and supposed that since

EXAMINATIONS AND ORTHODOXIES

fundamentally all the sages were alike, those who studied them would all turn out the same as well. 35 Moreover, assuming an inherent, integrated order to the world and the sages' perfect mental apprehension of that order allowed Wang to assume that the Classics as a product of the sages must reflect that order if properly understood. 36 As his cousin and intellectual companion Zeng Gong (I o I 9-8 3) explained, in the ideal world of antiquity, "morality was the same for all, and customs were uniform"; those who "spoke about inherent/normative patterns [li]" always reached the same conclusions. This was "because when li fits there is no legitimate alternative" (li dang gu wu er). Thus the Classics are coherent, like "the theorizing of a single person"; the intellectual diversity of later periods resulted from "not fitting the li," from having values that "did not fit the intentions of the sages." 37 Yet Wang, who first made a name as a guwen essayist, was alert to the problem of imitation: he called on literati to grasp the intentions of the ancients for themselves and thought the Zhou dynasty began to decline when men started imitating forms and forgot how to act creatively as sages. 38 Wang's approach to the Classics, to writing as the foundation of the civilization created by the Former Kings, and ultimately to the phenomenal world of his own times was to establish meanings by supposing that there was a single, inclusive system constituted by all the parts and pieces before him. To define the value of any particular piece, Wang asked what its function was in the integrated system of which it was a part. Supposing that there was a natural coherence in the arrangement of hexagram lines, poems, character strokes, institutional systems, and so on allowed Wang to determine the functional value of each part by explaining why it should appear where it did in the system he was looking at. For Wang, knowing how a part could function in the whole was knowing how it ought to be. "Every one of the ten-thousand things has a perfect pattern [li] to it," he wrote in an essay, and "the way to attain a refined grasp of its pattern lies simply in attaining its unity. " 39 The Classics thus reflected an integrated system that inhered in the natural world. And thus, Wang noted, one could make up for any deficiency in the Classics by taking phenomena into account. "For long the world has not seen the complete Classics. If one were only to read the Classics, it would not be enough to know the Classics. I thus read everything ... and I inquire of everyone, down to the farmer and the craftswoman. Only then am I able to know the great system [dati] of the Classics and be free of doubt. The later ages in which we learn are different from the time of the Former Kings. We must do this if we are fully to know the sages. " 40

Peter K. Bol

43

Were it not for the fact that Wang had the power to translate his vision into policy, his resemblance to other thinkers of the day would be more apparent. First, he supposed the existence of an integrated order in both nature and antiquity. Second, he made coherence the test of speculation. Third, he was inclusive of all phenomena. Fourth, he acknowledged the fact of historical change and the relativism that follows from that. Fifth, he held that each man can enact this approach to determining values for himself. But what each literatus was supposed to do for himself Wang had done better than almost anyone else, and he became increasingly convinced that, like Confucius, he had a Heaven-protected mission to ensure the transmission of his understanding of the true meaning of ancient culture. Writing toward the end of his life, he introduced his glossary of morals, the Explanations of Characters: "Heaven, not being about to allow This Culture of Ours [siwen] to perish, has used me to illuminate the beginning [of This Culture]. Therefore, teaching and learning must start from this [book]. Those who are able to understand it will have gotten nine-tenths of the ideas of morality. " 41 Wang Anshi's Critics

Among Wang's critics were groups around the literary intellectual SuShi, the moral philosopher Cheng Yi, and the historian and statesman Sima Guang (ror9-86). Although they dominated the court only briefly, between roSs and 1093, New Policies regimes hounded them and their followers for decades. There were internal divisions among the critics, but they generally agreed on two objections to the new curriculum. First, Wang was "making others be the same as himself" by demanding adherence to his teachings. 42 Su Shi spoke for many when he objected that the new curriculum encouraged imitation: When literati are not able to complete themselves, the problem lies with customary learning. The problem with customary learning is that it corrupts men's capabilities and blocks men's ears and eyes. They recite what their teacher said about the construction of characters and follow the customary literary style. After only several ten-thousand words of this is the task of being a literatus finished .... The learning of Mr. Wang is exactly like striking prints: [the students] come out according to the block, and there is no need to decorate them before they are usable. How could they ever be made into rare and precious implements? 43

The second criticism was that Wang's program encouraged students to go outside texts and history in search of some ultimate, mysterious key to understanding. The director of the departmental examination of ro88,

44

EXAMINATIONS AND ORTHODOXIES

Sima Guang's fellow historian Fan Zuyu, contended that literati under the influence of the New Policies had increasingly turned to Buddhist and Daoist ideas about human nature and the cosmos. 44 Somehow Mencian zide, or "getting or apprehending it for/in oneself,"45 and dogmatism coexisted as literati took the need for external public norms and internal individual thought to their extremes. The quandary of Wang and his followers is captured in a letter Zheng Xia (104 r90) wrote to Wang as chief councillor. Some literati adhere to Wang by imitating his writing and conduct, Zheng noted, and others do so by sharing Wang's universal inclusiveness ("Comprehending heaven above, reaching the earth below, and at the sides extending without limit in all directions"). The problem is that the spiritual resource that enables Wang to see the whole is beyond facile imitation. Just as with Confucius, "Your wenzhang can be heard [and imitated], but your discussion of human nature and Heaven's dao cannot be." But, Zheng continued, slyly introducing his critique, because Wang had wisely appointed men who shared his creative quality rather than those who could merely imitate him, those he has appointed are unable to avoid differences and mistakes. And-to get to the point-if it is possible for his assistants to make mistakes, then Wang needs to listen to criticism. 46 Wang's successors combined dogmatism with a claim to be in accord with the ultimate mysterious source of creativity. They ordered students to study Wang's texts and the cosmic dao, and they proscribed the dissemination of the writings and teachings of Wang's foremost critics, among them Sima Guang, Su Shi, and the Cheng brothers. 47 Wang's partisans held that his dao and xue (learning) continued the "dao of Confucius and Mencius" and provided the model for all scholars. 48 But others, such as the renegade Chen Guan (ros7-ri22), complained that the court had made Wang's teachings into the "national correctness" (guo shi), and was using appeals to "the innate principles of human nature" (xingming zhi li) to demand conformity in morality and customs. 49 One consequence of the imposition of a unified curriculum was that some began to defend the idea of diversity in a manner akin to a defense of intellectual pluralism. SuShi was the leading spokesman for those willing to defend the idea of multiple ways of being literati and multiple points of view. From his perspective, Wang marked the decline of literature, for while Wang's wen was good, Su admitted, "the problem is that he likes to make others the same as himself." Su's solution to the problem of how to allow individuality without giving up on common values was to locate commonality in the creative process, not in the results: "The goodness of

Peter K. Bol

45

the earth is uniform in bringing things into being; it is not uniform in what it brings into being. It is only on barren, brackish soil that there are yellow reeds and white rushes as far as the eye can see. This is the uniformity of Mr. Wang." 50 Su's allies echoed his complaint within the examination system by asking questions that encouraged candidates to see the existence of contending schools of thought as a beneficial development rather than as a sign of decline. 51 The Southern Song court ceased trying to force through the New Policies educational program and eventually decided to offer literati a choice between a Classics jinshi without orthodox commentaries and a revived literary jinshi. The decision not to approve any particular commentaries but to allow students to choose their own, and even add their own opinions, also served to announce that the court would not favor any single school of thought. This was a response to the ideological battles of early Southern Song, when advocates of Cheng Yi's teaching sought to have Wang Anshi's learning officially condemned and Cheng's teachings officially recognized. 52 Without trying to trace the history of intellectual culture from the twelfth century on, I shall conclude this section and prepare the grounds for a discussion of I 3 I 3 with a reflection on the difference between Su Shi's and Cheng Yi's circles as opponents of Wang's orthodoxy. For Su and his circle, human understanding of dao and /i would always be mediated by one's education, experience, and personal character, and a person's expression of what he had understood would be mediated by his facility in language and wen. "Learning" as the literati enterprise that led to conclusions about how one ought to act could never be perfect, and one's grasp of the universal would in practice be particular. Thus, it would always remain the task of individuals to make persuasive arguments for their own ideas about the necessary patterns of things and events. Su at least was willing to redefine writing that was "wen" along these lines, as writing that succeeded at the difficult task of accurately communicating an understanding of the intentions and patterns of things. 53 In the Southern Song the most successful alternative to the New Learning was Daoxue Nco-Confucianism. For Cheng Yi and his disciples, learning was above all a process of self-cultivation, leading to the recovery of an innate moral nature obscured from consciousness by selfishness. In theory the moral principle of any thing or affair was perfectly accessible to the fully cultivated mind. However, the good student gave public expression to what he had learned through ethical conduct (dexing), not literary composition. For Daoxue followers, literary writing and cultural forms

EXAMINATIONS AND ORTHODOXIES

could never be more than a vehicle for conveying values that existed independently of culture. Moreover, literary skill could never serve as a test of human worth because, in contrast to the ability to know right and wrong and behave ethically that all human beings possessed, literary skill was a talent given only to a few. In mid-twelfth century, Zhu Xi and Li.i Zuqian ( II3 7-8 I) created a Daoxue curriculum that recognized the utility of the broader cultural tradition without undermining Cheng Yi's view that all humans possessed ultimate values, which could only be actualized through individual effort. Daoxue in the Examination System of 1313 The reinstitution of the examination in I3 I 3 put into effect a policy that had been drafted in the I27os but abandoned. The new examinations favored the Classics over literature. As the court announced, "Classics learning is in fact the dao of cultivating the self and governing others. Literary composition is the learning of stealing stanzas and decorating phrases. Since the Sui and Tang, the selection of men [for office] has singly valued literary composition, therefore literati practice has been frivolous [fuhua]." 54 This represented the triumph of Daoxue over the legacy of the ]in dynasty, whose intellectuals had preferred Ouyang Xiu and SuShi to Cheng Yi and Zhu Xi. The examination distinguished among ethnic groups, but all were to master Cheng-Zhu thought. Mongolians and Inner Asians (semu) were to study the Four Books with Zhu Xi's commentaries. The literati from north and south faced a more elaborate test. The first session consisted of one question on an apparent contradiction in one of the Four Books and another on the meaning of a passage from one of the Five Classics. Students followed Zhu Xi's commentaries in their essay on the Four Books but were to "conclude it with their personal opinion." Each candidate specialized in one of the Five Classics and its commentaries (with the exception of the Book of Rites these were Cheng-Zhu school commentaries). Literary composition was tested in the second session through the composition of a rhapsody and decree in ancient style and memorial forms in modified parallel prose. The third session called for a treatise written "straightforwardly, without embellishment," on current policy or an issue from the Classics or Histories. 55 With minor adjustments, the Ming dynasty adopted this system and gave further support to NeoConfucianism with the Great Compendium of the Five Classics and Four Books, a set of subcommentaries for the Neo-Confucian interpretations based on Yuan works. This ended official sanction for the Han-Tang

Peter K. Bot

47

exegeses collected in the Tang dynasty Correct Significances of the Five Classics. Similarly the Great Compendium of the Principles of Human Nature, organized according to the Neo-Confucian philosophical system and citing some I 20 thinkers, defined Song thought as the basis for literati thinking about values. The restoration of examinations was good for literati morale-! would read the claim that the exams "changed customs" as meaning that they gave literati a reason to continue to act as literati 56-but the exams did not grow into a major source of officials. Nevertheless, as Xiao Qiqing has demonstrated, the Yuan government generally was not hostile to the literati, some of whom they protected with the "Confucian household" classification and nurtured with the most extensive school system to date in Chinese history. Literati were generally prepared to serve the Yuan dynasty. 57 The examinations recognized the literati as a group defined by its learning, to be distinguished from the clerks, those literate and politically experienced subjects whom the Mongols willingly appointed to responsible positions in the civil administration. 58 Yet there were literati who became clerks and clerks who became officials and transformed themselves into literati. Some advocates of Daoxue warned that its institutionalization would not necessarily advance its goal of moral transformation. 59 As a proponent of the teachings of Lu Jiuyuan (I I39-93) objected, the examinations tested writing when the real issue was the mind-and-heart, a sentiment the great Neo-Confucian teacher Wu Cheng (1249-I33 3) shared. 60 But by and large Neo-Confucians believed they had brought about a great historical change. Cheng Duanli (1271-1345) made this clear in an inscription for an academy founded in the Southern Song for holding "discourses on learning" (jiangxue) in opposition to examination learning. Now in the Yuan, Cheng explained, this distinction between true learning and examination learning is no longer valid. Since Xu Heng [I 209-8 r] aided the age with the learning of Master Zhu ... all the scholars under heaven have known they should honor the Classics upon which Master Zhu commented in order to find their way back to Confucius and Mencius. Xu's merit is great. Our examination system also follows Master Zhu's "Private Proposal [on the Examination System]." The Classics section takes the interpretations of Cheng and Zhu as principal but uses the old commentaries and subcommentaries at the same time, while the essays on the meaning of the Classics do not have to conform to a set form of composition. The essays [these students] write on the Classics are able to compare the Cheng-Zhu interpretation with the old commentaries and subcommentaries item by item and define the failings and achievements

EXAMINATIONS AND ORTHODOXIES

of the Han scholars' exegeses. We have completely washed away the faults of the late Song, their meaningless emptiness, literary excess, and despoiling of the Classics. We have united Classical studies, the learning of principle, and examination training, thus to make it easier for literati devoted to the Dao. How could the examinations of Han, Tang, or Song compare! 61 It is the claim that the core of the examination curriculum-recognized as the particular views of certain thinkers-should have intellectual authority for all literati that makes the Nco-Confucian examinations comparable to the New Policies curriculum of the Northern Song. Wu Shidao, who took a degree in I 3 2 I, used language reminiscent of Wang Anshi when he said that the decision to recognize the "learning of moral principles" (yili zhi xue) and the Song masters as authoritative was intended to "unify morality and make customs the same. " 62

Daoxue as Its Advocates Saw It To address the question of why literati did not protest the imposition of the Daoxue curriculum we need, first, to see what its most devoted advocates thought they were imposing and, second, to see what those who ought to have been least accommodating to the change thought they were accepting. For Yuan Daoxue advocates, the literary culture and guwen writers of the Tang and Song provided a foil against which they could define themselves. Zheng Yu (I298-I358), the master of a private academy who declined an appointment to the Hanlin Academy, introduced his own literary collection in I3 so, the "Left-over Strength Manuscripts" (from the Analects passage: "If he has any strength left over [after learning to behave morally], he studies wen") with the following statement. 63 After Mencius each scholar treated his own opinion as learning and wenzhang as dao. Therefore Han Yu, Liu Zongyuan, Ouyang Xiu, and Su Shi all gained fame in their time in this manner. Since the authors said, "I am able to fully realize the transmission of This Dao of Ours," those who came later and honored them said further, "These are the men who are to be known as having transmitted from the right source." [In fact,] they blocked the perception of society and forced our populace into ignorance. Thus the Dao was not clear because wenzhang had beclouded it; the Dao was not practiced because wenzhang had entangled it. The crimes of wenzhang were greater than bears saying. In the early Song the Cheng brothers from Henan appeared. Society called them the two Masters Cheng and, for the first time, came to understand that the transmission of This Dao of Ours does not depend upon language and writing; it is complete within my own nature. It does not exist in a lofty and distant realm of nothingness; it operates in the midst of daily practice and common conduct. With this [understanding] one fathoms principle,

Peter K. Bol

49

with this [understanding] one illuminates the Dao; with this [understanding] one purifies the body and transmits it to posterity, with this [understanding] one resolves doubt and enlightens the confused. With this the composition of wenzhang was discarded and became a useless thing. The heavy rain now ceased; once again heaven and sun became visible, and the awful apparitions from crevice and crag were no more. Then Master Zhu of Xin'an completed a grand synthesis of the great scholars. When he discussed moral principle [daoli], he always recorded it in wenzhang, and when he composed wenzhang, he always based them on moral principle. Those who had hitherto been shackled walked free, and the beclouded saw. It is true indeed that "those possessed of virtue will speak words of truth." Wenzhang became a device for threading the Dao and not a term for empty words. When I was in my teens and still ignorant, having not settled on any "former sayings and past deeds" [as my guide], I heard others reciting Master Zhu's words and felt as if they were coming from my own mouth. When I heard them speaking of his dao, I felt as if it were expressed from my own mind. Since I liked it so much, I gave it all my strength-! took him as the authority not only for moral principle but also for wenzhang. Once I became thoroughly versed, I did not know I was studying the ancients but felt as if the ancients were like me. A man laughed at me: "Taking Han and Liu as authority for wenzhang and Zhu and Cheng as authority for moral principle is one thing everyone agrees on; isn't the way you learn mistaken?" I said, "There is no wen outside dao. To go outside the Dao of the Sages and Worthies to do wen is not what I mean by wen. There is no dao outside wen. To go outside the wen of the Six Classics to seek dao is not what I mean by dao. For this I refer to Master Zhu." 64

Wen matters, Zheng Yu insisted, for what men write creates self-sustaining traditions, enabling writers to use their own values to shape those of society. Historically the importance of wen has had the unfortunate consequence of keeping people from seeing that dao has always been located inside human nature, that it is accessible through learning, and that it is applicable to the activities of daily life. Zheng was particularly exercised by the great guwen writers, for the obvious reason that in their writings they laid claim to the dao of the sages without seeing that its grounds were in human nature. Following Cheng Yi, Yuan moralists also blamed the loss of "our Dao" on the elaborate commentaries of Han and Tang Classicism and the rise of Buddhism and Daoism in addition to literary composition. 65 But Han Yu, who had popularized the Mencian paradigm of the loss and recovery of the dao of the sages, was a more immediate problem for many of them. 66 Han could only be saved to an extent. As Liang Yin (1303-89) explained, "Yu sought dao through learning wen. He knew its function

50

EXAMINATIONS AND ORTHODOXIES

included all affairs, but he never understood that its substance was complete in my mind. He knew it could be practiced in society, but he never understood that its basis ought to be first [realized] in my own life. " 67 Zheng Yu's preface contains the great claim of the Neo-Confucians: the dao-to be understood both as the dao of the sages that Confucius and Mencius spoke of and the dao of the self-sustaining, integrated universe-existed complete in the self. In this sense "seeking the dao" (qiu dao) meant seeking to actuate in daily practice something one already possessed. Among believers there was argument over the appropriate means of seeking. The most important Yuan debate was summed up by the two phrases "honoring the moral nature" and "the learning of study and inquiry," which were associated with the more intuitive approach of Lu Jiuyuan and the intellectual, "book-reading" style of Zhu Xi, respectively. This was, I think, simply another version of the guwen problem of fusing individual engagement in thinking about values with an appropriation of the cultural tradition as the public manifestation of values. There were many who devoted considerable effort to defending Zhu Xi's teaching and his textual legacy through the creation of anthologies, curricula, and commentaries. 68 There were also men who converted to Lu's "learning of the original mind. " 69 Wu Cheng is the most famous example of a man who tried to hold the two together in creative tension. 70 Another example of this effort comes from Cheng Duanxue (1278-I334), brother of the Cheng Duanli quoted above. One works back through cumulative layers of interpretation, he explained, from Cheng-Zhu to Confucius and Mencius to the Classics, in order to seek the "means by which" the sages transmitted and Confucius taught. As one does this, he is acquiring knowledge of "the models of the sages." The "means by which" is the existence of the faculty of moral judgment in the self ("the steelyard and measuring rod in me"), which one does his best to actuate and extend. The models of the sages are the forms and texts of what the Neo-Confucians saw as the true cultural tradition. In effect, one persuades himself that he has the ability to judge at the same time he is learning the sagely models with which to judge. Through cumulative practice, Cheng promised, one ultimately attains coherence, the "realm where all is threaded together. " 71 If we wish to speak of a Nco-Confucian "orthodoxy" in a Yuan context, we probably should keep in mind that its content is less definite than Wang Anshi's commentaries and glossary. The Nco-Confucian idea that each individual is endowed with a moral faculty beyond intellectualization opens the door to considerable diversity in practice. Clearly many

Peter K. Bol

literati held that there were morally responsible models of speech and action that mediated the search for the dao. Here again it seems to me that the richness of the Song cultural legacy alone mitigated against an easy definition of the range of correct behavior and belief. Another way of looking at the problem is to ask whether Daoxue proponents could allow that there were ways of thinking about values that had to be judged according to standards outside Daoxue. In the end, could they allow literature autonomy-? Xu Youren, who passed the first examination in I3 I 5, suggested that ultimately there would be no autonomy. The spread of Cheng-Zhu learning in the Yuan had improved the quality of wen, he contended, for "when a unifying principle [yi-li] dominates, wen is included in it. " 72 Gong Shitai (I 29 8-r 3 62) reached a similar conclusion by a different route. Since the dao was based on the correctness of universal principle in the human mind, the person who had cultivated an understanding of dao internally before he spoke would effortlessly write things that were wen. 73 This implies, I think, a view that good literary standards have a real foundation that Daoxue cultivation attains without intent. Such a view would account for the claim that a Neo-Confucian teacher transformed the imitative, shallow wen of the local literati by teaching them that in "understanding both the Classics and Daoxue, one has to take Master Zhu as authority. " 74

Intellectual Diversity in a Daoxue World It is easy to find Neo-Confucian literati sharply attacking traditional literary intellectuals and literary endeavors, but rare are cases of Yuan literati criticizing Daoxue as a movement. One of the few outspoken critics writing after 1313, Yuan Jue (1266-1327) asserted that its rise had led to (r) neglect of the practical work of government, (2) a decline in literary accomplishment, and (3) a willingness to sacrifice facts in order to attain mental coherence. 75 Yet Yuan also admitted that the spread of Zhu Xi's writings and ideas about human nature had brought a degree of cultural commonality-"literati have been able to unify what they see [in books] and hear [from teachers]"-just as the dynastic founders had unified all China.76 Yuan was exceptional in my view. Some literary men simply did not mention the existence of Daoxue, but neither did they establish themselves as rna jor intellectuals. 77 But what are we to make of the fact that some of the most influential defenders of cultural accomplishment and Tang and Song guwen thinkers also promoted the idea that Song Neo-Confucian thought was a necessary foundation of Yuan literati culture? The best-known examples of this are

52

EXAMINATIONS AND ORTHODOXIES

the "Jinhua" scholars from Wuzhou lu in the Jiang-Zhe Branch Secretariat (modern Jinhua in Zhejiang). These scholars are also notable for the selfconsciousness with which they promoted "Wuzhou learning," as heir simultaneously both to the pragmatic statecraft of the Song Wuzhou scholars Lii Zuqian, Chen Liang (1143-94), and Tang Zhongyou (1136-88) and to the moral philosophy of Cheng Yi and Zhu Xi. 78 The first generations of Jinhua scholars were not of a single sort. HuangJin (1277-13 57) favored statecraft studies; Xu Qian (1270-1337) actively promoted Daoxue; Liu Guan (1270-1342) was closer to being a Song-style guwen intellectual.79 Song Lian and Wang Wei (1323-74) lived through the fall of the Yuan and became advisers to the founder of the Ming dynasty. Although we have more collected writings fromJinhua than any other place in the Yuan, there were literati with similar attitudes in other regions: Yu Ji (12721348) and Wei Su (1303-72) from Fuzhou in Jiangxi and Ouyang Xuan (1283-1357) from Liuyang in Huguang. 80 I shall suggest three general explanations all these scholars gave for accepting the imposition of the Nco-Confucian curriculum. The first was that Daoxue could be seen as an evolving historical tradition whose texts and teachings were useful as a tool for thinking rather than a dogma to be learned. HuangJin passed the first Yuan examination in 1315 and was well acquainted with both Nco-Confucian texts and scholars. 81 As a teacher at the Imperial College, he objected to empty talk about the principles of heaven and man, nature and decree, associated with Neo-Confucianism. 82 He was not a spokesman for Cheng-Zhu teachings or, for that matter, the teachings of any figure.s 3 None of the ten inscriptions for schools and academies in Huang's collection even discusses Daoxue teachings or texts. 84 The intellectual world we encounter in Huang ]in's writings does not seem transformed by moral philosophy, yet he does not object to the Nco-Confucian examination curriculum or to the pre-eminence of Zhu Xi. One explanation for this is Huang's conviction that he was part of an evolving intellectual tradition. "Those good at saying things worthy of being remembered [li yan] do not need to come from antiquity, nor do they need not to come from antiquity. If there is not some difference [with earlier views], there is no need to write a book, but neither is there a need for it to be transmitted if there is not some agreement. "S 5 His contemporary Liu Guan similarly treated Daoxue as part of a larger historical tradition of learning for self-transformation, which began with Classics, was extended by Mencius, Xunzi, Yang Xiong, and Han Yu, and was brought back to the essentials by the Song philosophers. 86

Peter K. Bol

53

Neo-Confucianism was broad enough to include their particular interests. For Liu the evidence that one's spiritual and intellectual discipline had improved was that one would "naturally get at moral truths and one's poetry would naturally get at emotional correctness. " 87 Huang Jin's concerns were more practical. He compiled a six-chapter anthology and study of the six arts (archery, mathematics, calligraphy, etc.) in which he drew on multiple sources, including Zhu Xi. It was, he said, a book not only for the antiquarian and erudite but also for the literatus of his own day who favored the grand but ignored the details, who pursued nature and destiny at the expense of institutions. Thus, he concluded, "I make clear the means with which the Former Kings taught and realize what Master Zhu wanted to accomplish. " 88 I am not sure that Huang was being ironic. His Jinhua contemporary Xu Qian, an outspoken proponent of Cheng-Zhu who published a number of interpretations of the Classics and Four Books, stressed both thinking for oneself and investigating the diversity of "the many" rather than unity of "the one." Cheng Yi's teachings should not be understood as "knowing that li is one and not having to seek it in the diversity of functions." 89 · Even Yu Ji, one of the most influential of the southern officials and a tireless advocate of Neo-Confucianism as the recovery of true learning, 90 agreed that Daoxue was part of an evolving historical tradition. The NeoConfucians did not make literati who gained fame before them worthlessmen like Ouyang Xiu, Su Shi, and Zeng Gong-nor did it mean that later men had nothing to contribute. The Neo-Confucian commentaries did not exhaust the sages' meanings. 91 Zhu Xi's books made it possible for a literatus isolated in a provincial backwater, without teachers and friends, to understand and actuate the way of the sages for himself. 92 The point was to "comprehend the sages' minds and attain the sages' dao," not to make sure that whatever one said agreed with Cheng and Zhu. 93 He charged the literati of the day to continue seeking the values of the sages for themselves and not stop at Cheng and Zhu. They taught men how to seek; they did not finish the seeking. 94 A second reason was the usefulness of the Neo-Confucian idea that all people are inherently capable of acting morally. For Liu Guan the innate goodness of human nature guaranteed the success of efforts to transform the populace through education. 95 For Ouyang Xuan, also a jinshi of 1315 and a successful court official, the idea of a moral nature provided grounds for redefining loyalty to the ruler as "realizing the principles of what ought to be so, endowed in us by heaven." 96 The private scholar Wu Lai, the son of a clerk who had become a scholar-official and

54

EXAMINATIONS AND ORTHODOXIES

mentor to the Mongol official Toghto, warned against attempts to reduce Confucianism, the moral teachings for an inclusive society, to an exclusive dogma. 97 He depicted Daoxue as a historical movement that suffered from divisions and vituperative internal debates between followers of Zhu and Lu. Yet he also held that the universality of a mind possessed of moral principles meant that commonality was possible, not only between past and present or self and other but also between Chinese and barbarians. 98 The Neo-Confucian understanding of human nature allowed for political independence for the sake of principle while providing a foundation for social and political unity. A third reason literati found to accept the Daoxue curriculum is that some literati persuaded others that it was possible to historicize Daoxue, viewing it from the perspective of the guwen tradition, while accepting its philosophical premises; in short they integrated the reasons discussed above. More than anything else, I suspect, this is what enabled the "Jinhua school" to believe it had a claim on the center of intellectual culture. Song Lian and Wang Wei, who lived through the Yuan-Ming transition, represent a self-conscious revival of guwen as a mode of thinking and writing about values. John D. Langlois, Jr., speaks of Song as having a "cosmology of wen." 99 Another way to put this is to say that Song Lian used the term wen as the overarching concept to hold together the connections he saw among culture, learning, morality, and responsible social and political action. Song's "cosmology of wen" is found in his two-part essay "On the Origins of Wen" ("Wen yuan"), in which wen is both the origin of civilization and the external expression of the individual's nurturing of his vital energies. The essay concludes: "Outside the Six Classics one ought to take Mencius as authority, then Master Han [Yu], then Master Ouyang [Xiu]. Then the thoroughfare for the state will be freed of brambles and snakes, and one may hasten directly along the great way of the sages and worthies." 100 Like the guwen writers of the Tang and Song he admired and against those who were merely literary, Song Lian believed that the literary intellectual is responsible for values. "In general, in doing wen, we wish our words to get the point across and our dao to be clear; what more is there to ask for? However, the dao is not easy to make clear; before we can succeed, we must be able to understand words and nurture qi." 101 But at the same time these writers claimed to be the true representatives of Daoxue. Wang explained in a preface to one of Song Lian's collections: since the worth of wenzhang stems from the fact that it illuminates moral principles (li), it follows that to do writing of value one must, like the

Peter K. Bol

55

sages, have the knowledge first. Song Lian's great achievement was to see that "wenzhang is the means for conveying scholarship [xueshu]." But this is not all. Wang Wei argued that Song Lian was heir to two Song traditions in Jinhua. The first reached back to Lii Zuqian, as heir to Zhou Dunyi and the Chengs, and Tang Zhongyou and Chen Liang, as utilitarian statecraft scholars. This had continued through the Yuan in Jinhua with such men as Liu Guan, Huang Jin, and Wu Lai. But Song apprehended the purport of Daoxue from a second tradition that went back to Zhu Xi and Huang Gan (II52-I22I) and then spread to Jinhua, so that it had now become a center of Nee-Confucianism (lixue). 102 Song's writing, Wang concluded, reveals his scholarship and illuminates li; it makes him a sue. cessor to all those who went before. 103 What should we make of this fusion of guwen and Daoxue-a fusion that clearly was not acceptable to some advocates of Daoxue but was influential nevertheless? My answer for the moment is that Song and Wang wanted to historicize thinking about values, and thus free themselves from having to be loyal transmitters of any school (I assume that, as is often said, the Classics can be made to contain everything) and allow them to speak to the problems of their own times, but they also wanted to insist that there was a real foundation in the individual for thinking about values, a necessary underpinning for the claim that literati outside government could judge and guide their times. There was, Song concluded in a discussion of his teacher Huang Jin, an inevitability to cultural and intellectual change and diversity. Had not the Six Classics differed from each other from the start? Had there not been multiple schools of thought? Did not the great prose writers-Sima Qian, Wang Tong, Han Yu, Ouyang Xiu, and the Chengs-differ from each other? This was inevitable, Song wrote, for wen, like the energies of heaven-and-earth, was "daily renewing" without exhaustion, constantly varying as a result of the individual spirit. And Huang "himself said, each piece of literary writing conveys my learning; dare I agree with another for the sake of getting along? It is enough not to turn against the sages." 104 But for Song himself the individuality and creativity of the literary intellectual had a universal basis. "Wen would not be established but for dao, nor filled but for dao, nor practiced but for dao. Because one's mind and dao are one and dao and heaven are one, therefore all the words that come forth are constants [jing, also: 'Classics']. " 105 For Wang Wei also, writing about values is the most important of all wen. Writing in the philosophical schools category is the best of all writing, better than elaborate parallel and regulated literary forms, examination

EXAMINATIONS AND ORTHODOXIES

writing, writing that spreads the ruler's influence, and historical writing. It is wen that "conveys dao," for although the ideas of thinkers differ, they are ultimately one, even if none of them is "the great whole of the dao." Above this are only the Six Classics, the sages' "perfect wen" and "devices that convey the dao." Yet for Wang the Classics convey attitudes, not dogmas: constancy (Odes), variability (Change), substantiality (Documents), decisiveness (Annals), inclusiveness (Music), and differentiation (Rites)-all values that can be represented through literary style. 106 These perhaps are the values that, taken together, make it possible for Wang to see the principles of things. As he wrote elsewhere, in wen one need only be concerned with illuminating li; if he can do so in his own mind, his qi will provide the necessary energy and his words will convey the li. 107 It seems to me that Song and Wang and their predecessors in Jinhua were neither true "literary men" in the manner of some fourteenth-century literati nor good Neo-Confucians. They were post-Daoxue guwen intellectuals. Wang himself was willing to be called a shi (literatus) or a ru (Confucian) who participated in "the learning of the sages and worthies." This learning, he explained in "On the Origin ofRu," differs from Classical exegesis and literary composition, the two traditions of literati learning that had dominated intellectual culture since the Han. It shows literati how, through learning from the textual tradition, everyone can establish his 0wn internal foundation while developing the talent to accomplish things in response to external events. It shows that one can, when mature, by thinking in his mind, fathom the li of all things and attain knowledge. Clarity of mind is equivalent to the realization of one's nature; insight into li makes it possible to see that all things are connected to oneself. There is thus a "single basis" that ties investigating things to bringing peace to all under heaven. In short, this learning teaches men how to think for themselves in a socially responsible way. To reduce it to a set of texts to be memorized or to treat it as nothing more than a form of literary writing is to fail to see that "the learning of sages and worthies" differs from the learning of Classical exegesis and literary composition, even as it includes them. 108 This essay has asked why the literati most concerned with defending the mediating role of literary intellectuals and of intellectual diversity in thinking about values of culture resisted the government's imposition of a national New Policies examination curriculum in the Northern Song but accepted the imposition of the Nco-Confucian curriculum in the Yuan, even though many committed Neo-Confucians intended it to be inimical to their interests.

Peter K. Bol

57

I have argued in essence that these literati ridiculed Wang's curriculum as an attempt to force intellectual uniformity. They accepted Daoxue, however, because they believed its fundamental message was that the ultimate grounds of moral value lay in the individual and thus it provided them, once they chose to see Daoxue as an evolving historical tradition, with the authority to claim intellectual and moral leadership in their own times. This suggests that in fact what literati had begun to fight for in the I07os-independence from the court's authority in order to claim for themselves cultural and moral authority over society-was being realized in the fourteenth century. The curriculum being imposed from above was a curriculum that both Daoxue advocates and scholars of the Jinhua sort understood as locating moral authority with the literati. No one was arguing that the court, through either the ruler's unique authority or imperially sponsored scholarship, was the final arbiter of common values. 109 Learning as "seeking the dao" and "apprehending something of oneself" supposed a process of thinking about what one ought to value that could be individual and open-ended. Even those who insisted on a narrow textual tradition to mediate that search could not reduce Nco-Confucianism to rote learning. The idea that the individual scholar has to engage in this process of learning became the bedrock of literati intellectual culture; Yuan literati of all sorts thought this shift had taken place in the Song, although they disagreed about who should get the credit. These ideas were in place before the Yuan; they cannot be explained simply by the not surprising desire of literati under foreign rule to maintain some degree of cultural independence. I am not sure, in fact, that the Mongols cared one way or the other how literati thought about such matters; it may be more appropriate to see literati as speaking to each other about the need to maintain their independence. The Ming founder Zhu Yuanzhang, however, was sensitive to such claims and was not inclined to accept even a hint of opposition, as the murder of tens of thousands during his reign reminds us. Still, I would question whether he changed the course of later imperial society. A better conclusion to draw from this chapter is that from the Song into the Yuan literati secured a basis for their cultural and intellectual authority and independence and in this manner became largely impervious to the imperial court's occasional attempts to legislate their values. Surely such a picture fits better with our sense of late imperial local society. The Nco-Confucian curriculum may have been enshrined in the examinations, but the message it was thought to carry denied the authority of the state and its examiners over those they examined.

CHAPTER 2

The Formation of "Dao Learning" as Imperial Ideology During the Early Ming Dynasty Benjamin A. Elman

/\f;//hen the Prince ofYan in Beiping (Northern peace), Zhu Di (I36o-

f/f/ I424), installed himself as the Yongle (Eternal happiness) emperor (r. I402-24) of Ming China (I368-I644), it is said that he personally asked both the eminent Confucian scholar Fang Xiaoru (I3 57-I402) and then capital vice-censor Lian Zining (d. I402) to serve him. A controversial secundus in the I 3 8 5 palace examination, who had dared to criticize in his examination essay the executions of many officials, Lian had loyally and bravely served both the founding Hongwu (August military; r. I 3 6898) and the succeedingJianwen (Establish culture; r. I399-I402) emperors. When Lian contemptuously refused and berated the Prince for his immoral actions, Zhu Di had Lian's tongue cut off to silence him. The Prince of Yan then justified his military revolt, his occupation of the Ming capital, Yingtian (present-day Nanjing), and his ascension in place of the ]ianwen emperor by saying: "My only desire was to emulate the Duke of Zhou, who came to support young King Cheng." 1 Lian put his finger to his mouth and, using his blood, traced on the ground: "Where is King Cheng?" 2 In Fang Xiaoru's case, the confrontation was equally chilling. The Prince of Yan, politely at first, demanded that Fang, a confidant of the Jianwen emperor and most likely the last official to see him, draft the announcement of the Prince's succession to the throne. When Fang refused and labeled Zhu Di a criminal, the two men got into a heated argument: 3 "I modeled myself on the Duke of Zhou, who served King Cheng and no more."

PRINCE:

Benjamin A. Elman

59

"Where is King Cheng?" "He burned himself to death [in the palace]." FANG: "Why don't you establish King Cheng's son as emperor?" PRINCE: "The nation requires a mature ruler." FANG: "Then why don't you establish King Cheng's younger brother as ruler?" PRINCE: "These are my family's affairs and that's all." FANG XIAORU: PRINCE:

The Prince became agitated and gave Fang Xiaoru a writing brush to prepare the announcement of his accession. Fang threw the brush to the ground and in tears scornfully continued the argument: "If I must die, then so be it. I will not write the draft for the announcement." PRINCE (loudly): "How can you expect to die so suddenly? In dying, are you not concerned about your relatives to the ninth degree?" FANG: "What does it matter to me if you make it to the tenth degree?" FANG:

The Prince of Yan, realizing that Fang Xiaoru would never acknowledge him as the new emperor, ordered his attendants to use knives to slit open Fang's mouth on both sides up to his ears. Then, Fang was tossed back into prison, where his friends and followers were brought to him one by one. When Fang refused to see them, all were killed. It is said that Fang was in agony for seven days, but until his death he continued to mock Zhu Di for his pretensions and left a lyric that became famous. 4 "King Cheng," that is, theJianwen emperor, was probably dead. Only the remains of the empress and their eldest son were found in the debris of the burned palace. Zhu Di conducted a funeral service for the fallen ruler on July 20, one week after the fall of the capital on July I 3, although rumors that the Jianwen emperor had escaped the burning palace continued throughout the Ming dynasty. 5 Lian Zining was executed, along with his entire family and lineage. Of the latter, over I 50 people were murdered, some only distantly related (the ninth or tenth degree of kinship). Several hundred others were banished. In Fang Xiaoru's case, 873 relatives were also executed. Besides Fang and Lian, the kin of other Jianwen loyalists were also eliminated; estimates range as high as ten thousand for the total number of officials and members of their families murdered in 1402. 6 The Jianwen reign was expunged from the historical records, becoming instead the thirty-first to the thirty-fifth years of the late (d. I 398) Hongwu emperor.? The "Veritable Records" of the early reigns were twice tampered with, and in the final version of the account of the Hongwu reign the "Veritable Records" were doctored with falsehoods to confirm Zhu Di as the legitimate and sole successor to his father, the Hongwu emperor. 8

6o

THE FORMATION OF "DAO LEARNING"

As the Eternal Happiness emperor, Zhu Di became a powerful and influential monarch, known posthumously as Taizong (Paramount scion) and Wen huangdi (Emperor of culture) since his death in 1424, the latter an ironic-even if clever-choice given the military manner in which he had removed the Establisher of Culture (Jianwen). And in r 53 8 Zhu Di was granted the additional temple name Chengzu (Formative ancestor) by the Jiajing emperor (r. 1522-66), who like Zhu Di established a new line of imperial succession. In effect, Zhu Di was the second founder of the Ming dynasty (he moved the capital from Nanjing to Beijing [Northern capital] in 1415), after his father, posthumously known as Taizu (Paramount ancestor). Moreover, after the usurpation, Zhu Di actively promoted Confucian studies, especially the Southern Song (I I 2 7-I 2 79) Confucian persuasion known as Daoxue (Dao Learning), which had become the core curriculum of the civil service examinations in I 3 I 3 during the Yuan dynasty (!28oI368) and since 1384 during the Hongwu reign. Zhu's own Shengxue xinfa (The methods of the mind in the sages' teachings), completed in I409 and presented to his designated successor, was emblematic in his mind of the unity of the daotong (orthodox transmission of the Dao) and the zhitong (statecraft legitimacy) that he claimed for his reign. 9 The "sagely Duke of Zhou" had by his own hand become a "sage-king," a paragon of Cheng-Zhu (Cheng Yi, I033-I IO?, and Zhu Xi, II30-12oo) ideals. In addition, Dao Learning moral philosophy was successfully utilized as political ideology to draw attention away from the events of 1402. 10 The "Jianwen martyrs," however suicidal their actions, were motivated by morally compelling Confucian political ideals that had preceded but were still part of Dao Learning in the tumultuous early years of the fifteenth century. At the same time, however, the political usurper, Zhu Di, culturally usurped the Dao Learning strand of Song moral philosophy and became a great Ming emperor. Both sides claimed orthodoxy, the authorization of the past to legitimate present actions. 11 Zhu Di was the victor in the world of power, but Lian Zining and Fang Xiaoru became historical legends. Both the martyrdom of loyal Confucian officials and the sageliness of the new emperor were woven into the historical tapestry known as the "Ming dynasty." Who was the real Confucian? Zhu Di? Fang Xiaoru? Who had the right to judge? Was Zhu Di's support for Dao Learning merely a ploy? Or was it his only way psychologically to cleanse himself and the officials loyal to him for the brief period of bloodletting he had personally ordered? As historians, not philosophers, we must deal with both sides.

Benjamin A. Elman

61

To be sure, most Mandarins, as we will see, were indispensable handmaidens in the wedding between Dao Learning philosophy and state autocracy during the Hongwu and Yongle emperors' reigns. Moreover, the repeated ideological uses of Dao Learning by rulers such as Zhu Di were not accidental or fortuitous. The Confucian canon and its commentators, after all, had since the Han (206 B.C.-A.D. 220) and Tang (6 I 8-907) dynasties been supported by and in turn been supportive of the imperial system. Li Shimin {the future Emperor Taizong of the Tang, r. 626-49 ), for instance, had assassinated his brother, the chosen successor, in 626 and then forced his father, Emperor Gaozu (6 r 8-26), to abdicate in his favor. Later in 63 8, Li Shimin authorized the compilation of the Wujing zhengyi (Orthodox meanings in the Five Classics) to provide the definitive textual basis for Confucian learning. 12 After the fall of the Song dynasty, Yuan, Ming, and Qing rulers wisely chose the moral philosophy of Dao Learning to serve this ideological function. Given the alternatives offered by Buddhism, Daoism, or popular religion, which Mongol, Han Chinese, and Manchu emperors also utilized to assert dynastic legitimacy, 13 their appeal to Dao Learning put them in touch with the most lettered and influential elite: the Han Chinese literati (shi) .14 In Weber ian terms, some sort of "selective affinity" between the Ming state {the ruler and his bureaucracy) and Confucian philosophy (Dao Learning) likely existed. In particular, we should note .the hidden affinity that united the purely moral values of Dao Learning and the social values of local Han elites, an affinity that gathered covert strength during the Yuan dynasty when Han Chinese were marginalized under Mongol rule. When Mongol rulers in the early fourteenth century turned somewhat from coercive to cultural forms of control vis-a-vis the Han Chinese, they reproduced a structure of political relationships that during the Song had empowered Han elites through the civil service examinations in the state bureaucracy. In effect, by the early Ming, the values of the state educational system for elite men were automatically linked with the Dao Learning sympathies of local elites. The latter thus were advantaged in their quest for prestige and status because the civil examinations tested what they took as their cultural birthright: mastery of Dao Learning. A marriage of convenience between the ruler and his Confucianized elites lurked beneath the events of 1402. 15 Such a view, however useful, is still too one-sided. On the other side were the few martyrs like Fang Xiaoru and Lian Zining who defied imperial ideology disguised as orthodoxy and chose the Confucian path of martyrdom. Dao Learning served both sides. The raison d'etre for both imperial

THE FORMATION OF "DAO LEARNING"

power and literati idealism was located, selectively to be sure, in the Confucian canon. Both sides could appeal to important aspects of that canon to legitimate their actions and claim the "orthodox" legacy of the past on their behalf. What I hope to provide here is an analysis of the cultural content of state ideology in early Ming China, or in other words, how Dao Learning orthodoxy, as the state defined it, over time served imperial purposes. At the same time, however, the legends of Lian Zining and Fang Xiaoru remain anomalous surds in the historical record. If the early Ming is a frightening period in Chinese history because of the terrible slaughter of officials by the Hongwu and Yongle emperors, it is also frightening because a few Confucians were willing to die, and sacrifice almost all their kin, rather than submit to them. How doctrine becomes ideology is an important historical question. In raising this question, we move immediately from the internal integrity of philosophical positions to the political, social, and economic uses of ideas in particular historical contexts. How ideas inform and authorize action is a question that carries us beyond the domain of "pure" philosophy and the traditional history of ideas. Instead of interrogating ideas in "texts" for their universal "meaning," we decipher how they reveal the particular "contexts" of those whose actions were informed and served by references to those ideas. In the contemporary turn from the history of ideas to cultural history, our role as intellectual historians shifts from trusting in the ideals of philosophy to distrusting their historical uses. 16 The Dao Learning orthodoxy in court politics and elite society (usually mispackaged as "Neo-Confucianism") has hitherto been praised by intellectual historians for its philosophical vision as a sophisticated and multidimensional set of metaphysical doctrines and moral teachings, which emerged during the Song dynasties (960-I279) and were later systematized by Zhu Xi, arguably the greatest philosopher in Chinese history. But outside the domain of its contemporary champions, the Cheng-Zhu school of Dao Learning has also been blamed by social historians for its political uses as an autocratic state ideology. 17 Song Dao Learning philosophy and political autocracy became dubious partners during the Mongol Yuan dynasty when, at the urging of Confucian (many non-Han Chinese) advisers in I 3 I 3, the interpretations of the great Song philosophers Cheng Yi and Zhu Xi were for the first time made the orthodox guidelines for the imperial examination system belatedly resumed in I 3 I 5. Thereafter, the reproduction of the bureaucracy was premised on the mastery of Dao Learning moral philosophy. In addition to this educational function, the

Benjamin A. Elman

Cheng-Zhu school also provided the cultural language of imperial power, as emperors like Zhu Di and his successors claimed the mantle of the sage-kings for themselves. They repossessed the "orthodox transmission of the Dao," and thus their "statecraft legitimacy" before the Confucian literati was reestablished. 18 This brief partnership in turn led to a long-term political and cultural relationship that was consummated in a formal wedding between Dao Learning teachings on the Classics and imperial state power during the Ming and Qing dynasties. Han-Chinese Ming and Manchu Qing emperors, like their Mongol predecessors, believed that the Cheng-Zhu school provided the cultural and political justification for their rule. When emperors selected Cheng-Zhu learning as the verbal machinery of their rule, they in effect tied the constitutionality of their dynasty to that philosophy and committed the state to its educational propagation in schools and on civil examinations. How did this come about? "Orthodoxy" and "ideology" are frequent terms in this paper. Although there were many political and cultural definitions in Ming China, by state "orthodoxy" I mean what the state publicly authorized as orthodox. At times, such as during the Song dynasty, the state orthodoxy came from wider literati circles; at other times, such as in the early Ming, it was made the core of the civil service examination curriculum and thereby influenced literati culture. Thus, Dao Learning moral philosophy chiefly concerns me here not as an autonomous field of inquiry with its own inherent intellectual integrity and growth, which of course it had, but rather as a system of concepts, arguments, and beliefs endorsed and manipulated by the state for its larger political purposes. That process of manipulation-when concepts, arguments, and beliefs selectively serve to legitimate political sovereignty-is what I refer to as "ideology." In this chapter, I will try to show how the late imperial state in China successfully incorporated Dao Learning philosophy into the civil service examination system to enhance its larger agenda of training loyal officials who would share power with the ruler and serve the larger interests of the dynasty. The political coherence of imperial ideology derived from its intimate ties to and selective reproduction of Cheng-Zhu learning. That ideological coherence, however, was more extracted from than reflective of the philosophical doctrines on which it was based. State ideology may have had many selective affinities with Confucian moral philosophy, but the political purposes to which those affinities were applied were determined by the needs of the state rather than the integrity of the philosophy. The emperor (or those who spoke for him), not the philosopher, had the

THE FORMATION OF "DAO LEARNING"

final say on how Confucian concepts, arguments, and beliefs were put into educational practice through the civil examinations. Hence, the actions of the Jianwen martyrs were never part of the examination curriculum or an object of official study. Their legends, however, lived on, just as candidates relearned, generation after generation, the sagely model set by the Duke of Zhou in dutifully serving his brother's son, King Cheng. As a carefully crafted Confucian disguise worn by an autocratic but not yet totalitarian state, Dao Learning when translated into imperial ideology helped justify, that is, induce public acceptance of, the bureaucratic and military forms of power on which the Ming and Qing empires were largely based. This chapter, then, seeks to identify during the early Ming the political and cultural uses of that imperial disguise in the required educational curriculum of the civil service examinations. Many helped to create it. Many saw through the disguise. After 1425, most, fortunately, never had to face cruel but charismatic sage-kings like the Hongwu emperor or the Yongle emperor again. Hence, they could live with the disguise, and literati during the late Ming were able to change and modify it to serve their needs, thereby reopening the tragic cases of Fang Xiaoru and Lian Zining for re-evaluation.

From Usurper To Sage: The Yongle Emperor and the Ideological Uses of Dao Learning The Eternal Happiness reign began with a bloodbath. It ended with Zhu Di, the Prince of Yan, a declared sage-king in the line of Yao, Shun, and Yu. Unlike Han and Tang emperors, who had not received the transmission of statecraft legitimacy, the Yongle emperor, like his father, had repossessed the orthodox transmission of the Dao through his support for Dao Learning. He could not have accomplished this feat without Confucian collaborators. If the suppressed memories of Lian Zining and Fang Xiaoru lived on as legend, the cultural work needed to augment Zhu Di's imperial legitimacy and to domesticate Dao Learning and turn it into orthodox ideology was carried out by men who made the transition from serving the Jianwen emperor to submitting to the Yongle emperor with few overt qualms of conscience. For every Lian Zining, there were many others like Hanlin academician Yang Rang (1371-1440) who greeted the Prince of Yan when he entered Nanjing and chose to serve the new ruler. Yang changed his name from Zirong to commemorate the occasion. Yang Rang had taken his jinshi in 1400, placing sixth, under the tutelage of the Jianwen emperor. 19

Benjamin A. Elman

The examiners and I IO graduates of the 1400 jinshi examination were representative of the fact that most officials did not heed the Confucian injunction that having served the Jianwen emperor, they could not very well serve his murderer (bu erchen). 20 Records from the palace examination of 1400, for example, show that Fang Xiaoru was one of the Jianwen emperor's readers for the examination, and Xie Jin (I3 69-14I5) was one of the officials in charge of collecting the candidates' papers. Both were Hanlin academicians. Rather than martyr himself in I402, Xie Jin, a jinshi of I 3 88, chose the path of least resistance. For his loyalty, Zhu Di reappointed Xie to the Hanlin Academy and immediately delegated him in I402 to go through the Jianwen emperor's papers and remove anything that might be used to challenge the usurpation of power. Next, and still in I402, the Yongle emperor put Xie Jin in charge of the first revision of the Hongwu emperor's "Veritable Records," which had to be altered to confirm Zhu Di as emperor and denigrate the Jianwen emperor as ruler. Among the changes introduced was the claim that Zhu Di was born to the Empress Ma (he was likely born of a concubine). The version of history that Zhu Di had Xie Jin help prepare (there was a second revision; see below) thus presented Zhu as the eldest surviving son of Zhu Yuanzhang. Zhu Di rightfully should have been designated the heir apparent in I392 when his older brother died, but through the machinations of unscrupulous advisers serving the future Jianwen emperor, the successor became instead the unqualified son of the original but now dead heir. 21 Hu Guang (I 3 70-14 I 8) was selected as optimus on the I400 palace examination. Like Lian Zining, Hu was one of the Jiangxi elite so prominent in the Jianwen emperor's court. Hu had finished eighth on the metropolitan examination the same year and second in an earlier Jiangxi provincial competition. Curiously, the Jianwen emperor had Hu change his name to Jing before he entered the Hanlin Academy because the original was the same as that of a Han dynasty official. 22 Another Jiangxi native who took his jinshi degree in 1400 wasJin Youzi ( I3 68-I43 I), who finished seventh on the palace examination and thirteenth on the metropolitan. Earlier he had finished ninth on the Jiangxi provincial examination. 23 Jin was a fellow townsman of Lian Zining. Both had grown up and studied Confucius' Spring and Autumn Annals together as young men in preparation for the civil examinations. 24 After I402 both Hu Guang and Jin Youzi loyally served the Yongle emperor. Zhu Di did not have to worry about any regional opposition from Lian's cohorts. Hu immediately changed his name back to Guang. Changing names perhaps mitigated the moral dilemmas Hu faced.ZS Reappointed by Zhu Di as a Hanlin academician, in

66

THE FORMATION OF "DAO LEARNING"

I4 I4 he was placed in charge of the Wujing Sishu daquan (Complete collection [of commentaries] for the Five Classics and Four Books) project that the Yongle emperor authorized to define Cheng-Zhu Confucian orthodoxy (see below). Jin Youzi joined Hu Guang and Yang Rang on this influential cultural project. 26 In addition to ordering changes in the "Veritable Records of the Hongwu Reign," whose first revision was completed in 1403, Zhu Di also instructed Xie Jin in 1404 to employ some I4 7 scholars to bring together all extant classical, literary, and historical writings in a single collection, which when completed in 1404 the emperor named the Wenxian dacheng (Great collection of written documents}. In some ways this project continued similar projects initiated during Zhu Yuanzhang's reign. Xie Jin in 13 88, for example, had suggested to the Hongwu emperor that he authorize a compilation of essential Song Confucian writings on the ClassicsY Previously in I 3 73-7 4, Zhu Yuanzhang had already indicated his interest in copying Tang Taizong (see above), in compiling definitive records of classical learning and ancient institutional works. At that time a work entitled Qunjing leiyao (Classified essentials of the Classics) was prepared to explicate the Four Books and Five Classics. 28 The haste with which this project was undertaken, so soon after the usurpation, and the fact that Xie Jin, who had just completed an initial whitewashing of the "Veritable Records," was chosen to collect and edit the materials suggests that Zhu Di had political as well as cultural motives in mind when he said to Xie: 29 The world's affairs and matters from antiquity to today are scattered throughout many books. They are not easy to examine or read. I want to gather all affairs and matters recorded in each book and unify all the various compilations by using a phonetic scheme to facilitate study. Whatever contains words dealing with the techniques and crafts of the classics, histories, philosophers, litterateurs, the hundred schools, astronomy, geography, yin-yang, prognostication, medicine, Buddhism, and Daoism should all be collected into a single work.

In addition to collecting documents, the compilers could ferret out materials damaging to the legitimacy of Zhu Di's accession to the throne. It is likely, then, that the compilation had its darker side, political shadows that reached deep into the eighteenth century when the Qianlong emperor, who well understood Zhu Di's motives and knew the records of the Yongle reign very well (see below), authorized in the 1770s and r78os the voluminous Siku quanshu (Complete collection of the four treasuries) project in part to ferret out anti-Manchu writings. 30

Benjamin A. Elman

The political implications of the project are clearer when we take into account Zhu Di's unhappiness with the Wenxian dacheng when it was completed in 1404Y In 1405, the Yongle emperor instructed his close confidant Yao Guangxiao (1335-1418) to undertake the project. A staff of 2,169 compilers, among whom were Buddhists and medical specialists, was placed at Yao's disposal to comb all known works and copy them for the project. A Buddhist monk critical of Dao Learning, 32 Yao had been instrumental in encouraging Zhu Di to revolt against the Jianwen emperor in 1399 and remained until his death one of the emperor's closest advisers. Moreover, shortly after completing this encyclopedic project in 1407, which became known as the Yongle dadian (Great compendium of the Yongle era), Yao was put in charge of a second revision of the Hongwu emperor's "Veritable Records" that lasted from I4II to 1418. All copies of the first revision completed by Xie Jin, like copies of the original completed during the Jianwen reign, were eradicated. All that survived was Yao's second revision. 33 When we note that there were 472 graduates of the 1404 metropolitan civil service examination, the first under the Yongle emperor, and that this was the highest number since I 3 8 5, when Zhu Yuanzhang needed to fill his depleted bureaucracy after the Hu Weiyong (d. 1380) affair (in which Zhu eliminated those he charged with treason), then we can agree with Danjo Hiroshi's conjecture that the sudden increase in jinshi reflected the emperor's need to produce, immediately, literati loyal to him and not the Jianwen emperor. When asked by his examiners what the quota should be in 1404, for example, the emperor replied that he wanted to set the quota at its highest level to date but that this should not be taken as a precedent. 34 After the completion of the palace examination, the emperor further ordered that all candidates who had earlier failed the metropolitan examination should be re-examined in a special literary examination. In this way, another 6o loyal students were chosen to enter the National School to prepare for the next metropolitan examination. 35 Xie Jin, an examiner in 1400 under Jianwen, was again chosen examiner in 1404 under Yongle. 36 The optimus for the 1404 palace examination was Zeng Qi (I372-I432), from Jiangxi. In fact, the top seven places went to candidates from Jiangxi, the top three from Jishui county, Xie Jin's and Hu Guang's own home county. As in 1400, the Jiangxi graduates quickly filled the Hanlin Academy. In total, 24 percent of the 472 graduates came fromJiangxi, compared with I 8 percent from Zhejiang and I 5 percent from Jiangnan. In 1406, 25 percent of the jinshi came from Jiangxi; in 141 I this figure rose to 3 2 percentY Later in 1404, Xie Jin was promoted

68

THE FORMATION OF "DAO LEARNING"

to the rank of grand secretary. 38 Loyalty to the new emperor, particularly among the Jiangxi elite, was clearly rewarded. 39 Throughout the Yongle reign, Jiangxi natives took 25-30 percent of the jinshi degrees. Not surprisingly, the palace examination questions and answers for the 1404 and 1406 examinations, and thereafter, made no mention of the Jianwen reign or even that there had been a civil war. As in the "Veritable Records," the Jianwen reign simply disappeared. Zhu Di himself prepared the policy question for the 1404 palace examination, although on later palace examinations he delegated the responsibility to Hanlin academicians. In 1404, the emperor asked candidates to explain the different institutional systems of antiquity; in 1406 he required them to comment on the changes in schooling during the Han, Tang, and Song dynasties. His focus in 1404 was on the political order; in 1406 the emphasis was the role of education in ordering society. 40 Zeng Qi's policy answer in 1404, which earned him first place, simply conceded that "the emperor had received the mandate to rule and now occupied the position of a sage." Zeng, paraphrasing Zhu Di's question, continued: 41 Your humble servant recognizes that the emperor's mind-heart is the mindheart of the Yellow Emperor, Yao, and Shun. All those sages before and after have had this mind-heart. The Paramount Ancestor [i.e., Zhu Yuanzhang], as sage and worthy, [wisely] wielded both civil and military power. The exalted Ming was set in motion, and its great virtue was accomplished. He was the esteemed ruler who unified heaven and magnified filial piety and thereby exemplified in reality this mind-heart. That is why Your Majesty has felicitously continued the intentions of the people and spoken of their affairs. How can [the needs of] this age and this people be disregarded?

Lin Huan, the top graduate in 1406, who went on as a Hanlin compiler to work on the Yongle dadian project, similarly made the ruler's mind-heart the central theme of his answer: 42 I have heard that order has its origin. It emerges after one first seeks illumination in the mind-heart. Enacting order has its model. It resides in examining widely into antiquity. Consequently, seeking illumination in the mind is the origin of order, and examining into antiquity is the trace of [that] order. Sages in ordering the world never failed to take the examining of antiquity as the Way. Moreover, when have sages not based the occurrence of order on the mind-heart? Only the Paramount Ancestor [Zhu Yuanzhang], as sage and worthy, [wisely] wielded both civil and military power. The exalted Ming was set in motion, and its great virtue was accomplished. He was the esteemed ruler who unified heaven and magnified filial piety. He initiated the building of the broad foundation, taking care that the six schools would

Benjamin A. Elman

be united and the political unification would last ten thousand generations. His achievement has been rarely duplicated. Your Highness has succeeded with precious intent in thoughtfully forging ahead of past glories and continuing the beatific plans [of before] and opening tomorrow's glory. Lin's verbatim duplication of part of the earlier pledge of loyalty prepared by Zeng Qi suggests that such pledges had political rather than cultural meaning. Throughout the Ming, formulaic paeans to the emperor became a major aspect of the civil examinations. 43 In Yongle's case, they papered over his 1402 usurpation of power. Imperial focus on the unified mind-heart of the emperor in 1404 and 1406, which became one of the most widely wielded slogans in Ming civil examinations, dated from the Yongle reign, when Zhu Di successfully used this Dao Learning image after the "terror of 1402" to co-opt moral legitimacy of Dao Learning for his usurpation. The xinfa (methods of the mind-heart) of the sagekings had been a frequent subject in literati writings to that point, but it was during the Yongle reign that this philosophic doctrine was turned into imperial ideology on the civil service examinations. From 1371 to 1400, not a single policy question on the palace examination had addressed the sage-kings' methods of the mind. 44 After the 1404 and 1406 palace examinations, the subject appeared repeatedly in policy questions on the Ming metropolitan examinations. 45 Zhu Di's Shengxue xinfa, completed with the help of Hanlin academicians in 1409 and presented to the designated successor, the future Hongxi emperor (r. 1425 ), for his moral cultivation, preached the unity of the orthodox transmission of the Way and statecraft legitimacy. 46 In many ways a prelude to the Dao Learning compendium known as the Xingli daquan (Great collection of works on nature and principles), compiled in 1414-15, Zhu Di's own selection of Cheng-Zhu Confucian commentaries represented imperial reauthorization of Dao Learning as cultural and philosophic orthodoxy. Zhu Di's elucidation of the famous doctrine of the daoxin (mind of the Dao) serves as a representative example. Citing Zhu Xi and his Song disciples, the emperor demonstrated that he was in complete agreement with Zhu Xi that the mind of the Dao, as the venue for moral principles, should be the master, while the "human mind" (renxin), the venue for selfish desires, should take its orders from the former. The emperor, in effect, took Zhu Xi's place in educating his son. Moreover, Zhu Di claimed that he had repossessed the "xinfa of the kings and emperors of ten thousand generations. " 47 Perhaps by 1409 the Yongle emperor had put behind him his brutal actions in 1402, but even if he had, the almost obligatory use of the xinfa

70

THE FORMATION OF "DAO LEARNING"

doctrine as early as the I404 and 1406 palace examinations suggests that more than just the imperial conversion to Dao Learning orthodoxy was involved. As a true sage-king, Zhu Di could have it both ways: he used Cheng-Zhu Learning for political legitimacy; at the same time he became perhaps the greatest imperial patron of Song Dao Learning. Yet, his preface to the Shengxue xinfa did have some darker elements. Speaking of the principles of loyalty the ruler expected of his officials, Zhu Di wrote: 48 When the ruler becomes a ruler like Yao and Shun, the people become like the people of Yao and Shun. As for illustrious officials such as Gao [Yao], Gui, [Hou] Ji, and Xie, 49 can they not be called glorious? Therefore, in upholding high principles without submitting to threats, nothing is greater than loyalty. Those who receive their positions and salaries from the ruler should concern themselves with the nation as they do their families. They should forget about themselves and follow the nation. They should not avoid difficulties or dangers. Nor should they make plans according to their own benefit. They should strengthen their resolve and ascertain that their resolve is unchangeable.

Lian Zining, Fang Xiaoru, and the Jianwen martyrs, not Zhu Di's well-placed collaborators such as Xie Jin or Hu Guang, had lived up to this ideal. But the former remained anathema in public life, although the emperor did on occasion, as in I4I3, wish that Lian Zining had submitted to him. On the other hand, Hu Guang, when he died in J4I8, was greatly honored for his service. Earlier, however, Xie Jin, hated by the Yongle emperor's designated successor for Xie's opposition to Zhu Di's choosing him over the Prince of Han, was eventually thrown into prison in I4 I I, where he died in I4I5, on charges of lese majeste. 50

Establishing a Classical Curriculum: The Wujing Sishu daquan Projects In addition to the "correction" of the historical record, Zhu Di's regime required educational legitimation. The Yongle emperor wished to "appear as sage ruler, a teacher of his people, and a patron of learning. " 51 These cultural endeavors were brought to a climax in I4I5 by publication and dissemination of three classical projects, the Sishu daquan (Complete collection [of commentaries] for the Four Books), Wujing daquan (Complete collection [of commentaries] for the Five Classics), and Xingli daquan, whose purpose was to define and print for use in all government schools down to the county level the sources candidates should use to prepare for the civil service examinations.

Benjamin A. Elman

7I

Confucian scholars first prepared two major projects, which enshrined the Cheng-Zhu school of lixue (studies of principles) as civil service examination orthodoxy. Entitled the Complete Collection [of commentaries] for the Five Classics and Four Books, these two anthologies were compiled in great haste. It took only nine months, from I4 I4 to I4 r 5, for Hu Guang and his staff to prepare and blend the mostly Yuan scholia (which frequently cited earlier Han, Tang, and Song commentaries) into a coherent passage-by-passage commentary for the Five Classics and Four Books. The haste with which the commentaries were compiled, perhaps to add cultural luster to the events surrounding the move of the capital to Beijing in 1415, elicited much later criticism for their lack of comprehensiveness. Because the three collections were compiled by Hanlin academicians who had previously been employed in revising the "Veritable Records of the Hongwu Reign" and had also helped edit the Yongle dadian, later scholars such as Gu Yanwu (r6r3-82} suspected that the real purpose of the Sanbu daquan (Great collections trilogy) was to cover up the accomplishments of the Jianwen reign in classical learning. 52 The ruler [that is, Zhu Di]lied to the court, and those below [that is, the bureaucracy] swindled the literati. Was there ever anything like this in Tang or Song times? Did they not compromise honest and upright officials while replacing the ]ianwen emperor? Moreover, when the writing of [eight-legged] examination essays began, literati all at once discarded the "practical learning" transmitted since the Song and Yuan dynasties. Those above and below were mutually gullible and became fixated on careerist interests, never stopping to ask why. Alas! The demise of classical studies in reality began from this. Other Qing Confucians also frequently blamed the Yongle projects and their pervasive influence on the examinations for the decline of classical studies during the Ming dynasty. The compilers of the Ming History had similar complaints. 53 Zhu Di's intent becomes clearer when we read the preface he immediately prepared when the last part of the trilogy, the Xingli daquan, was completed and all three works were officially authorized for printing. Zhu wrote in I4I 5 that upon succeeding to the throne, all sage-kings had "used the Way to order the world." Hence, he himself, "as successor to the illustrious foundations established by the great emperor, the Paramount Ancestor," had ordered the Hanlin academicians to prepare the three works "to include whatever had clarified the meaning of the Classics and to exclude whatever was contrary to the lessons of the Classics." 54 In their own statement, Hu Guang and the compilers (including Yang Rong and

THE FORMATION OF "DAO LEARNING"

Jin Youzi) echoed Zhu Di's pretense that the Jianwen emperor had never existed and that Zhu was the legitimate successor by praising him in no uncertain terms: "Never before has there been a ruler of such great action, who has been able to clarify the Way of the Six Classics and continue the [national] unity from our previous sages like this." 55 The Great Collection of Works on Nature and Principles represented curriculum support for the moral philosophy of Song Dao Learning and was required reading for the first session of the provincial and metropolitan civil service examinations for the duration of the Ming dynasty. The Complete Collection for the Four Books, interestingly, included a full and unexpurgated version of the Mencius for the first time since the 1370-72 provincial, metropolitan, and palace civil examinations. Zhu Yuanzhang had been angered by the political philosophy of Mencius, whom the ChengZhu school had enshrined as Confucius' doctrinal successor in the Confucian lineage of orthodox teachings. 56 What infuriated the Hongwu emperor was the passage in the Mencius (one of the Four Books and hence part of session one of the Ming civil service examinations) in which Mencius defended an ascending view of political power based on the people and drew limits to the loyalty an official owed his ruler: When the ruler regards his officials as the ground or the grass, they regard him as a robber and an enemy.

Such questioning of the bonds between the ruler and his officials was tied to Mencius' claim that the ruler served the people: 57 Mencius said: "The people arethe most important element [in a state]; the spirits of the land and grain are the next; the sovereign is the least important."

In fact Mencius cited Confucius for his position: 58 Confucius said: "The Way has only two courses, that of benevolence and that of malevolence. One who carries the oppression of his people to the highest pitch will himself be slain, and his kingdom will perish. If [such a ruler] does not carry oppression to an extreme, his life will still be in jeopardy, and his kingdom will be weakened. He will be called [a tyrant] as You [r. 781-770 B.c.] and Li [r. 878-827] [in antiquity] were."

Such Confucian political parables, Zhu Yuanzhang had discovered, directly challenged the state's sovereignty, based since the early empire on a descending view of power emanating from the ruler. Instead, for Mencius the state derived its power from the "ascending" will of the people. These threats to imperial sovereignty were more than Zhu Yuanzhang could tolerate. Around the time he abolished the civil examinations in 1372,

Benjamin A. Elman

73

Zhu demanded of his ministers that the text of the Mencius be removed from the reading list for civil service examination candidates. In addition, Zhu called for removing Mencius from the sacrificial ceremony performed for Confucius and the official Confucian pantheon of sages, scholars, and martyrs. 59 Zhu Yuanzhang's ministers, particularly Qian Tang, tried to head off these dangerous precedents, which threatened Cheng-Zhu intellectual orthodoxy, by agreeing to remove those passages in the Mencius that the emperor found objectionable. They successfully prevailed on him, however, to keep the remaining text as required reading for the examination system. Because the examinations were soon stopped, however, the censored version of the Mencius was not used in the civil examinations until I 3 84, after Zhu had successfully purged his enemies from the civil and military bureaucracies. 60 Later, in 1394, Liu Sanwu (13 12-99?) was entrusted with preparation of a formal edition of the Mencius that deleted 8 5 sections that Zhu Yuanzhang found objectionable. The expurgated version was entitled the Mengzi jiewen (Abridged version of the Mencius). 61 Thereafter the censored edition of the Mencius was the standard text for civil examination candidates until !4!4-IS, when the original curiously was restored. The Yongle empero~ also restored Mencius' tablet to the Confucian temple as part of his efforts to patronize Confucian letters. Having usurped the throne from the Jianwen emperor, whom he accused of incompetence and heresy, and whose officials he described as traitorous vermin (jianchen), Zhu Di apparently had little to fear from Mencius' legitimation of regicide. He himself had just removed an emperor whom he accused of corruption and immorality. 62 In fact, perversely, what had offended Zhu Yuanzhang now could be welcomed, although still unofficially, by Zhu Di as justification for his military action. An unworthy ruler had been rightfully forced out of office. 63 Despite this concession, the Daquan collections clearly represented the changing political circumstances within which the Five Classics and Four Books were interpreted. 64 The early Ming compilers of the Complete Collection of the Four Books, which highlighted for examination candidates Zhu Xi's explication of the Great Learning, Analects, Mencius, and Doctrine of the Mean, fully accepted, for example, the rigorous moralism that derived from Zhu's bifurcation of heavenly principles from human desires. They were later accused by Qing dynasty Han Learning advocates of accepting a Buddhistic vision of good versus evil, which was more formalistic and inflexible than the more qualified and nuanced dualism that Zhu Xi himself had enunciated. 65

74

THE FORMATION OF "DAO LEARNING"

When they chose comments on the text of the Mencius, for example, the Hanlin scholars working on the Complete Collection of the Four Books, such as Hu Guang, still were careful to set limits to Mencius' discussion of the legitimate grounds for officials and the people to oppose an evil ruler. For the passages that earlier had enraged the Hongwu emperor but were tolerated by Zhu Di, Hu Guang chose commentators who stressed that Mencius' words applied only to the chaotic historical situation of the Warring States period. Accordingly, Mencius stood as a guide to the past. His criticism of ancient tyrants could not be translated into a binding precedent for later, "enlightened" ages. Moreover, the commentaries chosen indicated that he had set limits in his attack on evil rulers: 66 Without Mencius's theories, there would have been no way to warn those who during later generations served as rulers of the people .... However, Mencius said: "If one has the moral integrity that Yi Yin [renowned advisor of the first Shang ruler, ca. 1766-1753 B.C.] had [in urging banishment of the ruler], then it is permissible [to do so]. If, however, one does not have the moral integrity of Yi Yin, then a person [who speaks in this way] must be a usurper."

When the ruler's sovereignty was based on a descending view of political power, then the moral criteria of loyalty predetermined the limits of political criticism. If Confucians dissented from the policies of their ruler, the forms of dissent were ideologically circumscribed. In state affairs, high moral ground had been granted the ruler. He alone determined the acceptable limits of Mencius' theories in the civil examinations, although some late Ming literati outside the bureaucracy, many of them associated with the Donglin Academy in Wuxi and the Fushe (Return to antiquity society} in Suzhou, began to challenge the court's descending view of power at the same time that they historically rehabilitated the Jianwen emperor, Fang Xiaoru, and Lian Zining. Furthermore, the concern for self-cultivation, so prominent in the writings of Song Confucians such as Cheng Yi and Zhu Xi, and so heralded by the Yongle emperor after 1402, likely did play an important function in granting Confucians some political autonomy and moral prestige, which was institutionalized through the avenue of moral remonstrance. Remonstrance as a form ofpolitical dissent in turn served to measure the ruler according to universal Confucian standards. Zhu Di seems to have understood this, even if his motives for supporting Dao Learning had their opportunistic political side. During the Ming and Qing, however, rulers frequently closed this avenue of dissent or diverted it into a form of bureaucratic surveillanceY

Benjamin A. Elman

75

Moreover, even if the complete Mencius was rehabilitated in I4I 5, the passages that had troubled Zhu Yuanzhang in I372 were thereafter rarely chosen by examiners for essay topics in the provincial or metropolitan civil examinations. 68 Nor were these passages often raised in MingQing palace examinations. In uncensored form (a major concession by Zhu Di to Confucians to be sure), Mencius was allowed to speak to his readers. In substance, however, the passages critical of ancient rulers were usually left out of the actual proceedings of the civil examinations. 69 The Daquan trilogy was thereafter printed and distributed to all county and prefectural schools for use by students in their preparation for state examinations. A sage-king like Zhu Di was the beneficiary of all that the Confucian canon could offer. Even its dissenting portions had been turned to his favor. Epilogue The Yongle emperor, after moving the capital from Nanjing to Beijing in I4 I 5, eventually became one of the greatest rulers in Chinese history. Some have linked the move to Zhu Di's desire, in part, to get away from Nanjing, where the memories of the I402 martyrs stilllingered.7° Moreover, after I425, the civil examinations became the principal means for filling higher offices in the state bureaucracy. From I238 in north China and I 2 7 I in the south until I 3 I 5, there had been no state examinations of any kind in China. Between I 3 I 5 and I4 so, most official positions were still filled by recommendation or other special procedures. Only on the I 3 8 5 and I 404 metropolitan examinations, discussed above, did early Ming civil examinations ever approach in scope and magnitude the level of Song civil service appointments to the bureaucracy through examinations. For all intents and purposes, the hiatus in the institutional role of civil examinations as a major feature of Chinese political life lasted two centuries. Accordingly, although the designation of Dao Learning as the ~tate's ritual orthodoxy by the Southern Song court in I24I and its designation as civil examination orthodoxy by the Yuan in I 3 I 3 were important events, these were only steps to the eventual triumph of Cheng-Zhu Dao Learning during the Yongle reign. By becoming the principal road to official appointment, the civil service examinations thereby guaranteed that the state curriculum established in I4I5 based on Dao Learning and the Daquan trilogy, which became the key texts in the curriculum, would be studied and mastered by millions of examination candidates for the civil service

THE FORMATION OF "DAO LEARNING"

until I9oo. The cultural reproduction of Dao Learning in this manner was institutionalized for half a millennium. 71 The early Ming thus marked the unrivaled highpoint of the status of Cheng-Zhu Dao Learning in the "examination life" of Han Chinese literati. Put another way, although many Confucian believers had perished, Dao Learning overall had benefited from the sometimes tepid support of Zhu Yuanzhang and the "unswerving faith" of Zhu Di. So much so that Qing Confucians such as Zhu Yizun (I629-I709) and Quan Zuwang (17055 5) rightly, if naively, believed early Ming emperors to be stalwarts of Cheng-Zhu orthodoxy. These late imperial examples of a convenient, historical amnesia about the early Ming dynasty have lasted into our time. 72 But the triumph of Dao Learning as state ideology during the early Ming was achieved at a considerable price. In the process, the linkage between autocratic state power and Cheng-Zhu philosophic discourse was consummated in a wedding between perhaps the most powerful and expansive emperor in Chinese history, Zhu Di, and Dao Learning moral ideals based on personal self-cultivation. Zhu Di, using the sage-kings' methods for the transmission from one ruler to the next of the sagely mind of the Dao, became a Dao Learning exemplar. Ming Confucians thus served a state whose emperors at bottom remained insensitive to subtle shadings of moral theory or less than stark hues of right and wrong. Early in the Hongzhi reign (r. I488-I505), for example, the court attempted to eliminate what it considered "immoral shrines" (yinci). This action was not unprecedented. Early in his reign, as we have seen, Zhu Yuanzhang had wanted Mencius' tablet removed from the official Confucian sacrifices. What was new in the later policy was a proposal to remove all tablets commemorating Confucius' disciples and all Han dynasty Confucians from the Confucian Temple, a proposal that was actively debated. In I 5 30, the tablet commemorating the Later Han (25220) classicist Zheng Xu an (I 2 7-200) was removed from the Confucian Temple. Even though initially opposed and successfully mitigated by Ni Yue (I444-I50I), who contended that Han Confucians had been invaluable in preserving and transmitting the Classics, the subsequent removal of Zheng Xuan's tablet revealed the degree to which the Cheng-Zhu Dao Learning orthodoxy during the early Ming had begun to sanctify itself through policies of ritual exclusion/3 Universal "truth," legitimated by the state in the form of the hegemony of Cheng-Zhu Confucian discourse in late imperial official and educational life, drew the Classics and Four Books into its own conser-

Benjamin A. Elman

77

vative agenda, rather than following, for example, the more reformist agendas of earlier Northern Song or Former Han Confucians. Despite challenges in the sixteenth and seventeenth centuries by the influential philosopher-general Wang Yangming (14 72- r 5 29) and his vocal followers, such questioning of state orthodoxy never effectively penetrated the curriculum of the Ming civil service examinations. The classical slogan "exalt the ruler and expel barbarians" (zunwang rangyi) derived from the Spring and Autumn Annals, for instance, became in the hands of Ming officials a means, at times unsuccessful, to exalt the ruler's authority and to place upon his subjects demands for unquestioned loyalty. A descending view of political power remained orthodox. 74 How Dao Learning Confucian philosophy first took hold as imperial ideology in late imperial China is best explained by unraveling the interconnected political, moral, and institutional threads that were woven together into their final, imperial form under the Yongle emperor. Even if Emperor Wu (r. qo-87 B.C.) of the Han, Taizong of the Tang, and Taizu of the Song (r. 960-76) were his historical predecessors in such imperial cum Confucian cultural endeavors, and even if Emperor Lizong (r. 122564) of the Southern Song and Renzong (r. 13 12-20) of the Yuan had already placed Dao Learning on the imperial pedestals of first ritual and then civil service examination orthodoxy, Zhu Distill left a legacy for the Ming and Qing dynasties that overshadowed them. The imperial state and its Cheng-Zhu ideology that collapsed in the late nineteenth century drew its cultural lineage directly-Song and Yuan precedents notwithstanding-from the early Ming. Early Ming emperors chose Dao Learning to win the support of their most important social group, the literati. One of the exact uses of Dao Learning for an emperor such as Zhu Di who had usurped the throne was that support of Cheng-Zhu learning as a cultural vision of the late imperial state translated into a compelling raison d'etre for political legitimacy that induced literati obedience. When it was politically institutionalized, the cultural content of Cheng-Zhu philosophy allowed the Yongle emperor and his ministers to appeal to the Way of the sage-kings, which since antiquity had been the model for contemporary governance, and thus repossess the daotong. They claimed, moreover, that the moral principles of antiquity had been transmitted, mind to mind, from the sage-kings to the present emperor. Thereafter, emperors, Mandarins, and Dao Learning were inseparable. The social habits, political interests, and moral values inherited by Mandarins since the Yuan and Ming dynasties were then

THE FORMATION OF "DAO LEARNING"

officially reproduced (with much unofficial and official dissent) through a system of state schooling and civil examinations that acquired their mature form during the fifteenth century and lasted until the late nineteenth century/5 The memory of Zhu Di's violent usurpation weighed heavily on the conscience of the Ming, however. The famousJiangxi literatus Wu Yubi (13921469 ), for instance, refused to participate in early Ming examinations because he was unwilling to serve the Yongle emperor, whom he considered a usurper. Such refusal alienated him from his father, Wu Pu (I 3 631426), who had passed the 1400 metropolitan examination ranked first (he was fourth on the Jianwen palace examination) and had by 1403 become actively involved in revising and reissuing the "Veritable Records of the Hongwu Reign" for which the Yongle emperor quickly promoted him a grade in the Hanlin Academy. Wu Pu then served as a deputy chief compiler for the Yongle dadian. Wu Yubi's animosity toward the Yongle reign and the disgraceful actions of his father was echoed by his Jiangxi disciple Hu Juren ( 1434-84), who also refused to take the examinations, even though the Yongle emperor was already dead. Hu spelled out the subtext of his actions by appealing indirectly in his writings to the Mencian injunction against political usurpation. Both Wu Yubi and Hu Juren became model Dao Learning scholars honored for their integrity and moral cultivation by later Ming Confucians. Despite the instrumental uses of Cheng-Zhu learning by the early Ming state, its intellectual standing among dissenting Confucians like Wu and Hu would become a beacon for late Ming Cheng-Zhu scholars. 76 In addition, troubling legends of the heroism of Lian Zining and Fang Xiaoru remained, and their question "Where is King Cheng?" still went unanswered. Zhu Di himself only relented in his persecution of the families of the Jianwen martyrs in qr6, when he learned from Hu Guang, who had just returned from Jiangxi to attend his mother's funeral, that the people there were finally pacified. 77 Pardons were granted in succeeding reigns for surviving family members of the Jianwen emperor's executed officials. In 1425, for example, thousands of Fang Xiaoru's descendants were rehabilitated by the Hongxi emperor. 78 The next year, the Xuande emperor (r. 1426-3 5) pardoned Lian Zining's surviving family. 79 But not until 1573 were all the Jianwen martyrs pardoned. 80 Pressure on the Ming house to accord the Jianwen reign itself full legitimacy began to grow in the Wanli reign (1573-1619), during which Tu Shufang (fl. 1564-98) compiled the Compendium of Unofficial Records

Benjamin A. Elman

79

on the ]ianwen Reign. Tu's interest in the martyrs dated from I 579 when he had been appointed an official injiangxi and discovered that the martyrs' descendants were still subject to recrimination. He submitted a memorial in I 584 asking for their full pardon. Then in I 59 5, Tu began work on the Compendium, which was completed in 1598. About the same time a shrine for the Jianwen martyrs was erected in Nanjing, and a local shrine in Jiangxi was built for Lian Zining. 81 In an afterword to Tu's work, Yao Shilin (rs6I-r6sr?) wrote of the pain and sufferings the 1402 martyrs had undergone, which, he contended, had overturned the Way. 82 Moreover, in the I59os, Zhu Lu (IS 53-163 2), initiated a work (revised in 1615 and printed in 1621) entitled ]ianwen shufa ni (Draft of the principles for compiling documents of the Jianwen reign), in which he reconstructed, with annotations, the Jianwen benji (Basic annals of the Jianwen reign). Zhu Lu thereby established the historical legitimacy of the Jianwen emperor and publicly described the manner in which the "Veritable Records of the Hongwu Reign" had been censored and rewritten by Zhu Di's unscrupulous officials to expunge the Jianwen reign from historical documents. Although he blamed Zhu Di's officials and not the emperor, Zhu Lu made public the manipulations of history that had occurred: 83 Emperor Wen's [i.e., Zhu Di] public-mindedness managed to extinguish his private motives, but hadn't he revealed blemishes on his record? Those who polish jade try very hard to gloss over and thereby remove blemishes, but the blemishes are irrevocable. For a thousand years it cannot be completely clear. History today is just like this!

Under the Wanli emperor in I 59 5, the ]ianwen reign title was officially restored after repeated requests. 84 Li Zhi (I527-I6o2), in his iconoclastic writings, heralded the bravery of both Fang Xiaoru and Lian Zining, which caused Li to wonder about the political cost of the dynastic unity the Yongle emperor had achieved. 85 Even in official circles, the Hanlin academician Jiao Hong (I 54 I-I 620 ), for example, included official biographies for the Jianwen martyrs in his I6I6 biographical collection of noted Ming officials from 13 68 to I522 entitled Record of Verified Documents During the Ming Dynasty. 86 But the Jianwen emperor did not receive a posthumous temple name until 1644, and then only from Ming loyalists in Nanjing after Manchu forces took control of the northern capital of Beijing. The next year, the Southern Ming court hurriedly granted posthumous honorific names to thejianwen martyrs, including Fang Xiaoru and Lian Zining. Ai Nanying (I 58 3-I 646), an outspoken Jiangxi critic of late Ming political affairs who served the Southern Ming after the fall of Bei-

So

THE FORMATION OF "DAO LEARNING"

jing, prepared an eloquent defense in I63 I of both the moral stature of the ]ian wen emperor and the righteous legacy of the "Jianwen martyrs" :87 If the emperor wishes to make plans for the affairs of the empire, then he must first teach the empire about loyalty. If he wishes to teach the empire about loyalty, then he must first clarify the distinction between the ruler and his officials. Moreover, if he wishes to use those eliminated Uianwen] officials [as examples] to teach the empire about loyalty, then he first must be clear about how the Jianwen emperor went about being the ruler. Then he can show how the eliminated Uianwen] officials went about being officials. The Jianwen emperor was in fact the eldest grandson of the great emperor, the Paramount Ancestor [Zhu Yuanzhang], and the eldest son of the designated successor. He was in power for four years. Reverent and compassionate, he was never deficient in virtue. When he received [the mandate] and took the throne his name was truly orthodox.

Ai Nanying reassessed the legitimacy of the Ming in light of the tragic events of 1402, after which Zhu Di had moved the capital to Beijing. Using the form of a memorial sent from the Ministry of Rites in Nanjing to the emperor, Ai contended that the present reign must acknowledge its ties to the fallen Jianwen emperor and his martyred officials. Zhu Yuanzhang's successor had not been Zhu Di. Ai Nanying intimated that the succession of Ming emperors after Zhu Di was illegitimate. By writing the memorial from the Jianwen emperor's Ming capital of Nanjing, Ai urged the Ming court to acknowledge the mandate to rule that Zhu Yuanzhang had transmitted to his grandson some 260 years before. The ]ianwen martyrs should be vindicated. 88 Qian Qianyi (I 5 82-I 664), a noted Ming Confucian, was serving Prince Fu in Nanjing when the I645 pardons were announced. Earlier he had spent considerable time in his Mingshi gao (Draft history of the Ming dynasty; now lost) unraveling the events surrounding what he considered Zhu Di's immoral usurpation. Qian's elegy to Fang Xiaoru and the Jianwen martyrs said of Zhu Di only that "the Prince of Yan's grandsons were today's emperors." When Nanjing fell to the Manchus later in 1645, he was among the first important southerners to give allegiance to the Manchus. 89 Moreover, many late Ming scholars wrote of Yongle collaborators such as Hu Guang with scorn: "The Jianwen emperor personally selected Hu Jing [Guang] to be the [palace examination] optimus. The imperial grace granted him was incalculable, and yet Ching threw the emperor away as he would a ceremonial cap. " 90 The fall of the Ming dynasty forced many to reconsider the moral legitimacy of the dynasty they had served. After two centuries of suppression, in which the historical

Benjamin A. Elman

8r

records of their lives had taken on unofficial but legendary meaning as heroic Confucians, Lian, Fang, and the Jianwen martyrs were publicly apotheosized into sagely officials. Under the Qing dynasty, Ming emperors and ministers were initially discredited, and the Jianwen martyrs were openly praised. The early Qing Confucian Shi Runzhang wrote in his r663 preface to Lian Zining's collected works: 91 When I read history like this, I immediately threw away the rolls [of the book] and became speechless, unable to keep my eyes from opening wide [in anger]. By writing on the ground using the blood issuing from the tongue in his mouth [lit., "heavenly gate"], Lian enabled us to see how by energizing his tongue like the sun and moon he wielded his pen to scold [Zhu Di] with great wrath.

Meanwhile, Dao Learning survived to provide the Manchu Qing dynasty with its required Confucian legitimation to elicit from Han literati their voluntary submission. The civil examinations were immediately restored, although the Oboi regents for a time tried to put an end to the by then infamous eight-legged essays on the Four Books. After this attempt failed in 1667, the mastery of Cheng-Zhu learning again became the chief route to high office for Han Chinese males. 92 Ming loyalists, fortunately, were not as suicidal as their Jianwen predecessors. Instead, they lived to see the "enemy" wrap itself in Dao Learning garb and complete the transfer of the mandate to rule to the Manchu imperial family and its conquering armies. Moreover, the Manchus were sensitive to the moral sensibilities of famous Ming loyalists such as Fu Shan (1607-84}. 93 When the Kangxi emperor (r. 1662-1722) had the Hanlin scholars compile the Dao Learning tract entitled the Xingli jingyi (Essentials of works on nature and principles) (issued in 1715 ), and the Gujin tushu jicheng (Synthesis of books and illustrations past and present} encyclopedia, revised and printed in 1728 under the Yongzheng emperor (r. 1723-3 5 ), he was taking a page out of the Yongle reign to present himself as a Confucian sage-king. Likewise the Qianlong emperor (r. 173 6-9 5) in 1773, when he ordered the compilation ofthe greatest bibliographic project in history, the Siku quanshu, which was used in part to ferret out antiManchu writings. The r673 Kangxi Preface to the reissuing of the Ming version of the Xingli daquan, for a final example, linked early Qing Confucian legitimacy to the cultural policies of the Hongwu and Yongle emperors and based that legitimacy on the methods of the mind transferred from the early

THE FORMATION OF "DAO LEARNING"

sage-kings to their Qing peers. The Kangxi and Qianlong emperors became enlightened sage-kings too. 94 Harold Kahn has put it best: "Perhaps, after all, the historian in such cases was helpless: it took an emperor to know one. " 95 Although Fang Xiaoru and Lian Zining were legitimate heroes by the eighteenth century, their heroism was now part of the Qing cultural tapestry rewoven from early Ming cultural and political strands. ChengZhu learning was again at the center of the tension between imperial interests and literati moral values.

CHAPTER

3

Canon Formation in Late Imperial China Pauline Yu

i/tandard accounts of Chinese literary history have long told us that the supreme literary achievement of China's elite culture was the classical poetic tradition, whose highest expression was attained during the period corresponding roughly to the first half of the eighth century, later identified as the High Tang. 1 Both constituents of this commonplace have been subjected, appropriately, to scholarly interrogation in recent years. On the one hand, the critical lens has widened its focus considerably to reveal the increasingly wide authorships and audiences for other literary forms in late imperial China. On the other hand, the privileging of Tang-especially High Tang-poetic production has been destabilized not only by attention to the actual multifariousness of a mythically monolithic style and to the prolific poetic productions of later ages but also by research into a range of cultural activities, such as writings by women and non-canonical poetic forms, that were previously either consigned to the margins of scholarship or not recorded at all. These investigations have offered, and will continue to offer, a salutary counter-argument to the image of a late imperial literary world constrained within rigid, narrow, and pre-emptive orthodoxies. Before we rush to dismiss the historical valorization of High Tang poetry as a simple discursive formation, however, we should pause to consider the possible impulses and conditions contributing to its enduring power. Such an examination may provide insights not only into specific cultural values and the forces behind them but also into the broader process of canon formation itself. Some 2,200 poets and almost 49,000 works were collected in the Complete Poems of the Tang (Quan Tang shi), commissioned by the Qing emperor in r 70 5 and completed two years later. The figures for the following centuries are significantly more daunting, for works by more

J

84

CANON FORMATION IN LATE IMPERIAL CHINA

than four times that number of poets are extant from the succeeding dynasty, the Song, when individual writers produced up to tens of thousands of poems on their own, not to speak of the thousands of writers and hundreds of thousands of poems from the Ming and Qing dynasties. 2 It may well be that the sheer immensity of numbers from the Song onward militated against the achievement of an ascendancy by any later period analogous to that granted the Tang; the implications of this proliferation for the actual status of poetry writing are significant and will be taken up below. Yet the fact remains that, these numbers notwithstanding, the Chinese poetic tradition was dominated for centuries by a canon consisting of a handful of poems by a handful of poets, most of whom wrote during a fifty-year period over a millennium ago. Recent critical discussions of canon formation have engaged in extensive explanations and critiques of such cultural practices, which are of course by no means unique to China. 3 Some critics have asserted, for example, that identifying a small group of canonical works provides a mnemonic device essential to allowing a community to process a long and rich tradition. As Frank Kermode puts it, "Canons are useful in that they enable us to handle otherwise unmanageable historical deposits. They do this by affirming that some works are more valuable than others, more worthy of minute attention. " 4 Others have argued, on a more concrete level, that "there is a continuing need for the social and, in particular, educational uses of the literary canon as some convenient conglomerate of teachable texts. " 5 At the same time, however, the imperative-often moral and political, but certainly intellectual-to recognize such structures as undeniably contingent and constructed, rather than axiomatic, is a compelling one. Jane Tompkins has pointed out, for example, the striking variations among lists of "major" American authors published at different moments over the course of the twentieth century and notes that none of the editors of the anthologies in which these lists appear acknowledges an active hand in shaping a canon. Despite the fact that the instrumentality of their determinations "emerges clearly in the prefaces, where the editors anxiously justify their choices, defend them against other possible selections, and apologize for significant omissions," their "beliefs about the nature of literary value-i.e., that it is 'inherent in the works themselves,' timeless, and universal-prevent them from recognizing their own role in determining which are the truly great works ... as if they themselves had played virtually no part in deciding which authors deserved to be included, but were simply codifying choices about which there could be 'no question.'" 6

Pauline Yu

Thus, both the ongoing critiques of the canon and the benefits of the long historical view have served to bring the mechanisms of its formation to the critical foreground, where it may be seen as "an active and continuous selection and reselection, which even at its latest point in time is always a set of specific choices. " 7 Some of the motives behind such choices were suggested years ago by Northrop Frye, in his discussion of Matthew Arnold's "touchstones." Uncomfortable with one of Arnold's judgments concerning The Tempest, for example, Frye suspected that "some principle is clearly at work here which is much more highly selective than a purely critical experience of the play would be": Arnold's "high seriousness" evidently is closely connected with the view that epic and tragedy, because they deal with ruling-class figures and require the high style of decorum, are the aristocrats of literary forms. All his Class One touchstones are from, or judged by the standards of, epic and tragedy. Hence his demotion of Chaucer and Burns to Class Two seems to be affected by a feeling that comedy and satire should be kept in their proper place, like the moral standards and the social classes which they symbolize. We begin to suspect that the literary value-judgements are projections of social ones. Why does Arnold want to rank poets? He says that we increase our admiration for those who manage to stay in Class One after we have made it very hard for them to do so. This being clearly nonsense, we must look further. When we read "in poetry the distinction between excellent and inferior ... is of paramount importance ... because of the high destinies of poetry," we begin to get a clue. We see that Arnold is trying to create a new scriptural canon out of poetry to serve as a guide for those social principles which he wants culture to take over from religion. 8

Frye's salvo on the illusory impartiality of value judgments in Western literature has been followed by countless others exposing the political, social, and moral motivations behind such judgments as well as the fundamentally contingent and constructed nature of any cultural heritage. Similarly, in the case of China, the demarcation of a poetic canon and the arguments that activity generated can be traced in both critical essays and perhaps even more clearly in anthologies of the past millennium. My concerns in this chapter are with the arguments and methods by which such boundaries were conceptualized and the impulses behind them. The process was a protracted one; its roots may be traced to the Tang itself and its most influential articulation to the Song, but its first triumphs did not come until centuries later. And they rested on motivations that were ultimately as clearly social and pedagogical as they were aesthetic. Embedded on several levels within the elite culture of premodern China, poetry writing during the Tang was an accepted currency of personal,

86

CANON FORMATION IN LATE IMPERIAL CHINA

social, and political exchange. The cultural importance and institutionalization of poetic composition were peculiarly evident in its presence in the civil service examination for the most prestigious degree, the jinshi, or "presented scholar. " 9 The poetry requirement did not in fact last long and, indeed, was assailed intermittently throughout most of the dynasty. As David McMullen has pointed out, 10 more than one critic at the time complained that the ability to write poetry did not necessarily provide a reliable indicator of a person's future administrative competence. The section requiring poetic composition was thus removed at various times during the Tang and Song dynasties, and an interest in testing literary facility gradually lost ground to an emphasis on prose composition, classical exegesis, and policy questions.U From the beginning of the Yuan until the middle of the eighteenth century, poetry writing was not part of the examinations. 12 That a test of poetic composition did survive nonetheless in the most prestigious rite of passage toward civil office during the Tang, however, and one that was often sought even by members of aristocratic families for whom other routes to government positions remained open, attests not only to the vested interest of the educated elite in perpetuating the grounds of its own status but also to the simple, and ultimately perhaps even more important, fact-from the examiners' perspective-that fidelity to strict prosodic regulations was relatively easy to evaluate. 13 Later Chinese critics wondered whether the examination system's valorization of poetry during the Tang had in fact been instrumentai to that dynasty's literary achievement. Comments by the Southern Song critic Yan Yu (ca. r 194-ca. 1245) indicate that this question had already been posed at the beginning of the thirteenth century; Yan's response was unequivocal: "Someone asked, 'Why is poetry of the Tang superior to that of our own dynasty?' The Tang used poetry to select officials; thus many people devoted exclusive attention to it. This is the reason why the poetry of our dynasty does not attain [its level]. " 14 Other critics, however, disputed this assessment. The Ming scholar Yang Shen (1488-r 5 59), for example, argued that "the flourishing or decline of poetry is related to the talent and learning of individuals and is not based on its use as a means of selection by those above. Of the pentasyllabic regulated verse in eight couplets by which Tang officials were selected, those poems written to examination topics that have survived are largely lacking in skill, and those that are known throughout the world are not examination poems." 15 The unimpressive quality of poetry composed to set topics should not surprise us, but its presence on the examination nonetheless fostered con-

Pauline Yu

siderable attention to poetic composition and, in particular, to its rules and structures. The requirements for tonal and semantic parallelism and for balance between emotional expression and scenic description honed during the High Tang conveniently embodied and confirmed assumptions regarding the mutual integration of individual and cosmos articulated in both philosophical and aesthetic discourse. Even more important, their highly systematic nature was conducive to practiced execution, whether at court or in the country, and also offered significant pedagogical advantages in comparison with less regulated forms: rules could be taught, and pedagogical concerns took on increasing importance as the writing public expanded and the basics of cultural literacy could no longer be taken for granted as communal property. The most important early effort to delineate a canon of classical poetry centered in the High Tang must be credited to Yan Yu himself. In his collected remarks on poetry, Canglang shihua (Canglang's remarks on poetry), he provided for the first time a stylistic periodization of Chinese poetic history, 16 which, although criticized quite rightly from many quarters for its vague and unsystematic nature, laid the groundwork for later attempts to develop the theories of genre depletion and revivalism upon which canon formation hinged. Yan also argued strenuously against what he saw as an overemphasis by critics of his age on sterile imitation, although he by no means rejected the importance of methodical study of past exemplars. Rather, employing metaphors borrowed from Chan Buddhism, he singled out as the most suitable objects of study for latter-day writers the master poets of the High Tang, for they-in contrast to those of earlier and even finer ages-wrote in an era when poetry was becoming a self-conscious craft with externalized conventions. And poets like Du Fu (712-70), for example, served as ideal models for emulation not only because of the perfection of their style but also because of their admirable personal character. Yan Yu announced that he had managed "to reduce all poetry to one single set of standards" and settle "once and for all questions which have perplexed countless ages,'' 17 but the task remained for later editors and critics to develop and promulgate his views. The Yuan scholar Yang Shihong took an important intermediary step with his anthology, Sounds of the Tang (Tang yin) (completed in 1344), which he organized according to a revised version of Yan Yu's schema. Praised in a preface by the renowned Yuan poet Yu Ji (1272-1348) for powers of discrimination eclipsing those of his precursors, Yang Shihong articulated his operating principles unambiguously at the beginning of his own preface: in poetry

88

CANON FORMATION IN LATE IMPERIAL CHINA

there is nothing that surpasses the Tang, and "those who know poetry from a myriad generations thereafter all know that Li [Bo; 701-62] and Du [Fu] are its patriarchs [zong]. " 18 Notable here is his exclusive valorization of the Tang, in contrast to Yan Yu, who had still acknowledged the primacy of earlier poetry. Indulging in a rhetorical strategy followed by almost all editors to clear the necessary space for their own projects, Yang then recounts his many frustrations, because he lacks editions of poets' complete works, with anthologies of the preceding dynasties. He faulted five Tang collections, for example, either for not containing a representative sampling of both five- and seven-word regulated verse or for favoring the middle or late Tang periods over the High Tang. He took seven Song dynasty anthologies to task for collecting only late Tang poems, and/or for their unsystematic or undiscriminating selections. He, by contrast, over a period of ten years "copied out High Tang poems by hand and submerged himself in them day and night; he thereupon examined the norms and variations [zheng bian] in their prosody and selected the purest among them, divided them into 'incipient sounds' [shi yin], 'orthodox sounds' [zheng yin], and 'echoes handed down' [yi xiang], and called them Sounds of the Tang." 19 Yang's categories collapse Yan Yu's five period styles into three groupings of extremely unequal size: the first consists of only one chapter devoted to the seventh-century writers known as the Four Bravos of the Early Tang (chu Tang si jie); the second is much larger (six chapters) and includes poets from throughout the dynasty; and the last contains seven chapters of works by lesser talents, monks, and women. The "General Principles" ("Fan li") to the anthology offers explanations for these and other choices,Z0 but however painstakingly justified, Yang's categories are no less curious than Yan's for their unsystematic relationship to history, albeit for different reasons. 21 Nonetheless, Yang Shihong's clear concern with the potential pedagogical impact of his anthology suggests a certain lack of confidence in his readers' training. In the "General Principles," he noted, for example, that in cases of variants he chose what he considered the most secure version and omitted debatable alternatives, in order "not to delude the reader." In addition, although he purposely omitted poems by Li Bo, Du Fu, and Han Yu (768-824), because their complete collections were readily available, he announced that editions of the complete works of the first two arranged by prosodic category would be published subsequently, "as a convenience to the reader." Nevertheless, several later critics took him to task for leaving them out of Sounds of the Tang. 22 His efforts, however,

Pauline Yu

did not go unrecognized; the volume was heavily studied in the early Ming, when new and expanded editions were prepared. There is no question that the most important Ming reader of this Yuan anthology was, like Yan Yu, a native of Fujian province and one of its Ten Talents (Minzhong shi caizi). Gao Bing (I350-I423) compiled a Graded Compendium of Tang Poetry (Tang shi pinhui), in an attempt to correct what he perceived to be the flaws in Yang's framework and to implement and concretely represent the views articulated in Canglang's Remarks on Poetry. 23 His enterprise was at once more self-consciously inclusive, analytical, and critical than any prior collection. In addition to providing a more systematic periodization of the Tang as a whole into what became widely accepted as the four epochs of Early (6 I 8-7 I 3 ), High (71 3-766), Middle (766-83 5), and Late (83 5-906)/4 Gao Bing made it clear that although these demarcations were more chronologically consistent than those of his predecessors, they were also framed by an unambiguous system of values. The many forces motivating Gao Bing's delineation of the poetic canon are evident in both its overall organization and in his explicit statements throughoutthe compendium. He articulated these concerns in the "General Preface" ("Zong xu") to the volume, where he explained the need for such a collection. Having wandered and stumbled about for more than ten years through the uncharted terrain of Tang poetry, he tells us, he decided to fulfill a secret wish to "match minds with the sages of old, culling the glorious blossoms and weeding out the many dispensable ones, to assemble a collection to serve as a gateway and path for the study of Tang poetry. " 25 Providing this "gateway and path" required, first, arranging the 5,769 poems by 620 poets in the work according to prosodic form. 26 Although many Tang and Song anthologies of Tang poetry had used prosody as a standard of selection, choosing to include, for instance, only ancient-style or only regulated verse, Gao Bing's collection was the fi~st to contain both types and then systematically use specifics of prosodic form (type of poem, number of words per line, number of lines per poem) as its first ordering principle (Yang Shihong had arranged only one of his categories in this manner). His example was a compelling one, to be followed by countless other editors of subsequent collections, particularly those with explicit pedagogical aims; indeed, such a foregrounding of formal features makes perfect sense in a volume aiming to provide practical models for would-be poets.

90

CANON FORMATION IN LATE IMPERIAL CHINA

Gao Bing's pedagogical aspirations are further evident in his establishment of ranked categories for the poems in the collection. The tradition of grading individuals-particularly government officials-or works had deep roots going back to the Han,Z 7 and the nomenclature of Gao Bing's categories underscores both the political implications of such evaluations and his esteem for the High Tang. Like the civil bureaucracy, Gao Bing's system has nine ranks, and within each level the poems are grouped by author, with the poets listed in chronological order, primarily by birthdate. For any particular prosodic form, such as five-word regulated verse, all the poems by one author are assigned to one rank, although the order may differ from form to form. The categories correlate, with a few exceptions, with historical periods, the "Pioneers of Orthodoxy" being located in the Early Tang, the "Orthodox Patriarchs," "Great Masters," "Famous Masters," and "Wingmen" in the High Tang, the "Footstep Followers" in the Middle Tang, and the "Orthodox Innovators" and "Lingering Echoes" in the Late Tang. The poets labeled as "Overflow Currents" fall outside the grading system, for they are individuals, like women, monks, and foreigners, for whom collections and biographical information are too incomplete to make a reliable judgment. 28 As with Yan Yu, for Gao Bing canonical values were as ethical, in their links to personal character, as they were aesthetic. In discussing canon formation in early modern Europe, Hans Ulrich Gumbrecht has observed that from antiquity until the eighteenth century, "we still encounter names of authors rather than titles of works. The discursive constitution of the canon as a series of authors' names can be associated with the fact that the imitation of exemplary action, objectified in texts and constantly being reproduced in new texts, created the perspective of relevance that made the canon an element of education and socialization, for competencies of action are bound to people. " 29 Gao Bing's reference in his preface to his secret desire to communicate with the ancient sages of poetry, selecting the glorious and weeding out those without merit, points to similar aims. This search for cultural ideals was very much rooted in the intellectual and political climate of the early Ming, consumed as it was with the recuperation of a lost national culture and the need to validate its claim to power. Thus the founding emperor "and his advisors made great efforts to utilize notions of legitimacy, based upon claims of cultural and political orthodoxy, as means to bolster political, social and economic control and overall reconsolidation and unity. " 30 The government sponsored a number of other codifying projects as well and eventually adopted Gao's version of Tang poetry as the officially sanctioned one.Jl Tang literary culture,

Pauline Yu

91

after all, appeared to have institutionalized more dramatically than that of any other era the mutual implication of self and society, the links between the individual and the body politic that informed the identity of the elite as upholders of culture and the imperial order. The High Tang poet's face was at root a public one; there was an intensity and grandeur to this stance and to the obsessiveness of the issue of service itself. During succeeding dynasties like the Song, this grandeur inevitably seemed to have become attenuated as occasions for poems multiplied and poetry's increased dispensation to record the minutiae of the quotidian produced more intimate, casual, and apparently trivial pieces. Not only were later poets drawn from a wider social spectrum, with their works more numerous and more effectively circulated and preserved, but poetry itself had lost both its privileging role in the civil service examination and its privileged position within literati life; it was now one of many activities such as painting, calligraphy, and the writing of prose and fiction available as a means for self-definition. In promulgating a literary archaism that called for poetry to be written in emulation of High Tang models, Ming literary critics-many of whom held high offices within the government-were affirming, or perhaps wishing into being, other cultural and political homologies between the two dynasties as well. Equally important was Gao Bing's response to a concern already evident in both Yan Yu and Yang Shihong-a concern that the Tang did not have to worry about-teaching their readers how to be poets. We now recognize that the creation of what for decades was regarded as the standard, "canonical" college literature curriculum and the evolution of a New Critical methodology, for example, can be fully understood only in the context of the expanded and more socially diverse student clientele in British and American universities in the twentieth centuryY Similarly, the Ming "archaic language" (guwen ci) school, linked in particular with the two groups of scholars and statesmen known as the Former and Latter Seven Masters (qian hou qi zi), took root in an era of broadened access to prestige and power, owing to what some observers regarded as a simplified civil service examination and to the burgeoning and increasingly diversified economy. The archaists' schematic reduction of Chinese literary history to a handful of models to be studied and emulated exerted a powerful appeal on those whose social pedigrees were of fairly recent vintage. 33 Their revivalist theories dominated the cultural sphere more effectively than those of any other "school" before or after, 34 and Yan Yu and Gao Bing provided them with the conceptual and chronological framework on which their project was based.

92

CANON FORMATION IN LATE IMPERIAL CHINA

One particularly influential anthology that replicated Gao's configuration of the poetic canon (forsaking, however, his exclusive focus on the Tang) 31 was the Correct Edition of Ancient and Modern Poetry (Gu jin shi shan) edited by Li Panlong ( r 5 14-70 ), one of the Latter Seven Masters. Of Li's selection of 2,229 poems from ancient times through the Ming, slightly fewer than one-third are pre-Tang, one-third are from the Tang, and over one-third from the Ming (whose poets were by this time writing with their eyes on Tang models). Notably unrepresented are the Song and Yuan dynasties. Moreover, of the 742 Tang poems, which are arranged according to prosodic form, well over two-thirds are examples of regulated verse and quatrains, and 309 are by High Tang poets. 36 Li's anthology did not lack for critics: some took issue with his apparent disparagement of Tang ancient-style verse, 37 others deplored his consignment of Song and Yuan poetry to oblivion, and still others found his selection of Ming contemporaries self-serving and overly generous to his intellectual allies. Ming and Qing critics alike objected to the well-known dictum promulgated by the Former and Latter Seven Masters as a group that "prose must be like the Qin and Han; poetry must be like the High Tang." At the same time, however, Li's seven juan of Tang poetry were highly enough regarded to be extracted and published under his name, posthumously, as Selections of Tang Poetry (Tang shi xuan), a volume that circulated widely, in seventeen editions, in both China and Japan. To be sure, other Ming anthologies of Tang poetry embraced standards of selection different from Li Panlong's. 38 However, his Correct Edition of Ancient and Modern Poetry exerted far-reaching influence on subsequent anthologies. It served, for instance-along with Gao Bing's Orthodox Sounds of Tang Poetry (Tang shi zheng yin)-as a basis for Tang Ruxun's much larger (so juan) Explications of Tang Poetry (Tang shi jie), printed in r6r5 although more widely circulated in a r659 edition and then reedited at the beginning of the eighteenth century by Wu ChangqiY These two collections together left an imprint on later anthologies evident not only in their arrangement by prosodic form but more importantly in their specific choices of poems. Ling Ziliu has conveniently, for our purposes, provided a tabulation of appearances of poems by Li Bo, Du Fu, and Wang Wei (7or-6r) in anthologies from the Tang dynasty through the twentieth century. 40 Although these three poets are now the most widely anthologized High Tang writers, such was not the case in the Tang, 41 nor do the selections of Song and Yuan anthologies suggest any clear consensus on the corpus of any one of them. Ling's charts indicate, however, that when a group of

Pauline Yu

93

approximately sixty anthologies spanning thirteen centuries is scrutinized as a whole, for each poet and within each of seven prosodic categories (pentasyllabic and heptasyllabic ancient verse, regulated verse, and quatrains, and pentasyllabic regulated couplets), only one, two, or at most three poems appear with a significantly impressive frequency, and in some cases twice as often as the next poem on the list. And with but one exception for each poet, only the anthologies of Li Panlong and Tang Ruxun include all these chestnuts. 42 In addition to decisive selections of particular poems, Li Panlong's general valorization of Du Fu, Li Bo, and Wang Wei as the three preeminent poets of the Tang continued to be supported in both critical discourse and poetic anthologies over the next few centuries, even by those who disavowed Ming archaistic principles. In the preface to his collection of Tang poetry, for example, the Qing scholar Shen Deqian (1673-1769) called attention both to what poetry teaches (which he sums up in the four-word phrase first associated with the Classic of Poetry, "geniality, mildness, honesty, generosity" [wen rou dun hou]) and to how poetry should be taught. He opened by pointing out how varied Tang poetry is: there are both good poems and bad, and one must take care not to begin a course of study with the latter. To do so would be "like heading south to go to You and Ji or heading north to go to Min and Ao-you'll never get there. Thus distinguishing between what to discard and what to select, so that the minds and eyes of those coming after will have standards and not be deluded, is the sole responsibility of editors of poems." However, contending schools since the sixteenth century (i.e., the heyday of the archaic language school) have established no reliable principles, and the resulting degeneration of poetry's didactic lessons may well be attributed to the errors of such scholars. "Now an editor of poems is obliged to be able to discard [the songs of] Zheng and retain the Elegances,43 but one who misuses this function will, on the contrary, cause people to discard the Elegances and rush en masse to Zheng. Thus how can the distinction between what is to be discarded and what is to be selected not be considered weighty?" 44 Shen assured his readers that he, by contrast, had taken proper care to scrutinize the language, prosody, and meaning of the Tang corpus in making his selections, so that "those traveling by land will have cart and horse and those traveling by water will have bark and oar." He then chronicled, in his "General Principles" to the collection, historical oversights going back to the Tang (already a familiar litany of failures) evident in the inadequate representation of both Li Bo and Du Fu in other anthologies, stated that even Gao Bing's broad selection in his

94

CANON FORMATION IN LATE IMPERIAL CHINA

Orthodox Sounds "does not exhaust their riches," and announced that his collection (echoing Yang Shihong) will "regard Li and Du as patriarchs [zong]," a statement confirmed by a count of the poems he included. 45 Although a long-standing cultural penchant for binary formulations led many other critics to echo this pairing of Du Fu and Li Bo as complementary ideals, Wang Wei had his partisans as well. Not only was he placed at the head of such collections as Lu Shiyong's Tang shi jing (Mirror of Tang poetry), but Wang Shizhen's (I634-I7I I) popular Samadhi Collection of Tang Worthies (Tang xian sanmei ji) also explicitly foregrounded the works of Wang Wei and other poets in his "school. " 46 His anthology in fact contains no poems by Li Boor Du Fu but has I I2 poems by Wang Wei alone, well over one-fourth of the poet's total corpus. All these judgments coalesced in the well-known mid-eighteenth-century selection, the Three Hundred Tang Poems (Tang shi san bai shou), where no other poets are better represented than Du Fu (30 poems), Wang Wei (29 ), and Li Bo (27}. 47 Needless to say, challenges to this canon were mounted periodically and systematically, indeed almost as soon as its outlines became clear. Shen Deqian's explicit declaration of veneration for Li Bo and Du Fu suggests that contrary opinions also flourished. Esteem for the High Tang, after all, had simply displaced a passion for late Tang poetry going back to the Song, which enjoyed a significant resurgence in the seventeenth century. Thus Feng Shu and his brother Feng Ban (I 614-7I ), who were associated with the Yu Mountain school led by Qian Qianyi (I582-I664), issued a new edition of the early tenth-century Collection of Talents (Cai diao ji), compiled by Wei Gu, with new commentaries employing the qi cheng zhuan he mode of analysis made popular in discussions of bagu wen (about which more later}. 48 Similarly, the Jin dynasty anthology spuriously attributed to Yuan Haowen (II90-I257), The Drums and Pipes ofTang Poetry (Tang shi gu chui), a collection of approximately 6oo heptasyllabic regulated verses drawn primarily from the late Tang, was reprinted with a new preface by Qian Qianyi. 49 Other new anthologies devoted to heptasyllabic regulated forms of the same period were also published at this time. Jin Shengtan (I6Io?6I}, for example, compiled a selection of poems and commentaries in I 66o that was issued posthumously by his son Jin Yong as A First Volume of Poems by Geniuses of the Tang (Tang caizi shi jia ji). According to the preface to the collection (whose authenticity is somewhat in doubt), this volume of 59 5 poems had been assembled in response to a plea from his

Pauline Yu

95

son for an elucidation of Tang seven-syllabic regulated verse; the vast majority of the 145 authors included are ninth-century figures. Written apparently also at Yong's request, the preface itself is a lush exposition of the ubiquity of poetry in all natural emanations and the universal penchant for poetic self-expression in all humankind. 5° Jin described regulated verse as a natural evolution in the history of poetic forms, in fact their pinnacle, and he employed a number of images and analogies from nature to emphasize this point; he likened the four couplets of an eight-line poem, for example, to the progression of the seasons from spring through winter. 51 What proved particularly controversial in this volume, however, was ]in's method of explication, which involved dividing each poem into two halves and commenting exhaustively on each; I will return to this below. Another new seventeenth-century collection devoted exclusively to heptasyllabic regulated verse of the Tang was the Glorious Blossoms of Tang Poetry (Tang shi ying hua), one of the many anthological endeavors of the scholar Gu Youxiao (Maolun) ofWujiang. Gu arranged his authors in chronological order according to Gao Bing's four periods and provided considerable background information on each poet from the Tang dynastic histories and other critical sources. As he explained in the "General Principles" to the anthology, such data will elucidate what he felt are the actual historical incidents and political situations inspiring the composition of the poems, thereby "broadening the learning of all literati and, moreover, fulfilling the intentions of the illustrious and virtuous explicators of poetry." The collection is a large one, containing the majority of the over 3,ooo examples Gu had considered; he provided an explanation for the collection's breadth: Contemporary styles may be divided into Early, High, Middle, and Late [Tang]; human characters may be differentiated as upright, licentious, honest, and insincere; poetic modes may variously be virile and lofty, frivolous and indolent, intricate and elegant, and plain and impoverished. Each has become a school of its own. I have selected and compiled them in their entirety to afford poets of a thousand generations thereafter an inspired understanding. Whether by following the current to trace the source or by going from the roots to reach the branch tips, personal nature will be seen to have a basis, and poetic style will truly have an origin. 52

A preface to the Tang shi ying hua by Qian Qianyi (dated 1657), which consists essentially of an extended diatribe against Yan Yu, provides insight into the critical temper of the times. Qian began by castigating Yan Yu and Gao Bing for developing an overly precise schema for periodizing the Tang that makes it difficult, for example, to categorize

96

CANON FORMATION IN LATE IMPERIAL CHINA

poets whose lives happen to cross the artificially imposed datelines. Furthermore, an exclusive valorization of the High Tang is patently absurd and denies the continued development of new and praiseworthy forms from the very beginnings of the tradition. Turning to Yan's use of Chan Buddhist terminology to characterize styles, Qian argued that many of his key concepts have been used erroneously. He then took issue with Yan's oft-quoted position that poetry eschews discursive language and rational argument by citing passages from the Classic of Poetry, the undisputed fount of the tradition, that are clear counterexamples. "Someone whose eyes are filmed over," he concluded, "will distinguish empty illusions; someone suffering from fever will point out spectral beings to his side. Mr. Yan's discussion of poetry is also an illness of film and fever, and his disease has been transmitted and infected generations after him. They lift up eyes fully clouded over with Mr. Yan's cataracts and speak words replete with Mr. Yan's delirium." The only cure for these maladies, he declared, is Gu Youxiao's anthology. 53 Qian's fulminations against Yan Yu, Gao Bing, and the Ming archaiclanguage school appear in several other essays as well (such as his preface to the Tang shi gu chui) and represent a movement on the part of many late Ming and early Qing critics to ridicule the obsession with High Tang models, lavishly praise poetry from the Song over that from the Tang, or reject the canonizing process altogether, on the grounds that dynastic eras bear no necessary relationship to the quality of poetry or that slavish imitation of archaic models can only inhibit the free expression of a poet's own nature. 54 In addition to resurrecting an interest in other periods of the Tang, such critics also began to turn in earnest to the virtually uncharted terrain of Song dynasty poetry as well. Such path breaking scholarly endeavors, as Benjamin Elman has pointed out, were to a large extent impelled by the contemporary development of increasingly sophisticated methods of evidential research and textual study. 55 The Kangxi reign period in particular witnessed a number of such efforts, of which the best known are the Collection of Song Dynasty Poetry (Song shi chao), a selection of works by a hundred poets undertaken by Lii LiuJiang (1629-99) and several other scholars (preface dated 1671), 56 and the Anecdotal Events Behind Song Dynasty Poetry (Song shi ji shi) compiled by LiE (16921752). Yet another index of the prevalence of this trend can be seen in a preface by Jiang Chenying to Wang Shizhen's Samadhi Collection of Tang Worthies, which notes that the anthology had in fact been compiled in opposition to this current of interest in the Song, with the aim of distin-

Pauline Yu

97

guishing the High Tang style from that of the other three periods and Tang poetry in general from earlier and later forms. 57 Such arguments could not avoid being predicated on the conceptualization of difference established within contemporary critical discourse, which characteristically invoked a binary opposition between two hypostatized constructs, the Tang and the Song. 5 8 Perhaps as a mediating force, the Yuan dynasty anthologyprimer of Tang and Song penta- and heptasyllabic regulated verse forms compiled by Fang Hui (1227-I307), Ying kui Iii sui, completed in 1283which had attracted little attention during the Ming-entered the critical arena full-blown at this time. Highly debated by schools that were obsessed with disputing, among other issues, the relative merits of Tang versus Song poetry, it was a lightning rod of controversy throughout the dynasty and was re-edited-by both friend and foe-over ten times. 59 The preface attributed (though not altogether securely) to Fang Hui briefly but cogently articulates his premises and goals: "The essence of letters is poetry, and the essence of poetry is tonal regulation. What I have selected are the forms of poetry [shi ge]; what I have annotated are critical discussions of poetry [shi hua]. If the student seeks them out, he will reach their core. " 60 Throughout the anthology, Fang Hui's comments and evaluations set out a critical and pedagogical program associated with the Jiangxi school, 61 and a specific concern with an appropriate balancing of emotional (qing) with scenic (jing) elements in a poem. 62 Within the context of what had by then become a consistent obsession with the teaching of poetry, and with anthologies as a tool toward that end, the intense interest in the Ying kui Iii sui makes eminent sense. Moreover, advancements in two-color printing were marshaled to highlight his critical opinions, and lessons, in editions printed during the Qing. 63 In addition to this renaissance of Song shi poetry, Qing dynasty scholars also successfully revived and rehabilitated other poetic forms, such as the song lyric (ci), that had been discredited from their origins by associations with eroticism, excessive sentiment, and denizens of the entertainment quarters. 64 Many of the most accomplished practitioners of this genre were women, and recent scholarly research has begun to unearth what appears to be an extensive surviving body of writings by women-over 2,ooo anthologies of poetry alone-that did not always merit the attention of mainstream critics and bibliographers but clearly enjoyed a wide and appreciative readership. 65 We might argue, of course, that such a pluralism of critical interests could thrive precisely because a High Tang poetic canon had been securely estab-

98

CANON FORMATION IN LATE IMPERIAL CHINA

lished. Whatever the case, despite the lively contention of critical interests and allegiances during the Qing, the fact remains that the pedagogical concerns articulated by Yan Yu and critics after him remained paramount in importance-particularly, perhaps, in view of the increasing elusiveness of success on the civil service examinations as the numbers of eligible candidates continued to burgeon. 66 Certainly, the disciplined and regulated style perfected in the High Tang could be incorporated into a standard curriculum with considerably less difficulty than doctrines of sheer inspiration, expressionism, or iconoclasm, and such utilitarian concerns played a crucial role in this process. Following Gao Bing's example, one anthology after another highlights this important educational mission. Shen Deqian wrote of his collection, for example, that "this edition will aid the student of poetry to start his journey" (literally, "remove the skid" [fa ren] to allow a cart to proceed forward)Y The preface to Sun Zhu's (r7II-78) Three Hundred Tang Poems is even more emphatic: Throughout the world when children begin their studies, they are taught the Poems of a Thousand Masters [Qian jia shi, a primer popular since the Ming]; because it is easy to memorize and chant, it has circulated widely and not been discarded. But its poems have been quite randomly assembled, with no distinction between the skillful and the clumsy. Moreover, it only includes the two forms of five- and seven-word regulated verse and quatrains, and Tang and Song poets are freely intermingled within a remarkably eccentric format. Thus I have focused on the works of Tang poetry that have pleased all tastes and selected the most essential examples. There are a pproximately ten examples of each poetic form, over three hundred in all, collected to form a volume. This will be a textbook for home and school that children can study and that will be remembered even in old age. Will it not be on a par with the Poems of a Thousand Masters? According to the proverb, "After thoroughly studying three hundred Tang poems, even someone who can't chant poetry will be able to do so." I offer this volume as a test. 68

Although Sun Zhu explicitly addressed his volume to a youthful audience, his concerns were shared by other equally popular Tang anthologies compiled during the eighteenth century. All these volumes, like Shen Deqian's, follow the lead of the Ming archaists in foregrounding prosodic form as the fundamental principle of arrangement, and all of them showcase the luminaries of the High Tang. Wang Yaoqu's Combined Explications of Ancient and Tang Poetry (Gu Tang shi he jie; completed 1732), for example, reveals the influence of Li Panlong's anthology in his many duplications of selections from the poems of Wang Wei, Li Bo, and Du Fu. 69 Wang employed the punctuation dot method that had come to be associated with collections of bagu wen to mark meaningful phrases, and

Pauline Yu

99

his commentaries on individual poems, like those in Jin Shengtan's anthology, similarly divide the poems into two sections and walk the reader through the generation of meaning in a painstakingly clear-cut way. 70 Finally, another Qing anthology, the Combined Selections of Tang Poetry (Tang shi he xuan) of Liu Wenwei, was compiled at approximately the same time as the Three Hundred Tang Poems and enjoyed a comparably large circulation, particularly through the nineteenth century. A slightly larger collection (3 68 poems versus the 321 eventually included in the Three Hundred Tang Poems), it reveals the influence of the Ming archaist movement in the critical comments that preface each section-most of which are drawn from the writings of Yan Yu, Gao Bing, and the Former and Latter Masters or critics associated with them-as well as in the very division of the volume's contents according to prosodic form. Two-thirds of the poems are different from those in Sun Zhu's anthology, primarily because Liu included almost twice as many poems by Du Fu (85, as opposed to 30); because his selection of poets was much broader, with many instances of a single work by a poet; and because he added a section of regulated couplets (pailii). Nevertheless, the preponderance of poets are High Tang figures, with the three most amply represented being Du Fu, Li Bo, and Wang Wei; the overlapping poems constitute what has become a recognizable canon; and regulated verse forms clearly dominate the collection. 71 The development and consolidation of a relatively self-contained, identifiable group of canonical texts and of methods of communicating their technical intricacies addressed the need to provide a systematic and codifiable poetic curriculum that could be taught and mastered. As has been suggested, this was related to the increasing breadth and diversity of the poetry-writing class. 72 High Tang poetry in both form and substance proved particularly appropriate to these ends. As we have seen, increasingly heavily annotated and punctuated anthologies proliferated, along with manuals providing simple and specific pointers on the writing of poetry. 73 Although the reinstitution of a poetry section to the provincial and metropolitan examinations did not occur until 1757, where it remained until 1900, perhaps the most important fact about its presence on the examination was not when or for how long that was actually the case, but that it had once been so. Despite the relative briefness and inconstancy of its existence as an item to be tested, it played a powerfully appealing role in the construction of a literati self-image that was related to the even larger issue of the status of writing in general as a personal, social,

IOO

CANON FORMATION IN LATE IMPERIAL CHINA

and political form. Like literature itself, poetry was continually buffeted, as noted earlier, by intellectual challenges to its utility and was always susceptible to charges of triviality or preciosity. Yet somewhat paradoxically, to the extent that it was not part of the examinations for seven centuries, poetry could simultaneously take refuge in that distance as a guarantee of its aesthetic incorruptibility and encourage the belief that its implication in the political sphere was both appropriate and necessary. The notion that the formal discipline required of poetic compositionwhich consisted, as annotations to the anthologies discussed above indicate, not only of fidelity to prosodic rules but also of an ability to move successfully between one section of a poem and the next and to balance objective and subjective elements-was a concrete index of ethical and political leadership survived to the extent literature itself could succeed in making those claims. Poetry's quasi-mythical status as such an index afforded it the space within which it could perpetuate the illusion for itself and protected it against being discredited by being put to the test. Its success in straddling these boundaries may be measured, for example, in the flourishing of local poetry societies (she) from the Yuan onward. As part of a larger burgeoning of such social formations in the late Ming and Qing, these poetry clubs came increasingly to be seen as sites of resistance and threats to state hegemony; at the same time these associations often developed quite formal procedural protocols mimicking those of the state, including competitions reminiscent of the civil service examinations themselves. 74 We might also argue that poetry's function within a political context survived, spectrally, in the "eight-legged" examination essay (bagu wen) itself, which was required in the essays on the Four Books and Five Classics during session one of the provincial and metropolitan examinations from the late fifteenth century on. Although the essay was not, obviously, a poetic form governed by patterns of tonal alternation, line length, or rhyme, in other ways it can be regarded as poetry's sibling and substitute. 75 Space prohibits an extensive exploration of these affinities, but they rested principally in assumptions concerning the ideal speaking stance of the writer, the formal features of the two genres, and the critical strategies devised to analyze them. One obvious difference between an examination poem and a personal one was the fact that the author of the former was ostensibly writing to the subject set before him rather than directly of himself. The task at hand was to recognize the locus classicus if a citation from a text had been given as the topic and/or to marshal other relevant utterances on the

Pauline Yu

IOI

same theme. At the same time, of course, such impersonal manipulations of the words of others had to be conducted in a coherent and distinctive way/ 6 The examination essay required a similar movement between multiple subject-positions, for a crucial section of the essay mandated that the candidate temporarily abandon his own voice to speak in that of the sage-author of the passage on which he was writing (dai shengxian li yan), "a rhetorical posture supported by the mandatory use of classical exclamatory particles to evoke the sage's unalloyed earnestness. " 77 Much note has been taken of the links between such a requirement and the development of dramaticliterature, 78 but the extent to which such a deflection of the subjective speaking voice was a feature not only of examination poems but also of the central sections of regulated verse in general should also be recognized. Although scores of critics considered bagu wen a less rigorous form, and hence a less rigorous test, than poetry or caviled against its corruption of prose style, many others discerned important structural similarities among the examination essay, regulated verse, and what became the typica) examination poem.79 These centered on the fact that the essay had eight sections and/or eight subsections or "limbs" (albeit variously and loosely identified) and the examination poem eight rhymes or couplets, thus a simple doubling of the length of standard four-rhyme, eight-line regulated verse. The same terms developed for elements of one were applied by analogy to the other. Both were of course written in response to set themes, and similar rules governed the rhythms of exposition, the points at which candidates either had to mention the topic or refrain from doing so, and the sections-as just mentioned-in which the speaking stance shifted. Critical analysis of bagu wen could emphasize either its bipartite structure (first half and second half) or its dynamic progression from initial proposition to development to turning point to resolution, summed up in the formulation qi cheng zhuan he (beginning, continuing, turning, uniting). Although this four-character phrase is probably most familiarly associated with structural analysis of the examination essay, it was first developed as a means of discussing poetry, by the Yuan dynasty scholar Yang Zai (r27I-r323) in his "Essential Methods of Regulated Verse" ("Lii shi yao fa"). 80 Yang listed the four terms and then labeled the four couplets of a regulated verse that carry out their movement as, in order, "cracking the topic" (po ti), "jaws couplet" (han !ian), "neck couplet" (jing !ian), and "unifying verse" (jie ju). Versions of these terms were later applied to sections of bagu wen. Additionally, either essays or poems could be broken down into two halves for the purposes of analysis. Jin

!02

CANON FORMATION IN LATE IMPERIAL CHINA

Shengtan, as mentioned earlier, provoked considerable controversy for implementing this method in his Tang caizi shi, as the letters collected and appended to the anthology suggest. His preface to the anthology employs a plethora of images in an attempt to argue for the naturalness of the qi cheng zhuan he progression. 81 However, his correspondents were evidently not ~ntirely convinced, and Jin repeated and varied his claim, likening the bipartite sectioning of regulated verse, for example, to inhalation and exhalation in one letter, 82 and, in another, simply affirming that the naturalness of the regulated verse form had become apparent to him precisely through this mode of analysis. 83 As we have seen, all these terms, or variations of them, were used by critics to analyze both regulated verse 84 and the eight-legged essay.SS In addition, another method that had first been employed in thirteenth-century editions of poetrythe use of commentary and punctuation dots for emphasis-had in the meantime become associated with examination essay criticism, evident in "the practice of examination candidates marking up examples of essays with punctuation marks and comments as part of their education in how to write essays and the fact that part of the grading system of the examination papers involved the adding of emphatic punctuation and comments by the examination officials. " 86 These practices, enhanced by developments in multicolor printing, were then in the Qing transferred back to anthologies of poetry. Although many of these comparisons focus primarily on the examination essay and the examination poem (which, as noted earlier, took on a standard form twice the length of an eight-line regulated verse), it is nonetheless clear that all poetic forms were implicated in discussions of the latter. Thus when a section testing poetic composition was restored to the provincial and metropolitan sittings of the civil service examination in 1757, not only did a spate of anthologies of examination poemscrowding the already groaning shelves filled with collections of examination essays-follow shortly, 87 but publication, or republication, of many of the Tang anthologies I have discussed surged as well. Both the Three Hundred Tang Poems and Liu Wenwei's similar Combined Selections of Tang Poetry, for example, were compiled within a decade of the addition of poetry to the examination. Shortly before his death, Shen Deqian issued a second edition of his Discriminating Collection of Tang Poetry (Tang shi hie cai ji) in 1763, which, he noted at the beginning of the second preface, includes the works collected by Wang Shizhen in his popular Samadhi Collection of Tang Worthies and many others as well. Shen listed the titles of some of the poems added to this new edition, with the most

Pauline Yu

103

significant addition being a group of pentasyllabic examination poems (shi tie). "In the former selection," he explained, "these were scarcely to be seen. Now that they are required for the examination, I have collected and selected some excellent pieces that can be handed down as standards, so that they may show the way to those about to enter the gates of the palace [i.e., sit for the examination] each year." 88 Wang Shizhen's collection in turn was reprinted with new annotations by many different hands; Huang Peifang, in his edition of r8os, for example, explained that he had added commentary and punctuation emphases in order to make the difficult concept of samadhi or meditative concentration embodied by the poems apprehensible by the beginning student. 89 And finally, the reinstitution of the poetry section inspired the compilation of other Tang anthologies focused on the specific types of composition to be tested. 90 As the increasingly single-minded attention to pedagogical efficacy and methodological clarity in this group of texts should suggest, poetry in the context of the Qing civil service examination filled needs rather different from those it may have met in more private contexts, ones commensurate with the increased size and disparate levels of training of those sitting for the examinations. Alexander Woodside has pointed out that, with the examination system at its zenith during the Qing, "the numbers were causing China to embark on a transition from a society in which slower, more culturally reflective forms of literacy were dominant, to ones in which faster, more occupationally instrumental forms of literacy prevailed. " 91 For much of this population, overtly utilitarian anthologies such as those just mentioned might well have been the only literary fare consumed, a situation that was true of forms of writing other than poetry as well. "One provincial educational director," Woodside continues, "in 1764 calculated that sheng-yiian no longer read a work like the Li chi [Record of rituals] at all, but merely read popular digests that accounted for barely half of its contents. To real literati, this was a shock. " 92 Literati views on the value of poetic composition, moreover, continued to reflect the long-standing ambivalence about certain forms of writing in general. Wu Hongyi tells us that the status of poetry remained elevated throughout the Qing despite the many vilifications of the Ming archaists who had been most recently responsible for promoting it. Benjamin Elman, however, reminds us that various factions continued to harbor suspicions regarding poetry's threats to moral behavior or its at best tangential relationship to the really serious concerns of statecraft. More than one Qing scholar, for example, expressed dismay on being recognized for his poetic accomplishments, although, by the same token, the very fact that poetry

104

CANON FORMATION IN LATE IMPERIAL CHINA

brought such acclaim suggests that its value had not been totally discounted.93 Be that as it may, disdain for contemporary poetic activities by no means necessitated a rejection of the legacy of the past-quite the contrary. As institutionalized in the civil service examination, poetic composition could draw upon discourses of significance produced a millennium before, even as its practices and contexts had become more diverse and instrumental. Pedagogical interests joined forces with aesthetic and political concerns to establish a canon of poetry drawn from the eighth century of the Tang dynasty that remained stable despite varying and even contradictory interests. What persisted, however, was a conviction that cultural values were located most securely in a dim and distant past. Robert Weimann has written that "as a cultural institution, a literary canon may be defined as a publicly circulating usable body of writing which, by definition, is held to be as much representative of certain national or social interests and traditions as it is unrepresentative and exclusive of others. In fact, the very representivity of this privileged body of writing appears as a sine qua non for its function as a tradition or heritage, for receiving and projecting patterns of social, cultural, and national identity. " 94 Discursively constituted as the product of a circumscribed elite working with a shared poetic language and within relatively defined social and topical boundaries, the High Tang style offered a mythic image of cultural unity that must have seemed haunting by its very distance from the complexities of a much later reality. Both as a powerful icon from the past and as one lending itself to a course of systematic training, it afforded late imperial China a space within that myth.

CHAPTER

4

Salvaging Poetry: The "Poetic" in the Qing Stephen Owen

0The following is a passage from Wei Xi's (r624-8o) "Account of the Reconstruction of Level Mountain Hall" ("Chongjian Pingshantang ji"). The original hall, constructed by Ouyang Xiu just outside Yangzhou, had been celebrated by many Song literary figures and remained forever associated with Ouyang Xiu. It had several times fallen into ruin and been rebuilt; Wei Xi describes its most recent rebuilding by the seventeenth-century Qing governor Jin Zhen.

u

There was a terrace and the hall, behind which was a building of several stories, with two wings outspread, where offerings could be made to the spirit of Ouyang Xiu. It was an airy and spacious place, with a splendid beauty, giving access to thousands of scenic views. I don't know how it compares to Ouyang Xiu's building back then, but from it we can infer His Excellency Jin Zhen's intentions regarding the moral education of the folk and the promotion of virtuous customs. The customs of Yangzhou are, in fact, a melting pot for people from every corner of the land. It is a concourse for the fish trade, for the salt trade, and for money. Former officeholders and powerful families move here. Therefore its people have a great lust for gain, love parties and excursions, procure singers and pursue courtesans, wear fine clothes and live for the pleasure of the moment in order to show off their splendor to others. Only the most worthy among them still has any interest in things of culture. Once His Excellency had restored the ruin, he often drank and composed poetry here with members of prominent local families and those who came to visit him. This practice had the following result: what these people heard and cast their eyes on here gave them a joyous admiration for both the landscape and things of culture. Soon every family and household was reciting poems, until the way of literature and of the Classic of Poetry gradually changed the atmosphere of money and horse-trading. 1

!06

THE "POETIC" IN THE QING

Wei Xi praises Jin Zhen for re-enacting in the Qing present one of the essential purposes of poetry as set forth in the "Great Preface" to the Classic of Poetry, but something has gone terribly wrong. The distance between this passage and the "Great Preface" is a measure of the distance between classical poetry in the medieval world and in the seventeenth century. As stated in the "Great Preface" and developed in the standard exegeses through the Tang, one function of the Classic of Poetry, and by extension of poetry in general, was regulatory. By reciting the poems, one internalized their moral standpoint, and genuine feeling could be expressed in what was both a natural and a civilized way. 2 Through poetry, emotion, always threatened by dangerous repression or equally dangerous excess, found a legitimate outlet. By rebuilding a famous cultural monument near Yangzhou, Jin Zhen intended to guide the deracinated inhabitants of that notoriously sensual and commercial city away from their lusts, their ostentation, and their crass economic concerns toward a higher cultural experience. He invited them to his rebuilt hall where they could absorb its historical memories and the uplifting beauties of the Jiangnan mountainscape visible in the distance. By composing poetry, they internalized this valuable experience and enjoyed the benefits of its proper expression; they were transformed, continuing to compose poetry on returning to the corrupt city, which soon felt the general effects of their beneficent influence. Thus did poetry fulfill its ancient Confucian function of civilizing the folk and changing customs for the better. I imagine few readers would find Wei Xi's account anything more than a polite and sociable fiction. Just beneath the surface of this idyllic picture, we see clearly the role of poetry in social climbing. Members of the great families that populate that most iniquitous city of Yangzhou were interested only in making money, whoring, and generally having a good time-the reasons they moved to Yangzhou in the first place. The privilege of an invitation to a party with the governor involves the composition of poetry; it is a mark of social status. And the great families of Yangzhou learn to supplement the simpler pleasures of the flesh with the more refined pleasures of pretense. The "Great Preface'"s vision of humanity finding normative expression for its passions in poetry becomes instead a vision of assimilation of the urban bourgeoisie to the literary elite: "horse traders" and salt merchants are being uplifted by participation in the rituals of elite culture. What I have done here is to contextualize Wei Xi's account of poetic practice. Such contextualizing within a social whole is hard to avoid when

Stephen Owen

I07

reading poetry or reading about poetry in the late imperial period. 3 Although, in actual fact, the practice and study of poetry were far more restricted in earlier periods than in the seventeenth century, we easily grant poetry a centrality and importance in the Tang that we cannot grant in later periods. If poetry served as a means for advancement during the Tang, we consider such a social ground for poetry a factor contributing not only to the artistic legitimation of poetry but also to its very vitality. In the seventeenth century such social contextualization of poetry makes it seem somehow false, even ludicrous. Perhaps it is only Wei Xi's sneering reference to "horse-trading." Perhaps we are simply prejudiced against the seventeenth-century classical poetaster. Or perhaps something really changed.

What Happened? What happened to poetry and classical literature in general in the late imperial period is one of the most difficult questions of Chinese literary history. We need not too carefully define "late imperial," since we are considering an event that began in the Song and continued into the twentieth century, when, despite continued wide-scale composition, the possibility of classical poetry's continued viability as an art has ceased to be an issue. In the Song, classical poetry seems to possess somewhat less cultural importance than it had in the Tang, but in the Yuan, Ming, and Qing classical poetry entered a peculiar shadow life. Since poetry was supposed to represent genuine human feeling-" lust and rancor" kept in decent bounds, as Qian Qianyi ( r 582-r 664} put it-critics could only conclude (as many did) that it was not poetry itself that had declined but humanity. 4 The phenomenon of late classical poetry is peculiar. More poetry was written than ever before. More criticism and theory was written about it. There was immense' variety in this poetry, and it went through changes as profound as had occurred in the Tang and earlier. Educated people, men and women alike, memorized an immense store of texts. 5 And insofar as the less well educated urban population went to plays, read novels, or listened to storytellers, they were exposed to a rich store of classical poetry and song lyric, often as doggerel, but often as quotation of famous poems from the past. Late imperial poetry was far richer than Tang poetry in the range of its language, in the topics it treated, and in its historical depth. It was a much "smarter" poetry. Yet the predominant topic of thousands upon thousands of pages of traditional criticism was the failure

108

THE "POETIC" IN THE QING

of contemporary poetry (in prefaces usually politely allowing the author under discussion a success by precisely the criterion by which all others fail). Everyone recognized and still recognizes that poetry simply mattered less in the late imperial period, that it didn't work. It is important to clearly separate the actual failure of late imperial poetry from feeling that poetry had simply become irrelevant: at the time it mattered very much to have a poetry that mattered. Poetry was important; contemporary poems were not. Such a desire for an effective poetry touched on deep concerns of late imperial culture: it was, as I suggest above, a striking symptom of falseness and decline. 6 Huang Zongxi (I 610-9 5 ), the scholarly anthologist of Ming prose, admitted: Take a look at the collections from the past three centuries that circulate in the world or are kept in libraries. There are no fewer than a thousand writers, each of them having at the very least a few juan, with the more voluminous running to a hundred juan. Buried among the works of social exchange and all kinds of phoniness, there must be one or two genuinely moving passages. But they pile up on tables and desks, and no one looks at them. No sooner do you look at them than you find cliches all cut from the same mold, and you toss them aside immediately. 7

This is a strong admission for the anthologist who sought, in his own words, "to save those who are drowning." Huang's act of salvage did little good: it is safe to say that few read Huang's anthology Ming wen an, and fewer still recalled powerfully what they had read. Huang's motives for his anthology of Ming prose, to save it from being entirely unread, contrast strongly with those of the late classical anthologist of Tang poetry, who sought to provide proper models, with an implicit anxiety that the novice reader might be perniciously attracted to what is excluded from the anthology. Or consider Qian Qianyi: "As I get lazier in my old age, I can't stand to read poetry, and most of all I can't stand to read recent poetry. When I happen to cast my eyes on some volume of poetry currently circulating in the world, I get groggy, sinking down at my desk and going to sleep. " 8 A recluse immediately appears to advise Qian to experience poetry with his nose, by its "scent," rather than with his eyes. Late imperial criticism tended to suggest some remedy for current ills or to offer exceptions to the depressing norm; but implicitly or explicitly critics often agreed that there was a general barrenness between the Tang (or sometimes the Song) and the present. 9 Even the venerable adage that great distress made poets write well seemed to have failed, and critics were horrified to discover that the most grievous sufferings, such as those under-

Stephen Owen

109

gone in the Qing conquest, had not generally improved the literary caliber of recent poets.H1 Despite their feelings about recent classical literature, all these critics wrote in the genres of classical literature. Nor was there much agreement among them: they engaged in intense, polemical disputes on the nature of poetry. 11 What they shared was the issue over which they fought: how to write well in the present when so many were writing badly. If we simply read poems and poetic criticism, there is no way to understand the problem. The more one reads, the more one's sight is shaped by the blinders of their supremely intelligent response to the perceived crisis. Their answers to the question of what had gone wrong were often very good, and each can sustain a modern argument: the sheer volume of poetry, changes in the educational system, the fact that poetry was no longer part of the course of public advancement and had turned into a pastime, and everywhere, the spiritual subordination of the modern writer to his predecessors and the burden of the past. All of these are good, partial answers to the question why late imperial poetry failed, but none explains why anyone still cared. The same explanations apply just as well to twentieth-century classical poetry, but no one now cares or believes in the continued importance of the genre. Let us frame the question somewhat differently. Poetry was supposed to be the product of genuine feeling, and success in poetry was supposed to depend on genuine feeling: on this the critics were agreed, and they tell us so repeatedly. At the same time, as can be seen in the passage by Wei Xi quoted at the beginning of this chapter, the practice of poetry embodied participation in elite culture. How could an activity that epitomized elite culture be genuine in an age when so many manifestations of elite culture came to be potentially regarded with irony? This tension had existed in classical poetry from its beginnings and can be traced back to the unresolved conflict in the "Great Preface" to the Classic of Poetry between poetry as the involuntary expression of feeling and poetry as part of a civilizing program, bearing a moral, regulatory force. Something had changed to bring that conflict to the fore. A new class of representations, ironic, contextualizing representations, had appeared. These can be found in classical writing, but their special province was vernacular literatureY The special province of vernacular literature lay in demonstrating the frailties, follies, and hypocrisy of human beings. Prose fiction played an important role in this process, but the Ming and Qing (at least until the rise of the Honglou meng cult late in the eighteenth century) was an age of plays, and drama was the dominant form, both in the theater and in

IIO

THE "POETIC" IN THE QING

print. Formally and in its narrative conventions (e.g., the inclusion of low characters and subplots), drama was the primary ground on which high cultural norms confronted low cultural alternatives. Everyone recognized that a play like Xixiang ji (Romance of the western chamber) required Hongniang but the classical tale Yingying zhuan (Yingying's story) did not; without such alternative perspectives, recognized as having a truth at the very least equal to the representatives of high culture, the vernacular text fails. 13 The inevitable consequence, of course, was the flattening of representations of high culture. Irony in vernacular literature came in all flavors. There is the obvious irony in exposing hypocrisy and high cultural blindness, as the "low" character Chunxiang "exposes" the scholasticism of the schoolteacher in the "Schoolroom for Women" (Guishu VII) scene of Peony Pavilion (Mudan ting) as belonging to someone who teaches the Classic of Poetry but does not understand its basis in nature. In the figure of Du Liniang, the "poetic" is reconstituted against the failure of scholastic culture. Du Liniang needs the "low" character Chunxiang to guarantee her version of authenticity, to separate it from the socially sanctioned appearances. 14 But the low cultural perspective could add much finer forms of irony. One example might be Scene III ("Hongding") in Peach Blossom Fan (Taohua shan). This scene opens with the servants taking a comic inventory of the commodities necessary for a ceremony at the Confucian temple. Just as later in the scene Ruan Dacheng's hypocrisy in joining the Confucian ceremony guarantees the authenticity of this loftiest civic activity of high culture, the opening with the joking servants provides a perspective that contextualizes the ceremony and calls its perfect enclosure into question; it is a perspective that reminds us of the pragmatic world that invisibly supports the high cultural observance. The presence of the joking servants complicates our much manipulated sympathies; without it, we would have an uninteresting victory of high culture in the simple contrast of the genuine and the hypocritical. It is that act of contextualizing that gets beyond a simple binary structure of value and meaning. To understand what happened to late classical poetry, we must go beyond classical poetry itself (shi), even beyond the closely allied ci and qu, to the larger, though certainly vaguer notion of the "poetic. " 15 The poetic is what Jin Zhen was trying to stage at the rebuilt Level Mountain Hall, not just the composition of poetry but a quality of experience; arias and poems occur throughout a chuanqi drama, but there are scenes and moments that are recognizably poetic. Whether in life, as with Jin Zhen, or in drama, the poetic is a staged event (although the very idea of the

Stephen Owen

III

poetic, implying something spontaneous as the guarantor of the genuine, must run at cross-purposes with the idea of staging). This sense of the poetic includes modes of representation as well as forms of represented speech and behavior. In chuanqi we can find representation of the conflict between the poetic and the various alternatives that ironize or undercut the authority of the poetic. Since my interest is in the poetic per se, rather than its reduction to a place within the social whole (see the following discussion), I will concentrate on the attempt to redeem the poetic. In general Qing poetry was more often yinyong xingqing (to sing one's feelings) than yanzhi (to express one's abiding aims). 16 It was supposed to be the authorized manifestation of genuine feeling. The range of legitimate feeling was narrowly defined: primarily love, friendship, sadness, the appreciation of nature, and all the familiar responses that follow from the topics of classical poetry (the passions of vernacular literature-greed, cruelty, ruthless ambition, and unromantic lust-were somehow not acceptable as poetically "genuine"). When a male or female lead (and sometimes other characters) in a play find themselves in one of the legitimately poetic states of feeling, their language, particularly in the arias, jumps a register to one recognized as in some way "poetic. " 17 The poetic register was sometimes sustained and sometimes undercut by irony, but it was always contextualized, appearing and disappearing within the larger course of the play. If the Tang was the paradigm of the successful functioning of both poetry and the poetic (albeit in fact an illusory paradigm), genre was a means to create a separate and privileged social space. It could without hypocrisy be blind to contradiction. Tang poetry (and perhaps all socially effective lyric poetry) depended on the exclusion of many aspects of experience, things permitted in other genres of representation. No one who reads a Tang tale can avoid being shocked by physical objects, emotions, and events whose likes are never found in poetry. The world of Tang poetry is one constructed of exclusions. Such limitation must serve some positive role, because cultures so often demand it and assimilate new elements to the poetic repertoire only very slowly. 18 The act of limitation in a tradition of representation that is also a social praxis would seem to be the desire to create a protected space of experience. 19 In such a tradition, the paradox of the aesthetic becomes particularly clear: in Adorno's nice formulation, it is "the social antithesis of society." A need arises within the social whole and as a consequence of the social whole for a space of representation and of experience that is independent of the social whole, that excludes it. The frame of a work of art is a wall erected against its

II2

THE "POETIC" IN THE QING

inclusion in the world around it-a wall that academic literary and cultural studies often seek to breach. In China and in the countries within the Chinese cultural orbit, such a "separate space" was supposed to be characterized as genuine, natural, and primordial. 20 It is perhaps incorrect to term such a generic space "aesthetic" in the strict sense, because both in its constitution and often in its content it was intensely interested and unfree.zt But like the true aesthetic, such lyric space was essentially negative. During the course of the late imperial period, the demarcation of lyric poetry as a protected space remained constant, but it was increasingly represented as such; that is, it was contextualized. Contextualization undercuts the very essence of this particular social phenomenon, which is the negation of contextualization. That is, the paradox of art-the social antithesis of society-becomes increasingly immanent within the poetic act itself and the poetic moment. The context of the social whole can no longer be suppressed. What had previously been seen as a moment of stepping out of the complexities of the world into a protected space of genuine feeling and natural behavior (like the illusion one may have at a party) increasingly became a matter of leaving the "real" world and entering one of artificiality disguised behind the rhetoric of feeling. From this comes our historically coded distrust of the verse produced by Yangzhou burghers at Level Mountain Hall. The need to negate the increasingly voracious social whole did not disappear. If it had, poetry and the poetic would have disappeared altogether. Instead, people kept trying even harder. The fact that the attempt always failed was one factor behind those thousands of critical pages agonizing over what had happened to poetry, why it no longer worked. The struggle between the need for the poetic and its impossibility lies at the heart of Honglou meng, Six Accounts of This Life Adrift (Fusheng liuji), Peach Blossom Fan, and The Palace of Lasting Life (Changsheng dian) {the struggle between high culture as Neo-Confucianism and its impossibility seems to have been worked out in The Scholars [Rulin waishi]).

The Nobility of the Ridiculous Just as, from the Southern Song on, the communal practice of poetry gradually became only an elite pastime, there was a corresponding formation of a version of the poet (shiren) as a figure apart. 22 Such an idea of the shiren gradually approaches that of the Western "Poet." As the Poet struggles to create a separate space and separate identity, the return of contextualization makes him potentially a figure of ridicule. Indeed, being

Stephen Owen

113

an eccentric figure of ridicule (only one among many stylized poses of eccentricity in the late imperial period) guaranteed the poet's separateness and distinction. 23 The nobility of risibility is an important motif in late classical poetry. As examples, I quote two of Huang Jingren's (17 49-8 3) most famous quatrains. Offhand Compositions: New Year's Eve, guisi (1774)

-=f~~~~il\IH~iti@l§ ?f}(tlm'1~!fo/J:9'1-'!ia •f'lljJl:r!J;JiiilA.::f~

-£tZOJ=l ~~&if

Laughter and talk in a thousand homes, the waterclock drips on, yet I feel a misery coming unseen from beyond the world of things. Silent I stand on the market bridge, recognized by none, watching a single star like the moon such a long, long time.

II

lf:if.Jlt:Y'fi'IJtOEfl 5G:P:t1rmJQ~~ $dlfiiT~D-ft

§ •tiij

t.f tlift {,, jJ 1'F\¥1 A

Year after year! waste this eve reciting poems, by the lamp my children often secretly laugh at me. Yet how could the likes of you understand how much I regret wastefully spending the strength of my heart to be a poet.

Both poems are stylized dramatizations of the poet's alienation from his context, from the world around him. In the first poem, he stands apart in the larger world of Beijing, alone on a night when all others find themselves with their families; in the second poem, he returns to his family and finds himself no less an alien there. In the first poem poetic experience, seeing things as others do not see them and the very capacity to look alienate him from common humanity; in the second piece, alienation from family occurs through the act of composition that follows from poetic experience. In the first quatrain, the small thing becomes, through poetic attention, large and absorbing. On the one hand, there is no doubting the beauty of the "poetic image," whose very form embodies the enclosure of aesthetic attention, the focal point of light that excludes all else, a small thing becoming large by the very intensity of looking. At the same time the experience is mediated by placing the poet in a larger scene, by his seeing

THE "POETIC" IN THE QING

114

himself as he might look to others. Poetic experience is contextualized as such: the poem is not the poetic experience itself; rather, it is "about" poetic experience, stylized and on the very margin of the ludicrous. The laughter becomes audible in the second poem. The laughter of the children is clandestine but not unobserved; he sees himself being seen. By thematizing the vision of his own risibility, he makes an implicit claim of poetic compulsion, similar to claims made by the lyric poet and artist in the West. The explicit assertion of regret is coy. He is aware of his appearance, and there is pride in his shame. Questions of the absorption of art and its social contextualization are very much in play here. The only way to sustain the absorption of art within a world that contextualizes it is as a compulsion that survives even when one knows better. For poetry to survive as something more than an elite pastime, such an event of alienating compulsion and alienation must be dramatized again and again. The crowd outside may justly retain some skepticism about the force of aesthetic compulsion and absorption in staring at a star on New Year's Eve; it reeks of bad faith. To claim, however, compulsion in love and passion (qing) might be more persuasive, and it is there that the question of the poetic will be played out. 24 Can the poetic survive its contextualization, always seeing itself from the outside, from other perspectives? Or must it become, like leisure in our own world, a framed interval, an aesthetic vacation to forget what cannot be forgotten? The obvious answer is that it cannot survive as such; we always go home or go elsewhere; the world always undercuts a staged epiphany. As I suggested above, the poetic survives only as a desire that can never be fully realized. The ultimate failure of the poetic and of the world of love or feeling (qing), of which the poetic is so often the expression, is the conclusion of the two most influential works of Qing literature: Honglou meng and Peach Blossom Fan. Peach Blossom Fan's ironization of romantic comedy is deadly, and it reaches its culmination in "Entering the Way," Scene XL ("Rudao"). The two lovers, having long been separated by the machinations of the villain Ruan Dacheng and the upheavals occasioned by the fall of the Ming, come at last, by separate routes, to a complex of temples in the mountains. Each is standing in a separate wing of a great temple courtyard, awaiting a lecture by the abbot Zhang Wei; at that moment they suddenly recognize one another. [looks at Xiangjun from behind the fan and is startled]: That's Xiangjun standing over there! How did she come to be here? [He pushes forward urgently.]

HOU FANGYU

Stephen Owen

115

[sees him and is startled]: It's Hou Fangyu! I almost died of longing for you. [sings]

XIANGJUN

I think back on how you so suddenly left me, the silvery River of Stars was far and none could frame a bridge across, a wall higher than sky's very edge. There was no way to carry letters, and I struggled to reach you in dream in vain, yet this passion did not cease; and when I escaped, the road to you seemed even further away. [pointing to the fan]: Looking at these peach blossoms on the fan, I wondered how I would ever requite you. 25

HOU FANGYU

It has the trappings of a classic recognition scene, a grand tuanyuan, and it is wonderfully undone by Zhang Wei. The aesthetic-erotic is finally subverted by an unholy alliance of its two great competitors: political values and spiritual quietism. After discovering one another after so much suffering, Hou Fangyu and Xiangjun offer their thanks to all those who stood with them and made their reunion possible: When we return home as husband and wife, we want to repay you for everything you have done. ZHANG WEI: What do you think you are talking about with all this babbling!? Isn't it somehow ridiculous to be clinging to the love and passion that has taken root within you, now that the whole world has been turned upside down? HOU FANGYU: You're mistaken! A man and a woman founding a household has always been the primary human relationship, a focus for love through separation and reunion, through grief and joy. How can you be concerned about this? ZHANG WEI [angrily]: Ahh! Two besotted little worms! Where is this "homeland" of yours? Where is this "family?" Where is your "ruler?" Where is your "father?" Is it only this little bit of romantic love you can't cut away? [sings] HOU FANGYU:

Pathetic trifling of man and maidthe world turned upside down and you don't care; you babble on with wanton phrases, lurid words, tugging clothes and holding hands, declare a happily-ever-after to the gods. Don't you realize that long ago your fated wedlock was erased from registries in Heaven.

II6

THE "POETIC" IN THE QING

With thudding wingbeats mated ducks wake from dream and fly apart, the precious mirror of reunion lies in fragments on the ground, happy endings proved unsound. Blush at this bad performance of your scene, inspiring bystanders' laughterthe Great Path lies before you clear, flee on it immediately!

The conventions of chuanqi are initially satisfied by the lovers' moment of recognition, done in the grandest style. But moving toward this moment, the play has been preparing us to let that ending collapse, to swallow up the paltry good fortune of the lovers in a larger context of death, destruction, and national disaster. In his aria Zhang Wei invites us to see them as actors and to laugh at the bad romantic play they have been performing. Lovers may be, like poets, risible in their passion. Low characters, admirable at least for their common sense, often mock them. But in the final scenes of conventional romantic chuanqi, the lovers are allowed to win out. Here in Peach Blossom Fan the romantic ending is challenged, not by a low character but by Zhang Wei, one of the few entirely admirable characters in the play. Hou Fangyu, mocked in the wrong act for a romantic hero to be mocked in, responds with Confucian platitudes and is reminded that the world to which they intend to return and "live happily ever after" no longer exists. If a defense of the poetic, of qing, and of the rejection of context is to be possible, it must be found in a compulsion that survives all the mockery, common sense, and moral judgment that can be thrown against it.

The Palace of Lasting Life and the Defense of Poetry Hong Sheng's (I645-1704) The Palace of Lasting Life (final version r688) is explicitly a defense of qing. It is also, in many ways, the last great defense of the poetic and the high style, attempting to redeem them from irony. It is both a complex defense and a very uneasy one, deploying its divine machinery to create a literally separate space in which the poetic can exist without contextualizing ironies. It is not in any way a humanly credible defense; it is an allegory, working out a premise stated in the Prologue. The old story of Xuanzong and Yang guifei was appropriate for the dramatist who wanted to work out the absolute claim of qing announced in the Prologue:

Stephen Owen

II7

On passion's stage, in olden times as now, whose hearts stayed true until the end? If only lovers keep their faith and never fail, at last they will be joined again. No separation troubles them, though miles in thousands lie between from north to south; and life or death for two such hearts IS no concern.

"Keep their faith and never fail," is a rough rendering of jingcheng bu san, "essential authenticity not dissipating." Like the danger to moral perfection in Neo-Confucianism, the threat to love is the loss of cheng (authenticity), a double-heartedness of conflicting motives and multiple perspectives. In this venerable story of Xuanzong and Yang guifei, qing could be placed in direct conflict with absolute public duty without being subsumed by it or, as was common in chuanqi drama, reconciled with it. Hong Sheng used the autonomous scene structure that was a formal resource of the genre to work out a slow segregation of the conflicting elements, dissipating the tension between them without truly reconciling them. Qing and the poetic have their place, their absolute claims, but that place is altogether elsewhere. It was not a realistic answer, but, as the merchant says remarkably in "Ballad" ("Tanci," Scene XXXVIII): zhi yao changde haoting, guan ta huang buhuang, "so long as it's beautifully performed, it doesn't matter whether it's a lie or not." Someone at least recognizes the status of a story of absolute qing: a beautiful and compelling illusion. The Palace of Lasting Life is a very carefully constructed play. Contextualizing ironies depend on the presence of difference, a mixing of levels and points of view that makes the wholehearted absorption of high cultural experience impossible. 26 Unable simply to create a world of innocent absorption (much less to sustain such a world for so acts), the play creates a world of distinctions, and then moves from the transgression of distinctions to the restoration of segregated hierarchical spaces and finally to the absolute autonomy of a private space. Transgression serves as the guarantor of boundaries, just as cheating is necessary for the rules of the game in game theory; but once the central characters are drawn up into the heavenly world, which is the domain of the poetic, transgression ceases and the contextualizing world of articulated differences is annihilated. Before those final acts of the play, things are always moving between hierarchical levels, creating the articulation of division that will eventually permit the resegregation of levels. Sometimes this crossing of hierarchical

II8

THE "POETIC" IN THE QING

boundaries is transgressive, sometimes it is an event of grace. The "rise" of Yang Guozhong and An Lushan by their own machinations and the lack of caution in dispensing imperial favor (imperial attention being occupied elsewhere) are obviously transgressive and serve to show the difference of levels, that they are in places where they do not belong. In "Outing on a Spring Holiday" ("Xiyou," Scene V) the precious objects dropped by the Ladies of Qin, Guo, and Han on their way to an imperial banquet at Twisting River Park are recovered by commone~s, whose ignorance creates a similar situation of disjunction, an inappropriate crossing of levels. Grace in crossing levels of hierarchy may no less call to attention hierarchical difference, as in "Offering Food" ("Xianfan," Scene XXVI), in which an old peasant offers rice to the hungry emperor in his flight from the fall of Chang'an. The positive crossing of levels in "Offering Food" answers a parallel scene, "Bringing the Fruit" ("Jin'guo," Scene XV), in which envoys, rushing to bring lychees to Chang'an in time for Yang guifei's birthday, literally transgress boundaries, trampling a peasant's fields and riding down a blind old man. The way in which such parallels, juxtapositions, and crossings serve to create multiple perspectives and ironies is well illustrated in the scene that follows "Bringing the Fruit," "The Dancing Circle" ("Wupan," Scene XVI). Here the lychees cross from the violent, competitive space of the race between the envoys into the Palace of Lasting Life, where they are presented to Yang guifei on her birthday. Suddenly the lychees are seen only as a kind and thoughtful gesture of the emperor's affection: "My Lady, since you love this fruit, I sent a special edict to these places to have them brought to you with all possible haste. And today these fine fruits have arrived just in time for your birthday banquet." But the context of the preceding act does not permit the reader or audience to forget what might otherwise be easily forgotten. In the same way, we might give a loved one a gift of ivory or mahogany or an antiquity, while in the back of our minds are visions of rotting elephant carcasses, ruined forests, or looted tombs. The song that follows the presentation of the lychees tries to hold our attention on the "thing," luridly focusing on the fruit, going through their "red velvet" skin to the "crystalline white balls" of their flesh, which ends up dripping from Yang guifei's teeth as an "alabaster nectar." This is a countermovement of absorption, first of attention and then of ingestion, that opposes, and in doing so strengthens, the other, contextualizing level of public suffering that the lychees cause in their passage.

Stephen Owen

119

Perhaps the most interesting crossings in the play are those of the music "Coats of Feathers, Rainbow Skirts" and Yang guifei's stocking. The famous "Stocking Viewing" ("Kanwa," Scene XXXVI) is another good example of contextualizing by multiple perspectives-in this case within a single act. In "Stocking Viewing," a stocking, left behind after Yang guifei's forced suicide and used by a tavern keeper to attract business, becomes an object for viewing by different people and thus an object for multiple interpretations. Each interpretation has its own claim on truth. To the tavern keeper Granny Wang, it is a commodity guaranteeing her income; to the commoner Guo Congjin (the peasant who offered humble rice to Xuanzong during his flight), it is a stinking piece of footwear that recalls all the sufferings and wrongs done to the people before the An Lushan rebellion; to the Daoist nun, it is a reminder of the transitory nature of pleasure and glory (as well as being a relic that might give luster to her monastery); and finally to the scholar and musician LiMo, it is a poetic reminder of vanished splendors. See the thin lining of scented cotton, supple and light as a cloud of the gods. In golden palaces long ago its tiny paces were seen by none. How sad that today at a bar it is so casually displayed to the common eye. Traces of thread, needle marks, each layer wounds the heart. Too bad that this loveliest lady ever who suffered the greatest wrong ever, left nothing but this relic, eternally fragrant, to be passed on eternally.

Here is the mortal, contextualized poetic, only one truth among many, anxious at the contaminating presence of unpoetic eyes ("How sad that today at a bar/it is so casually displayed/to the common eye"). Things from higher levels of the hierarchy are always in danger of leaking down to lower levels and being defiled by the common eye. LiMo is, of course, unaware that the poetic, the love between the emperor and his consort, has not been lost but will simply go altogether elsewhere. It is Hong Sheng's task to reach that protected elsewhere from the crowded and various here. Scene II, "Troth" ("Dingqing"), which begins the play proper, announces high romance as all the palace attendants accompany Xuanzong and Yang guifei to the bedchamber:

THE "POETIC" IN THE QING

!20

ALL

[singing]: Flashing flames, a thousand rows, the crowding light of silver candles goes; And when we turn to look through beaded curtains drawn aslant, a silver river of stars out there faintly glows. Through tiered walkways and winding corridors everywhere the scented powder of petals wafts upon the air. How late does it seem? The moon stands high above the palms of the statue Immortal of Bronze. This is the night that reigns supreme in beauty of spring weather, red covers the azure screen, and inside brocade clouds will be a matched phoenix pair. "Bloom of Jasper," "Trees of Jade," "Night Moon and Spring River"each note sung in unison, as moonbeams pass the palace walls. We lift the hanging arras of lace and help the lovers, tipsy still, into the orchid chamber.

What is odd, of course, is that this opening scene has all the ceremonial trappings of a conventional final act, a concluding tuanyuan, but in this case the "wedding" occurs and the lovers go to bed at the beginning of the play rather than at the end. For, as the "Prologue" ("Chuan'gai," Scene I) tells us, this is a drama of the continuation of love rather than its consummation. Continuation becomes immortality, which lifts the high sentiment of qing beyond contextualization to another plane. It passes through the stage of the ridiculous and by sheer endurance transcends it. The final scene is not tuanyuan, "union," but "Chongyuan," "reunion" (a second ending). 27 We are permitted only this one scene of conventional romance before it is swallowed up in multiple contextualizations. In "Bribe" ("Huiquan," Scene III), An Lushan abjectly seeks pardon for a capital offense from the minister Yang Guozhong, a scene that adumbrates the dark consequences of imperial passion. But a few scenes later we have the unexpected: an imperial dalliance with Yang guifei's sister that results in a spat between

Stephen Owen

121

the lovers and Xuanzong's groveling desire for reconciliation. The political point of the reduction of the emperor to an ordinary mortal disempowered by love is not missed, nor is the functional purpose of these scenes, an initial crisis that first guarantees the continuation of love against roving male fancy. But most important, these scenes make love look ridiculous in something that approaches a doting parody of passion. Immediately following the mutual apologies of the imperial lovers is a whole new context. In "Omen" ("Yichen," Scene X), Guo Ziyi, the Confucian hero, arrives in Chang'an, where he goes to a tavern and witnesses the transgressive ostentation of the Yangs and An Lushan and responds with politically correct outrage. He reads on the tavern wall the prophetic poem that foretells the rebellion and the death of Yang guifei. Guo Ziyi has his own version of the poetic, a stylized loyalist machismo in the manner of Xin Qiji, which is the positive transformation of An Lushan's miles gloriosus. Coming immediately after scenes of corruption and the emperor's games of romance, Guo Ziyi's sense of commitment to the commonweal, loudly absent earlier in the play, immediately claims our assent. Guo closes his final aria in this scene with the lines: I rejoice that it is my lot to spurn the clouds and rise to Heaven's avenues, at last I will set the universe aright, and I intend a legacy of most glorious deeds to last a million years. And though foul witchery is abroad, I can do no less than on these shoulders bear the sun and moon, and with my own hand prop up the Great Tang.

Having fallen from the romantic heights of the opening wedding scene, at last we have a culturally approved place to stand. But then, with what can only be a calculated leap, the next scene, "The Music" ("Wenyue," Scene XI), takes us immediately to the moon, where ChangE sings in a high poetic language that is utterly without irony, very similar to the imperial language used in the opening scene, a mode of discourse that had just been contextualized and undermined by Guo Ziyi's appearance. In the earlier scenes of the play, absorption in the high poetry of love had been revealed as a blindness that led to corruption and folly. By displacing that poetry (but not yet love itself) to the world of the moon, it can be

122

THE "POETIC" IN THE QING

beautiful yet free of corruptive consequences: it is, in effect, decontextualized. Chang E sings: The finest nights lie under the sway of my clear beams alone, unsullied by the dust's least mote since time began. Wind-borne dews are strewn through crystal air and splash the moon's silver rays, in a stream of faintly tossing notes of music of the gods. Such radical juxtaposition, drawing on the formal conventions of chuanqi, is here used to create autonomous realms, each with its own internal legitimacy. Against the tendency to contextualize, swallow up, and deflate the high poetic, Hong Sheng deploys the powerful motif of the dream play, which promises, in its own way, to swallow up and contextualize what occurs below. This is the explicit content of the music "Coats of Feathers, Rainbow Skirts," which plays such a central role in the play. The immortal maidens sing: But though the agile tongue may trill the pipes and slender jadelike fingers pluck the strings with harmonies increasing, we waken not the nightmare dream in the world of mortal men, nor do we halt the motion of the waterclock in palaces of Heaven. It is, of course, a fine touch, characteristic of the period, that Yang guifei visits the moon in a dream, where she hears the music that tells her that this is the "real" world and that her mortal, waking world is one of dream. "Coats of Feathers, Rainbow Skirts," with its embedded message that poetically contains the unpoetic, plays a central role in the drama, crossing realms, articulating their difference, and guaranteeing their separation. Yang guifei memorizes the music of "Coats of Feathers, Rainbow Skirts" and, as she is instructed, takes it back to earth, where she copies it out as a score. Next she teaches it to her attendants Yongxin and Niannu, who pass it on to the imperial ensemble, under the direction of Li Guinian. As the imperial ensemble practices, the music makes another, unauthorized crossing, as it is overheard and memorized by the young scholar Li Mo, listening outside the walls of Huaqing Palace, who replays it on his

Stephen Owen

!23

iron flute ("Stealing the Music," "Touyue," Scene XIV). Eventually the music makes it all the way to the bottom of the hierarchy: after An Lushan takes the capital, he has it performed for himself in "Denouncing the Rebel" ("Mazei," Scene XXVIII). After the rebellion, "Coats of Feathers, Rainbow Skirts" reappears in a peculiarly elegaic context, as LiMo meets the aged Li Guinian in the south, in the famous "Ballad" ("Tanci," Scene XXXVIII). The penultimate music scene is perhaps the strangest of all: this is "Immortal's Recollection" ("Xianyi," Scene XL), in which Chang E sends her maid Hanhuang to the now deified Yang Yuhuan (Yang guifei) to get the scored version of "Coats of Feathers, Rainbow Skirts" that she had written down in the Tang palace. This was, of course, ChangE's own immortal music, which Yang Yuhuan originally scored from memory of her visit to the moon and now recopies from memory up in heaven. The final appearance of the music is, not unexpectedly, in "Reunion" ("Chongyuan," Scene L), where ChangE has it played for the reunion of Xuanzong and Yang Yuhuan. Detailing an unbroken chain of performance and learning, the music scenes are a small drama of transmission and the play's moment of self-referentiality: "Coats of Feathers, Rainbow Skirts" is something from another world that, by grace and accident, is allowed to "appear" in this world without being corrupted by it; then it returns to heaven changed, even enhanced. We eavesdrop on it or are voyeurs, as in the wonderfully lurid bath scene "Peeking at the Bathers" ("Kuiyu," Scene XXI), where the low figures watch imperial love-making in the bath. The representatives of Heaven-the Weaver and Oxherd, Chang E, and the apotheosized Yang Yuhuan-can speak high poetry in its purest form. The star-lovers, the Weaver and the Oxherd, who are both audience and stage managers of the events below, are the literal apotheosis of love. I think one might best describe the beatific state they have attained as being lovers without servants and family (the Weaver is permitted some handmaidens, but they serve as chorus rather than commentators with distinct viewpoints). Social hierarchy, and with it all context, is erased. And it is significant that these two deities speak to one another as equals. 28 Although he has some decidedly unpoetic tantrums, Xuanzong's high poetry is constantly undermined by context; he never, however, surrenders the poetic. An almost heavenly poetry of presence is found in the arias of the wedding celebration in the second scene, "Troth" ("Dingqing" ); but it soon becomes apparent that the very intensity of absorption corrupts the world around it, to the degree that it will eventually intrude upon and shatter the love song. Xuanzong's high poetry passes almost seamlessly from celebration to loss, from presence to absence, as when he laments

124

THE "POETIC" IN THE QING

his loneliness after the death of Yang guifei in "Hearing the Bells" ("Wenling," Scene XXIX). With imperial stubbornness he sticks to the language of love, and for this compulsion the machinery of the play rewards him by finally allowing him to emigrate to a land where he can speak in a high poetic register for eternity. "Infected" by the heavenly music, Li Mo becomes one of the most interesting and complex representatives of the "poetic" in the play. Li Mo's high poetry, in contrast to that of the emperor, Yang guifei, a.nd the deities, is reflexive, elegaic, and late classical. He is the witness of qing in music. He goes, singing high poetic arias, to listen at the walls of Huaqing Palace in anticipation of that absorbing experience, and afterward his own version of high poetry is always tinged with the sadness of loss. He is, perhaps, driven by qing no less than Xuanzong; but his compulsion, like that of the late classical critic, is not with the woman herself but with the representation and the trace. He journeys far to see Yang guifei's stocking, and he is always seeking the music "Coats of Feathers, Rainbow Skirts." And as the lover of traces and representations, Li Mo never makes it to heaven. We may read The Palace of Lasting Life as an allegory of the poetic as much as an allegory of the absorption of love. It is figured in "Coats of Feathers, Rainbow Skirts." The lovers that prove true in the end cannot be separated from the music, although the music and each of the lovers circulate independently. The lovers can err and be contextualized; their music can be performed even for An Lushan without being corrupted. The lovers achieve their separate space by enduring their trials, and thus it is only the transcribed, mortal rendition of "Coats of Feathers, Rainbow Skirts" that is finally appropriately played at their reunion in Heaven. That former "Rainbow Skirts" is here writ anew; and sung for those who understand, its import will be grasped, that love may last forever more.

"Reunion" At last the lovers renounce sexual love; apart from their crises and conflicts, they never had much to say to each other. Rising to Daoli Heaven, the Heaven of Heart's Desire, they leave behind not only mortal society but even the company of immortals who stage-managed their reunion. In the end all conflict and all context are transcended. No one will ever again

Stephen Owen

125

abuse the loan of their merely temporal power or peek at them when they bathe. They have achieved an unironic enclosure of mutual bliss; if they speak to one another in Daoli Heaven, their conversation will probably be very poetic and not very interesting. Long before, in the seventh century, Taizong-Tang emperor, fratricide, and usurper of his father's throne-could play the perfect Confucian prince before his ministers, bowing his imperial will humbly to their wise advice. The image of his virtuous government in the Zhenguan reign period was legendary. Perhaps the dark context behind the legends lends depth to the role he played, but we do not push it so far as to allow that context to undermine the role. However, in a world where the truth can no longer exist entirely in the moment of performance, but rather must be found in the larger context and multiple perspectives, such absorption in the single role depends upon a radical act of negation, leaving all the other worlds behind. It is almost impossible to imagine the discourse that could occur in such a state; we cannot guess what Xuanzong and Yang guifei will say to one another for all eternity in Daoli Heaven. But fortunately that was not the topic of The Palace of Lasting Life; the play is concerned with the process of transcending a world inhabited by others, who represented too many versions of truth and lie. The play closes with the lovers, Xuanzong and Yang guifei, soaring on phoenixes up toward Daoli Heaven, to the strains of "Coats of Feathers, Rainbow Skirts."

CHAPTER

5

A Jiao Is a Jiao Is a ? Thoughts on the Meaning of a Ritual Robert Hymes

{Ln a collection of Song dynasty miracle stories preserved in Verities of the Three Lords Fuqiu, Wang, and Guo of Mount Huagai, a twelfth-century sacred geography and hagiography for a cult of Daoist Immortals centered on Mount Huagai in Jiangxi, the following story appears:

J

Below Dafou Mountain there lived a certain Yanxian; I have foolishly forgotten his surname. In the year jihai of the Chunxi reign period [ri79], there was suddenly a noise behind his house. He rushed to look. It was an avalanche, of a force certain to crush the household with all its members. The ground shook so much that there was no sure place to hold on to. So he gathered his whole family, and they looked toward Dafou Mountain and appealed to the Three Immortals of Huagai to save them, vowing [to hold] a jiao to pay them reverence at the mountain. Their appeals were most pitiable. Suddenly the avalanche stopped. The whole family was saved. To this day the break in the earth is still there. The clear response in this case is particularly evident. Within half a li of Yanxian's house was a certain Zhang Bangchang, whose family encountered "the mountain's roar" on the same day, and half of them were crushed to death. Ah! When Yanxian, who appealed to the Immortals, receives good fortune in this way, while Bangchang, who had not appealed, receives ill fortune in that way, how can the people of this age not serve reverently the Immortals and Perfected? 1

The story contains what I would like to treat, for the sake of argument, as a paradox. The plot is simple enough: people in danger call out to certain divinities to save them, and they are saved. Others do not call out, and they are not saved. Although the attention to those not saved is unusual, in other respects the story resembles a number of others preserved in the same collection. The directness of the appeal-its immediacy in

A JIAO IS A ]lAO IS A ?

both senses-is important. In work still in progress, I have argued from these and other miracle stories that the cult of the Three Immortals of Mount Huagai, or at least that aspect or segment of the cult visible in surviving texts, upheld a notion of divine authority that involved direct and often personal relations and communication between humans and Immortals, and accordingly-despite the obviously Daoist character of its central figures-played down or simply ignored the bureaucratic model of divinity traditionally central to the texts and practices of professional Daoists. The Three Immortals appear in the miracle tales and other texts of their cult as inherently powerful, direct patrons of their human worshipers; as teacher and disciple or elder and younger brother to one another; and as teacher or even as employer/master to other gods or Immortals. 2 Unlike the divinities represented in professional Daoists' liturgical and ritual texts (at all periods, it seems) and sometimes in modern Chinese "popular" religion, the Three Immortals almost never appear as celestial bureaucrats, and they never serve as intermediaries or intercessors between human appellants and some higher divine authority. Based on my work on the Three Immortals, I have argued that we need to recognize at least two distinct vocabularies for representing gods and their relation to humans in Song (and, I would argue, later Chinese) religious culture: not only the bureaucratic vocabulary of god-officials and their derived or delegated authority that has been so prominent in the discussions of anthropologists since Arthur Wolf's seminal article, 3 but alongside this, and sometimes contending with it, a vocabulary of direct or personal relations and of an authority inherent in the gods who wield it. 4 The two vocabularies may of course combine and intertwine, but the concentration on the personal and the near exclusion of the bureaucratic in the Huagai cult texts exemplifies how the two may be used to articulate quite different notions of authority. I have argued that the Three Immortals became the most important local divinities for the secular elitethe shidafu stratum-of a part of eastern Jiangxi during the Southern Song precisely because they could represent a non-governmental, inherent, personal, and fundamentally local authority akin to the authority that local gentlemen in just that period were coming to vest (or to wish to see vested) in themselves. The absence of bureaucratic vocabulary, indeed the absence of any suggestion that the Three Immortals were (mere) mediators or intercessors to higher authority, was crucial to the cult's appeal to men of a local, and increasingly localist, elite. What, then, is the paradox in the story? Simply this: Yanxian and his family appeal directly, with no mediation but their own language, to the

Robert Hymes

Three Immortals. They ask for help and receive it immediately. But in asking, they make a promise: to hold a jiao for the Three Immortals, as a way "to pay them reverence." The jiao is the central sacrificial ritual of the Daoist liturgy. It is a ritual that, as I think all who have studied it would largely agree, seems to embody definite notions of authority: authority in which complex and almost obsessively mediated hierarchy is crucial, and which expresses itself in explicitly bureaucratic terms and images. Ultimately a jiao is not about anything personal, individual, or indeed particular at all: not about the individual worshiper's act of worship; not about the boon granted by a particular god. In jiao as we know them from studies by Daoist specialists and anthropologists in modern Taiwan, the gods of everyday life, the gods of temples and shrines people normally worship, are deliberately and obviously subordinated, through the physical repositioning of their images, to a vast celestial officialdom of which they are redefined as (at best) only a lowly part and ultimately to the supreme, unpersonified, and highly abstract god-principles that Daoists know as the Three Pure, as well as, nearer at hand, to the Daoist practitioners themselves, who hold their own positions in the heavenly bureaucracy. The purpose of the ritual, in the Daoists' view, is to reestablish the harmony of the cosmos on behalf of an entire community and more broadly on behalf of the whole human world. To sum up the problem, the miracle stories of the Three Immortals seem to tell us that people may communicate more or less directly with their local gods, who in response act directly, and out of a power and authority that are their own, to aid their worshipers. In contrast, the jiao seems to tell us that local gods merely convey people's appeals to (and intercede with) their celestial superiors, who are to them as the very highest earthly officials are to the lowest. That process must be further mediated by the Daoist specialist, a member of the same heavenly officialdom. Against this background, it seems fair to find it odd, at least on the face of it, that a family in danger, seeking immediate rescue by familiar local divinities, should offer as thanks-or rather, should promise as inducement that they will later offer as thanks-a jiao. Naive questions spring to mind. Are they promising to foot the bill for a ritual display of the utter subordination, indeed the marginal relevance to the real cosmic powers-that-be, of the gods they are asking to help them? Is it for this that the Three Immortals save them? Do Immortals enjoy being made to look unimportant? How does a ritual that proclaims bureaucratic hierarchy fit into a cult whose texts give little or no place to bureaucracy, or to hierarchy that is not personal and dyadic? How does worshiping

A JIAO IS A JIAO IS A?

the Three Pure thank the Three Immortals? How does a jiao "pay them reverence"? There is of course a larger historical question here, of which my own question is only a special case, and I should say that I will not try to offer a historical answer to that larger question. That is, I do not know when and how it happened that a Daoist jiao came to be a standard ritual vehicle for addressing, propitiating, rewarding, and thanking the gods worshiped generally among the lay Chinese populace at large. I suspect that it happened during the Tang-Song transition or the Song, and that it has its foundation in a commercialization of religious life that accompanied the commercialization of other aspects of daily life in the same period. But this is not the problem I am addressing here. Rather, I am querying the paradox of meaning, and my naive questions aim to clarify and sharpen the paradox. There are at least two ways to answer these questions quickly. One is to say that there is no problem here at all. Someone who argued this way might mean that the historical question is the only question to be asked: that once jiao became conventional vehicles for addressing gods, that's what they were for their lay users, whatever the Daoist officiants might have meant by them or thought they were expressing in the ritual performance itself; that there is no reason to expect consistency of meaning, in the observer's terms, between the different aspects or uses of a religious ritual-the argument is familiar enough. . The other quick answer-which might be elaborated at considerable length but which I think is still "quick" in its effect on our approach to this sort of problem-would be that both what the lay worshiper is doing (or seems to be doing) in his/her direct approach to a particular divinity and what the Daoist practitioner is doing (or seems to be doing) in his ritual performance and recitation are part of a single system of meanings, thus (again) that evident conflict or difference hides truer unity. The analyst, if she is brave, will then supply the single system of meanings that explains everything. To this end, one direction of argument might be that the apparent difference of meaning is resolved at a higher level of generality: that the things the worshiper and the Daoist do and mean, when we abstract from the details, are broadly similar in ways that show their membership in one community of meanings, one culture. (There is no logical reason why a "system of meanings" or a "cultural system" must mean a system of similarities of meaning, but this is in fact how the notion is commonly deployed.) For instance, both the Daoist and the god-worshiper believe in gods, and lots of them; both believe there is hierarchy among gods; both conceive hierarchy among gods (and between gods and people) in

Robert Hymes

133

ways that parallel hierarchy among people; both believe in relations of response, mutual effect, or resonance between the world of gods and the world of human beings; both believe that the two worlds are always or regularly in contact; both use physical images to represent gods; these images generally look much alike-the list could grow very long. Thus we see how very Chinese both the worshiper and the Daoist are, and everyone is happy. Another, rather different strategy, but still in the one-system-of-meanings ballpark, would be to resolve the difference entirely at a single level: the worshiper's gods and the Daoist's gods simply do fit together into one system, which is precisely the Daoist celestial hierarchy of deity-officials topped off by the three godly abstractions; the more fleshly Three Immortals find their appropriate place near the bottom of this structure and, as participants in it, both acknowledge and embody its rightness, beauty, and power (after all, they have chances for promotion!); when the common individual worshiper treats certain happenings as the direct acts of the Three Immortals, he does so either out of ignorance or as a kind of customary shorthand, because these are really the acts of the heads of the celestial bureaucracy, to whom the Three Immortals appeal on worshipers' behalf by appearing in a heavenly audience before the throne. Professional Daoists, in this reading, are the cultural experts who understand the system, and on whom worshipers rely for an understanding they do not have themselves. Thus one simply fills in the blanks in the Verities of the Three Lords. In what follows I avoid all these ways of arguing, not because I think all of them are equally wrong, or wrong equally often, but because I think they have the effect, in this case and others, of avoiding or shortcircuiting analysis. The first proposes that a seemingly meaningful ritual act or expression should be treated as meaningless in practice. This avoids many knotty problems, no doubt, but it suggests (in anything I've read) no way to recognize in general when the apparently meaningful is really not, unless the answer is: whenever seeing it so helps us avoid knotty problems. Arguments of the second class, the single-system-of-meanings arguments, may fail in other ways. Raising things to a higher level of generality indeed, tautologically, allows one to find a similarity missed at a lower level, but it remains unclear why and within what sort of unit similarity-seeking should be privileged; how one knows that a higher level of generality is appropriate to a particular question; and where, or why, the process of level-raising should stop. After all, even the hypothetical list of similarities offered above quickly ceases to seem so "Chinese" when

A ]lAO IS A ]lAO IS A ?

one considers the Indian parallels to virtually every item. One could even raise the level of generality so high as to construct "fundamental similarities" between Chinese and Christian religion; would this show that both were part of a unified system of meanings? The single-level solution instead abolishes difference by absorbing one member of a seemingly contrasting pair into the other: local gods are simply part of the larger Daoist system. This again is easy to do, but I am troubled in this case by the fact that to my knowledge hardly anyone in Song or modern China ever does this, ever seems explicitly to offer this sort of "synthesis," except professional Daoists; thus, in leaping to offer it oneself, one would seem simply to be taking sides. (There is after all, again, not a word about any higher celestial authorities, about any offering of memorials or attendance at audiences by the Three Immortals, about any intercession by anyone with anyone for anyone about anything, in the miracle stories in the Verities of the Three Lords.) To argue that lay worshipers of the Three Immortals understood only in part what professional Daoists understood as a whole invites the answer, "Who says so?" Well, professional Daoists say so, that's who. "Filling in the blanks" can mean that the people one studies can never, intentionally and knowingly, leave something blank: no silence is meaningful. I prefer, at least in the present context, to assume that there is a difference here worth examining, and to see where that assumption gets me. But I am a bit ahead of myself. It would be good to begin by showing that what I have proposed as a paradox-the absorption of the jiao into a context where direct and personal interaction between gods and individual humans is the point-recurs frequently in the Verities of the Three Lords. There are about 25 miracle stories in the book (all gathered in its last two chapters); of these, nine mentionjiao. Good method would probably require reproducing these in full so that the reader may judge them through no filter stronger than that of my translation. Space does not allow this; at the risk of the distortion that conciseness entails, and of a certain tedium (these are entertaining stories, and condensing them into a line or two sacrifices their flavor), I summarize each of them briefly here. In the first story that mentions a jiao, a certain Luo Bin, an officeholder in the regional military administration, prays daily to Mount Huagai for a son. "On the first and fifteenth of each month, he would go to the jiao at the Zhaoqing Abbey and offer prayers of various kinds, begging to be granted a dream in answer. " 5 Finally he dreams that a Daoist appears to him and gives him three plums, one slightly spoiled. In time he has three sons, two of them highly successful and one who dies young. In a second

Robert Hymes

135

story, the family of a man of Hunan is cured of plague by a strange wandering Daoist who says he is a certain Guo of Mount Huagai in Chongren county, Jiangxi. When he vanishes, the man travels to the mountain to find and repay him; the locals realize that the apparent mendicant was none other than Immortal Guo, one of the Three Immortals, and tell the man, "You should go quickly, offer incense and paper money, and repeatedly hold a jiao of thanks," which he does. 6 A man named Wang, in a third tale, seeks an heir at Mount Huagai. "After several days he reached the very peak and commanded the yellow-hats [i.e., professional Daoists] to set up a jiao. He burned a prayer, promising that he would pay one visit each year until he had a son and would then hold a great and splendid affair to repay the boon of the Immortals." That night three Daoists appear in his dreams, thank him for coming so far to see them, and tell him they are sending him a son in return. After the son (who narrates the story) is born, Wang holds a jiao on the mountain every year to express his thanks, and his son continues the annual ritual after him, remarking, "How can I ever repay in balancing measure the boon I have borne from the Immortals? " 7 In a very different sort of story, an assembly of worshipers (perhaps including professional Daoists), who have come to set up images to a certain group of Four Immortals at a different mountain, hold a jiao to dedicate the images, and in the course of the jiao-though not through the ritual process itself-are informed by the Four that they must place an image of the first of the Three Immortals of Huagai at the center of their altar, as he is the Four's superior. 8 A wealthy man in another story ascends the mountain and holds a jiao each year to pay his respects to the Three Immortals. 9 Six good-for-nothing rustics who climb the mountain in yet another tale are punished by the Immortals (with lightning!) for having eaten dog meat on the way up; terrified, they "made a vow to hold a jiao in order to offer grieving acknowledgment of their repentance." 10 The seventh jiao story is the tale of Yanxian and the avalanche. In the eighth, a gentleman visits Mount Huagai to obtain "sagely water" (sheng shui -holy water, if one prefers) for his mother, who is ill. After she takes it, three Immortals visit her in a dream and give her a pair of bronze chopsticks; she awakens and is well. When the family visits Mount Huagai to give thanks, she recognizes the images of the Three Immortals as the Daoists she met in her dream; "thereupon they held a great jiao and feast in thanks for the boon. " 11 In the last story, a drought hits Chongren county; the county administrator, his subordinates, and gentlemen and commoners of the county join in escorting the images of the Three Immortals to the

A JIAO IS A ]IilO IS A ?

county seat and holding a series of rites in their honor, culminating in a jiao; the rains come as soon as the jiao is finished. 12 Two things are interesting about these jiao. First, we are accustomed from ethnographic evidence to think of a jiao as a collective community ritual. But only two of these jiao seem to fit that description; the one performed on the first and fifteenth of each month at the Zhaoqing Abbey in the first story and the jiao performed for drought relief at the Chongren county seat in the last. 13 Aside from these, the jiao are offered by individual men or particular groups for quite personal or specific reasons and with no suggestion of benefit for a larger populace or the cosmos. Second, the stories treat the jiao as serving essentially three functions in relations between humans and the Three Immortals: as an object of exchange, or proposed exchange-as when someone promises a jiao if a favor is granted; as a token of thanks or of apology; or as a setting for personal prayers to the Three Immortals. In the first two functions, the jiao really seems to become simply a kind of gift from humans to the Immortals. To be sure, one cannot speak of unmediated contact here, since the jiao itself is mediating, even establishing or cementing, congenial relations with the Three Immortals, communicating or guaranteeing to them specific feelings or intentions of human beings. But the quite different sorts of mediation that the jiao as we see it in ethnography on Taiwan or in Daoist ritual manuals embodies and expresses-the mediation of local gods or immortals, as celestial officials, between human subjects and the truly authoritative highest gods; the required mediation of living Daoist professionals between worshipers and the gods; the bureaucratic and documentary character of all these processes-are missing. They are missing in large part because the jiao here is a sort of black box: we see it only from the outside, from the vantage point of the relationships it serves. We see that people offer it and that Immortals must be glad to receive itor how could it play the gift-like role it plays?-but we do not see inside it: we do not see what happens in the ritual itself. Even the professional Daoists who presumably are the physical performers of every jiao in the stories receive explicit mention only once, when Mr. Wang who wants a son "commands" (ming) the "yellow-hats"-an informal and anything but exalted term for Daoist practitioners-to set up a jiao at which he burns his own personal prayer and makes his vow. This is the jiao as mere setting; and the language ("commands") clearly subordinates the Daoists to the individual worshiper, turning them from mediators with the gods into instruments Wang uses to create the conditions in which he can communicate personally with the Three lmmortals. 14

Robert Hymes

137

Now, it would be wonderful to know "what was really happening" in these jiao to the Three Immortals. Other Song texts, notably the ritual manuals preserved in the Daoist canon of scriptures, do record something closely akin to what modern Daoists and their anthropologist observers describe today. During the Song as now, a written "memorial," modeled in many respects on the memorials that earthly officials used to communicate their wishes to the emperor, was crucial to the rite, 15 and then as now the deities honored and celebrated in the jiao, or in its sister rite the zhai, constituted an almost unimaginably vast pantheon-government, dwarfing and putting in the shade the particular gods worshiped in any specific place. Consider, for example, the grand listing of divinities in the Ceremonial Forms for the Performance and Completion of the Unsurpassed Grand Retreat of the Yellow Register ( Wushang huanglu da zhai licheng yi), compiled by Jiang Shuyu (ca. rrs6-r217). 16 This Southern Song guide to the performance of a Daoist ritual of the largest scale, such as might be performed at major temples on occasions of special weight, was addressed to the fullest assemblage of deities, numbering 3,6oo in all. The list divides the gods into complements of the right and of the left and, on each side, into three orders of protocol (ban) of successively descending rank. The orders of protocol in turn divide into level upon lower level of gods: sheer crowds of Imperial Lords, Heavenly Worthies, Primal Lords, Perfected Lords, Star Lords, Perfected Men, and so on, each belonging to a particular segment of the cosmos or region of the earth as well as to a particular rank and (though this is often unspecified) a particular function. The whole document is closely reminiscent of the protocol lists of earthly officials drawn up in the Song as guides for the positioning of participants in imperial rituals and recorded in the zhiguan (posts and officers) monographs of the dynastic history. What is striking in this list is the gulf that, with few exceptions, separates its immense and carefully ordered panoply of deities from the gods that we know were worshiped by the population at large in the prefectures and counties. The records that survive of the "popular" deities to whom the Song state granted titles of enfeoffment, for example, overlap in only the most occasional way with those in this Daoist listY Certain Daoist figures who were common objects of popular or elite veneration do appear-the patriarchs of the Quanzhen sect, 18 the Lingbao patriarch Lu Xiujing (406-77) 19 -although their rank is far from high. But the innumerable deities of local and regional cults are assigned as a body to some two or three of the 3,6oo "seats" (wei) available. "Gods of the Respective Shrines of the National Registers and Sacrificial Statutes" occupy, with

A JIAO IS A JIAO IS A ?

the anonymity of a crowd, the fifth place from the last in the complement of the left; 20 in the complement of the right the "Gods of the Shrines Within and Without the Capital" fall eleventh from the last, immediately after "City Gods of the Prefectures and Counties of the Empire. " 21 And while the gods whom, in other sources, one finds worshiped individually at shrines and temples locally or throughout the empire are almost always personal beings, with definite identities, names, and previous lives as historical or supposedly historical human beings on earth (this is as true for the Immortals and Perfecteds who emerge from the Daoist tradition as for the shen at large), most of the gods of the Ceremonial Forms pantheon appear as simple positions in a hierarchy, or as functionally defined posts, rather than as specific, personal beings. In sum, this is much the same picture of the divine world that is drawn in the liturgy of the modern jiao, and this picture is wholly absent, again, from the texts of the Huagai Immortals cult. What we do not see, however, in Song ritual manuals is how the liturgy they trace, and its performers, fitted into the larger celebration that from all modern examples one assumes a jiao involved, and that Mr. Wang's promise to hold "a great and splendid affair" suggests as well. While the Daoist practitioners were chanting their texts, delivering their memorials, and so on, where were they standing? When a jiao was offered to specific local or regional deities like the Three Immortals, what was the physical relation of the ritual performance to the temple and/or images of the gods who were receiving the gift? What role if any did the populace of lay worshipers play in the formal ritual? What else did they do while the ritual was going on? None of these questions find answers in any Song source I have seen. Yet it seems to me they are crucial. I am arguing that there was tension or difference between the representations of divinity offered in the texts of the Huagai cult and the representations constructed by the formal liturgy of the jiao and that the tension is evident even in the way the Huagai texts represent the jiao itself. It seems to me likely, in fact, that there should be such tension or difference wherever and whenever the jiao ritual became a prominent part of lay worship and celebration of specific local gods, since as forma/liturgy the jiao dealt with local gods largely by subordinating them. But if there was such tension, and if both the relatively unchanging character of the formal liturgy (which seems to be historical fact at least from the Song on) and its monopolization by professional Daoists prevented the liturgy itself from expressing, mediating, or resolving it, then I would expect to find that tension expressed or worked out precisely in those other aspects of the jiao about which the Song texts

Robert Hymes

139

tell us nothing: in the relation, physical and participatory, of lay worshipers to the ritual, and in the relation of the Daoists and their rite to the deities and their temple. Further, where different participants bring different views of a ritual to their participation in it, and where those views are not only different but different in ways that are potentially in conflict, I would expect them to be worked out or resolved variously in different times and places, depending on the power relations or other disparities of resources between the two parties, and depending on the past and current processes of negotiation, co-optation, monopolization, or domination that have produced the ritual as we now witness it. I cannot attempt to trace any such set of processes here, but I think one can make some progress by examining the products in a variety of settings. For its view of the jiao, the China field has been largely dependent on ethnography done in Taiwan. This work suggests a ritual that is very much of a piece, without much evidence of tension or contradiction. But I hope to suggest that even in this century the Taiwan jiao is not "the" jiao, and that comparing it to jiao in other places can at least hint at the sorts of differential power relations and processes of negotiation and the like that need to be explored, and so strengthen the notion that the tension I am pointing to is real. Perhaps at this point, when I am about to jump from the Song to modern times in seeming disregard of the methodological pitfalls that might lie in the way of using materials from one to talk about the other, I should pause to review just what it is I am doing and hope to do. Above all, I am not trying to trace the jiao through time, to show or to explain change or stability between the Song and this century. Nor, again, am I trying to explain how the jiao itself came to have the position it holds in modern god-worship and seems to have held already in the Song. Rather, I am using modern data to supply a lack in the Song sources, a lack that makes interpretation of what we do have difficult. In the first place I will be exposing a degree of variation, a span of possible forms, in the modern jiao that I think previous work has not remarked, although the evidence for it is clear in published ethnography. I certainly do not, and for my argument need not, assume that just the same range of variation or just the same set of forms existed in the Song. What is interesting is not the details of modern variants, but the principles or issues that may explain variation. Thus, in the second place, I will be proposing that a tension akin to the paradox I have pointed to in the Song sources is clearly present in the modern jiao and, through its different workings-out in different settings, accounts for modern variety: a tension, in fact, between the notion of direct and personal communication with gods on the one hand and the

A ]IAO IS A ]IAO IS A ?

notion of bureaucratic mediation by religious professionals on the other. Finding this tension as a source of variation and an object of negotiation in this century, I will argue, can in turn enrich our reading of the poorer Song materials. I will suggest that the portrayal of the jiao in the Three Immortals sources is itself a strategic move in a long-running argument over the jiao's meaning. Let us begin with the jiao in Taiwan. John Lagerwey's 1987 book Taoist Ritual in Chinese Society and History is an extraordinarily detailed and informative guide to the performance, and the meaning for Daoists, of major Daoist rituals; his study under Kristofer Schipper and his later work with the Tainan Daoist master Chen Rongsheng lend his account something of the authority of an insider's voice. 22 When he tells us in his introduction, for example, that the lay community representatives who pay for a jiao "watch it ... without really understanding what it is they are watching, " 23 one suspects one is hearing the professional Daoist's point of view. What I want to focus on here in Lagerwey's description of a jiao are two elements in particular: first, the bureaucratic character of the picture of the divine world the liturgy projects, and second, the subordination of local and community gods that it intends. None of this will be new to anyone who has read much about religious Daoism, but it is worth treating here, both as background for the rather different material from another part of China that I will draw on below, and as occasion for my own speculations about what might account for certain aspects of Taiwan practice. One document among many Lagerwey reproduces will be enough to convey the bureaucratic flavor of the Daoist priests' written communications to heaven in the course of the jiao. This particular document is an especially nice example because it lists a whole series of other documents the Daoists submit (by burning) to various divinities: Announcement (dispatch!): Office of the Great Method of the Numinous Treasure. This office today worships the Way and prepares an Offering [i.e., a jiao] to give thanks and pray for peace. Being charged with maintaining the order of the Three Heavens, the office-holder [refers to the Daoist officiant himself] is enrolled in the ranks of the five record-keeping bureaus. The antiquity of canon law ensures that sincerity penetrates the high (heaven) and the thick (earth); the luminosity of the ritual texts makes certain that the somber heavens (water) reach both the dark (earth) and the light (heaven). All our earnest petitions must be transmitted by an honorable office-holder if they are to be executed with dispatch. Those who must receive this announcement are as follows:

Robert Hymes -five box envelopes (fang-han) with invocations, invitations, and presentations are respectfully sent to: the Golden Gate in the Jade Capital: respectfully invoked; the gate of the Four Bureaus of the Three Realms: respectfully invited; the Bureau of the Great Sovereign of the Office of Heaven: respectfully presented; the Court of the Celestial Pivot in the Highest Purity: respectfully presented; the Office of the Eastern Peak Tai: respectfully presented. -two box envelopes with letters of presentation are respectfully forwarded to: the offices of the Six Masters of the Numinous Treasure: respectfully forwarded; the offices of the Four Saints of the North Pole: respectfully forwarded. -four box envelopes with missives of offering are respectfully presented to: the officers of merit who transmit symbols, the various gods who assist the Way, the gods of the soil and of the walls and moats, the gods to whom is rendered a cult recognized by law: to each, presented. -seven letters, cards, and notes are respectfully passed on to: the officers, generals, clerks, and soldiers of the various offices of the thunder and lightning of the Numinous Treasure in the Court of the Law of Highest Purity; the various officers, generals, clerks, and soldiers of the altar where gather the gods of the Three Realms in the Chamber of Spontaneity for Communication with the Perfected; the host of immortals who bring catastrophes during the various parts of this year; the venerable god of supreme virtue in charge of the Great Star (T'ai-sui); the venerable gods of the walls and moats of the district and the prefecture of Tainan; the gods and keen-eared ones from above, below, far, and near who receive sacrifices in this neighborhood; the six gods of the soil and directors of destiny of the incense burners of each family. The above documents have all been duly stamped, sealed, and completed. We look to the divine powers on high to transmit them for us. Let the mounted gods gallop each in his own direction and give (his documents) to

A JIAO IS A JIAO IS A ?

the right office, penetrating on high into the Golden Gate of the Jade Capital and below into the bright waters and dark earth. May our sincerity be communicated directly on high and a response be forthcoming. Let there be no delays: it must reach those concerned. The above is announced to the jade lasses of the Three Heavens charged with transmission, the messengers of the roads through the clouds of the Nine Heavens, the officers of merit of Orthodox Unity, the great generals of the molten and fiery hells, the potent officers in charge of memorializing on this day, the clerks mounted on the fiery dragons-all the gods of transmission. In the keng-shen year of heaven's revolution, the I Ith month, the 2oth day, dispatched. In charge of the execution of the rituals: Ch'en Ts'un-hsin (seal)24

Much of this, of course, does not sound especially bureaucratic to secular or Western ears-the dragons, the jade lasses, and so on-but there is a great deal that does: the name of the document itself; the identification of the Daoist priest as an office-holder; the Capital; the carefully graded series of bureaus, courts, and offices, and the careful grading of the sorts of documents that go to each; the box-envelopes, modeled on the boxes in which Qing dynasty officials sent secret memorials to the emperor; and the final date and seal. It is especially important that local gods of the kinds likely to be worshiped at or near the temple where the rite is performed appear (just as in the Song listing of gods for the Yellow Register Retreat) near the bottom: "the venerable gods of the walls and moats of the district and the prefecture of Tainan" and "the gods and keen-eared ones from above, below, far, and near who receive sacrifices in this neighborhood." Indeed, the fact that for these gods no "offices" are mentioned suggests that they may even fall outside-that is, below-the heavenly bureaucracy that is the main concern here; that they represent a kind of "commoner" stratum and so find their place even more distinctly at the bottom of the divine heap. After all, they receive no official box envelope, but simply "cards, letters, and notes." Lagerwey tells us (without elaboration) that "the primary hierarchical distinction is between gods who receive their invitation in a 'box envelope' and those who do not. " 25 If this is a proper reading, then it is worth pointing out that the subordination of the local gods, their distancing from the real structures of celestial authority, is even greater here than in the Song list, which did not set off local deities by any qualitative distinction but merely relegated them to positions at or near the bottom. Now this relative-and relatively extreme-subordination of gods of local appeal finds direct expression in the physical layout of Lagerwey's

Robert Hymes

143

(and other Taiwanese) jiao. The place of central importance in the ritual goes to the five highest Daoist gods: the Three Pure, flanked by the Supreme Sovereign, Jade Emperor, on one side and by the Great Emperor of the Purple Empyrean on the other. The portraits of these five hang at the north, or imperial, end of the temple, facing south as did all Chinese rulers. They are thus the primary recipients of the "offering" that the jiao accomplishes. Lining the walls to east and west, further to the south than these five but still in the northern end of the temple relative to the audience (on whom more shortly), are portraits standing for other recipients: the officers (guan) of heaven and earth. Still farther south on the east and west walls, in the "outer altar" area and so inferior to and serving as aides of the Daoist officiants themselves, are portraits of a variety of other specifically Daoist gods. Finally, in the southernmost, thus most inferior, position of all the deities present at the ritual stand a large number of images arrayed on bleachers, behind a table that holds local offerings (not the Daoists' offerings) to them, the Table of the Three Realms: "here are deposited the divinities brought by the people from their homes, as well as the divinities of the temple in which the ritual is being performed" 26 (my italics). These temple gods (and the others on the table), unlike all the other deities represented in the jiao, by their placement stand in the position of supplicant with respect to the five gods at the north end and in the position of audience with respect to the performance of the Daoists: they are manifestly and specifically not recipients of the offering. This represents a reversal of their usual role and a physical wrenching out of their normal position, since they would usually occupy the position of honor the five Daoist gods have now taken. (Lagerwey does not specify this, but see below.) The only beings present at the ritual whose position is inferior to theirs are the human "representatives of the community," the wealthy donors who have paid for the jiao. These face in the same direction as a subordinate before a monarch-north-but from a position at the southernmost end of the temple, subordinate to the images of their own gods. 27 Lagerwey makes perfectly clear that this represents a deliberate subordination of the temple gods and other locally worshiped divinities, not only to the Daoists' gods but to the practitioners themselves: A re-presentation, as we shall see, of the origin and structure of the universe, the ritual reminds the assembled gods of their proper role as local parts of a great Whole. In a certain sense, this goal is already achieved simply by setting the gods on the bleachers, for the Table of the Three Realms is also called the Table of the Three Officers, and the gods are thus at once placed

144

A JIAO IS A JIAO IS A ?

under the jurisdiction of and assimilated to the Three Officers [heaven, earth, and water, and their respective gods], that is to the deep structure-the order-of the natural universe .... The Offering provides the local godswho are none other than the "obstructed souls" of Feng-tu, the demons of the Six Heavens-with a glimpse, through this doorway, of how the Way works. Just to see it is already to be transformed, humanized, integrated. But they will get no more than a glimpse for ... they are not yet ready to be enrolled on the Daoist's register. 28 Thus the local gods become mere demons, eagerly seeking instruction from the ritual performed by the Daoists and, by their position, themselves worshiping the Daoists' universal gods. Lagerwey's account is especially eloquent on the meaning of this physical disposition of the gods, but others before him had remarked on it and interpreted it in roughly the same way. Thus according to Schipper, for a jiao "the statues of the deities normally worshipped in the temple are removed from their niches, situated at the noble side (opposite the entrance), and placed on the exact opposite side, with their backs turned toward the closed doors. " 29 According to Steven Sangren, "One of the most revealing and central of the many rituals included in chiao involves the physical manipulation of deity images in the temple. All the gods in the temple ... normally face south, absorbing yang. During the chiao, however, they are moved back toward the temple doors, facing north, in the position normally occupied by their worshipers. The gods' normal positions are then occupied by the Daoist gods of 'prior heaven' (hsient'ien) [xiantian ]. These gods, mysterious to nonspecialist Chinese, are thus yang in relation to the regular pantheon's temporary yin status. " 30 Other testimony could be adduced; there seems no doubt we are dealing with a general pattern in jiao in Taiwan. I would like, for ease of reference in the rest of this chapter, to call this pattern of jiao performance the "Daoist-on-top" pattern. I trust the reason is clear: the arrangement of the participants, both human and divine, puts the Daoists and their gods foremost, subordinating to them not only the local gods but the "representatives of the community" as well: both occupy positions to their south; both are mere watchers of the ritual the Daoist performs. While the ritual subordinates the Daoist too to the supreme gods he addresses, his subordination takes a form that simultaneously exalts him above gods and leaders of the temple community. Here physical layout simply confirms the message the words and documents of the liturgy assume: there is a celestial government; the Daoist is a part of it; the "representatives of the community" and, in Lagerwey's jiao, perhaps even their gods stand outside and below it. That the entire

Robert Hymes

145

ritual is at the same time represented as a rite of renewal of the whole cosmos through the reaffirmation and restoration of oneness and harmony cannot obscure the stark, unidirectional, differentiated hierarchy it imposes. As we will see, the Daoist-on-top pattern is not the only way to run a jiao. And this makes it even more interesting to ask why the wealthiest and presumably most powerful members of a Chinese community, those who initiate and pay for the jiao and in fact hire the Daoist officiants, should accept a ritual of this form: why they should willingly act out their own subordination and the subordination of the gods whom they and their neighbors worship daily. One part of an answer may be that there is more than one way of looking at what is going on even in this jiao, and that the "community representatives" do not have to view it in quite the same way as the Daoists do. It is interesting in itself, for instance, that the Table of the Three Realms bears offerings to the local bleacher-gods, to whom the Daoists themselves make no offering. That is, it is built into the very structure of this ritual that the community representatives are making offerings to different gods, gods the Daoists treat only as audience. There is at least a hint of differences of view too in Lagerwey's remark that the community representatives "watch it ... without really understanding what it is they are watching." Perhaps what to the specialist in Daoism seems a lack of understanding is simply a different understanding. It is interesting as well that Schipper tells us that the deities removed to the back of the temple "are said to be 'inspecting the sacrifice,' chien-chiao [jianjiao ]. " 31 The phrase could as easily be translated "supervising the sacrifice," which would suggest a very different picture of the local deities' role here than Lagerwey offers. One wonders just by whom these words "are said," and whether Schipper heard them from the community representatives rather than from the Daoists. Or perhaps there is more than one view even among Daoists themselves. Still, there is a question here. That the north-south orientation symbolizes political authority and subordination is a commonplace of Chinese culture. It would be amazing if the "community representatives" did not recognize that a jiao performed in Daoist-on-top fashion at least appears to subordinate, rather radically, their gods and themselves to religious specialists and the pantheon "mysterious to nonspecialist Chinese" (Sangren) that the specialists bring with them. I would argue that what makes this acceptable to men who would see themselves as community leaders is that it accomplishes a second subordination at the same time. The entire ritual in its Taiwan version, in fact, is founded on a primary act of exclusion. All the members of the community who would normally

I46

A ]IAO IS A JIAO IS A ?

have access to the temple as a place of worship, except the wealthy "community representatives" who have initiated the jiao, are forbidden to enter the temple or to witness the ritual. (This does not mean they are not in other ways part of "the jiao"; on this more below.) Thus, if the ritual reduces the local gods to mere audience, "audience" is here an exalted role in relation to the community as a whole, and those who fund the jiao are licensed to occupy that role and so to take up a position, again in relation to the community, precisely parallel to that of their own gods. We seem to be looking at a sort of bargain between religious practitioner and local powerholder (or perhaps sometimes only would-be powerholder) by which the latter accepts a position in a structure of authority that subordinates him and his gods in return for having his superordination over his own neighbors symbolized in the same ritual. To read the rite this way is not necessarily to claim that such a "bargain" is historical fact (although, since the "community representatives" do in fact hire the Daoists, bargaining of some sort is not simply metaphoric). I do think, however, that we probably cannot understand what happens here without knowing more than we yet know about the past and present working-out in this part of China of power relations, whether religious, economic, or political, between Daoists and local elites and between each and a larger populace. To put things in terms of my original paradox, I am arguing that the jiao and its representations have to be understood as the result of an interaction between different vocabularies of divine-human relations, carried at least in part by different participants in religious life. The form we see it take in Taiwan represents a particular way of bringing together into one ritual structure images of direct contact between individual humans and specific local gods, on the one hand, and images of a universal, impersonal, mediated and mediating bureaucratic hierarchy of divinities on the other. Even in the Taiwanese jiao, both vocabularies receive some expression (consider again the unmediated offerings by the "community representatives" to local gods on the Table of the Three Realms); yet the former is clearly and profoundly subordinated to the latter. I am arguing that we must understand this subordination simply as one possible outcome of the interaction between the two and urging that we begin to wonder about its specific historical roots in political and social contact and exchange, sometimes competitive, sometimes collusive, between Daoist professionals and lay worshipers. To make this plausible, of course, one must show that there are other outcomes. I think such evidence is readily available in work on the jiao in another part of China-broadly, the Hong Kong New Territories region. It is not clear to me why this other work has not so far had as deep an

Robert Hymes

147

impact on historians' and anthropologists' picture of the jiao as the work done on Taiwan. Consider first David Faure's description of a jiao he attended in 1980 in Fan Ling, a cluster of villages in the New Territories whose inhabitants share a single surname. In part of the material I quote below, Faure is concerned with quite different issues than those I am dealing with (in particular, with the relationship between lineage identity and territorial identity), and my quotations here reek of having been torn from context. In one place Faure provides a list of the gods invited to the jiao (in Cantonese, ta-tsiu): It is the practice in a ta-tsiu to invite to the altar all the deities held in

esteem by the villagers in common. On this occasion, the villagers invited the three deities of the village temple who, represented by their statuettes, were given central positions. In addition, they invited seventy-five other deities and spirits, represented by slips of red paper bearing their names. These obviously held secondary positions. Some of these secondary deities, such as the Jade Emperor or the Goddess of Mercy, were not related to any locality in particular. Some, such as the Ch'e Kung and the Hung Shing, were recognized as deities of south China. Others included the protector deities of Fan Ling and the vicinity, such as the rain god of Shui Moon Shaan, the earth gods of Tin Ping Shan, Wutong Shan, Ngau Lo Hau, Ngau T'aam Shaan, the Shui Long Kung, Mau Chau, the walled village of Fan Ling itself, the two other villages just by the village walls known as Naam Pin and Pak Pin, nearby So Kwun Po, inhabited by another segment of the lineage, and an unspecified "brother village," in addition to "the earth gods of the mountains" and "all the earth gods within our territory." Also invited were Governor-General Chau and Governor Wong, the two officials who petitioned the emperor to ask that the coastal evacuation policy in the late seventeenth century be rescinded. The founding ancestor of the lineage was also invited, but was represented as a single slip of red paper among the seventy-five slips .... What is significant here is that this list of spirits and deities represents, in religious terms, the villagers' mental map of their community and its vicinity. 32 (my italics) The centrality given to the villages' own temple deities, not only by position but by their representation by images while other deities are represented by slips, is striking here and contrasts sharply with the Taiwan jiao recorded by Lagerwey, Schipper, and Sangren. The other gods seem to constitute no clear functional or political hierarchy but rather a sort of loose network stretching outward to the larger region and ultimately to China as a whole. But where are the Daoists and their higher gods? In another section of his work Faure tells us: As a religious occasion, the foci of the ta-tsiu are the deities, and they are invited to an altar placed in a matshed especially constructed for the purpose.

A ]IAO IS A JIAO IS A ?

Near this matshed is one for the operas, built so that the stage faces the deities' altar. Nearby is also an altar for the saam-ts'ing . .. [i.e., the Three Pure], the deities of the village priests. This "hall" set up for the saam-ts'ing is appropriately adorned with the village priests' paraphernalia, notably pictures of the ten layers of the underworld and, nearby but not in the "hall," a picture of the Heavenly Master (T'in-sz). Because the "isolated spirits" (ku-wan), that is spirits unattached to ancestral shrines, are also invited to partake of the burnt offerings, an effigy of the taai-sz-wong [ta-shih-wang, "great scholar-gentleman king"] is erected to maintain order, a smaller effigy of the magistrate (also referred to as the city god) stands in a smaller matshed nearby accompanied by two yamen orderlies, and banners (faan-kon) are posted at the major crossroads leading into the site of the ta-tsiu . ... Within the confines so prepared, the rites are executed. In an earlier chapter, it has already been noted that the deities of the entire territory holding the ta-tsiu are invited to the deities' altar, the invitation being a representation of the territory. The presence of the city god, and his surrogate, the taai-sz, blends the territorial deities back into the official hierarchy that reflects the political order. The deities from the territory have been invited to the altar by the villagers themselves, but the effigies of the taai-sz and the city god, along with the representations of the saam-ts'ing set up in a special matshed, are brought along to the ta-tsiu by the village priests. Throughout the ceremonies, unnoticed by the villagers, it is the rites of the priests-such as bringing into the festival the presence of the taai-sz-that place the villagers' deities in a wider setting than the confines of the village. However, to the villagers, and to the priests themselves, the centering of their activities on the matshed of the saam-ts'ing only serves to demarcate them from the villagers. The villagers come into the site of the celebration, insert incense sticks into the burners in front of their territorial deities, and lose themselves in the crowds that stand around and chat, or proceed to the opera matshed to watch the opera, leaving the priests to their own rites. 33 (my italics) This is extraordinarily interesting. One longs for a map or chart: the opera matshed faces the temple deities' matshed directly, but how is the matshed of the "village priests," that is, the professional Daoists, oriented? How does it relate to the local gods and to the celebration as a whole? Part of the difference from Lagerwey's description, too, surely grows from difference of perspective: Faure is largely adopting what he sees to be the villagers' point of view, whereas Lagerwey takes that of the Daoists. Still, this is a very different sort of jiao. The unified, centered structure that in Taiwan emerges from placing the Daoists and their rite at the core, right in the middle of the local gods' temple itself, and assembling there not only the Daoists' gods and the reoriented and subordinated local gods, but the "community representatives" as well, is utterly missing here. The

Robert Hymes

149

local gods are again removed from their accustomed position, but only to be placed in a special matshed that, so far as the villagers' side of the ritual is concerned, becomes the central focus. But "so far as the villagers' side of the ritual is concerned" must be emphasized, since what Faure is really showing us is virtually two separate rituals, coinciding in time and adjacent (though apparently with no clearly meaningful ordering) in space. One is the villagers' ritual, centered precisely on individual prayers offered directly to their local patron gods and on the operatic performance for which the villagers and their gods join as audience. (Note the interesting difference here from the way the role of audience is deployed to construct a complex hierarchy in Lagerwey's rite.) The other is the Daoists' rite, which for all we can tell may resemble a Taiwan jiao, or a Song jiao, very · closely in its formal liturgy, and which, like those, focuses on the Three Pure above all. Most significant for my argument here, it is the Daoists and only the Daoists, as Faure tells us, who bring the symbolic vocabulary of celestial bureaucracy and of a hierarchy of gods into the picture, and they are virtually ignored by the villagers. Here, then, the pattern is not "Daoist on top" but, perhaps, "Daoist alongside." Whatever historical "bargain" may have been worked out here between Daoist and villager is far different from whatever produced the Taiwan rite. And yet there is a sense of a bargain here too: the Daoists have not simply come along, horned in as it were, and performed their own rite at the same time as the villagers are performing theirs, for the villagers do see the presence of Daoists and their rites as somehow essential for a proper jiao: To most villagers, the ta-tsiu is an event that must be realized every so often, usually every ten years, because it has to be done, because it has been promised (to the deities) by the ancestors, and because it is feared that illfortune will follow if the practice is abandoned without the deities' express approval. It is almost characteristic of the ta-tsiu that villagers are not bothered by their lack of understanding of what the priests perform. It matters, and the villagers are confident of the fact, that the priests do perform a set of ceremonies that amount to a ta-tsiu, and that the villagers themselves may integrate into them their own practices. 34

"Integrate" seems, frankly, a considerable exaggeration here: nobody is integrating his rites into anyone else's, as far as I can see in Faure's description, but it is interesting and important that villagers see the Daoists' presence as a requirement for their own celebration. "Lack of understanding" seems to me a less problematic notion here, where villagers (whether leaders or not) are simply not participating in the rites the Daoists

A JIAO IS A JIAO IS A ?

perform, than where they play a direct role as in Taiwan; but it is still an issue worth pursuing below. But the crucial point here is that the notion of direct, personal contact between men and local gods and the notion of celestial bureaucracy and official mediation here simply coexist, as the basis of entirely separate practices, within what at least advertises itself as one ritual, the ta-tsiu. There is neither spatial nor any other representation at the level of the whole, 35 of any subordination of one model to the other, or indeed of any relation between them at all. There is no indication here that wealthy or powerful members of the Fan Ling community are using their relationship with Daoist professionals as a way of symbolically clinching their own authority or dominance through the very structure of the ritual, as I have suggested for Taiwan. Yet the jiao in this case was organized and paid for, not by some spontaneous act of the villagers at large, but by the ancestral trust of the lineage that dominates the villages of Fan Ling, a unit of a kind generally managed by the wealthy and informally powerful of the lineage. This leaves one wondering again what processes of negotiation, cooperation, and competition have generated the present situation, in which Daoists are recognized as essential but do not get to impose their construction of the divine upon the ritual as a whole. Is an organized lineage leadership in a village cluster like Fan Ling simply more secure in its domination, less in need of Daoists' aid for religious confirmation of its power than the individual wealthy and powerful members of a templecommunity in Tainan? Is Fan Ling in fact a more strongly organized community, and does that affect its relations to professional Daoists? Notice also that for the villagers the entire ritual rests upon the promise made by ancestors to the local deities, strongly reminiscent of the personal vows made by Mr. Wang and others in the Verities to hold jiao if they received favors from the Three Immortals. The ritual, that is, commemorates and re-enacts an original direct, unmediated interaction between ancestors and gods. Since the lineage trust stands precisely for the relation of the villagers to these ancestors, one could find it unsurprising that unmediated contact with the gods becomes the focal point of the celebration the lineage trust sponsors. But what, then, is the function of the Daoists and their gods, and how have they been brought to accept the collateral position they hold? What is going on (and has gone on) here? Faure's data are not unique. Francis Hsu, in a book again devoted to entirely other issues {the relation of "magical" to "scientific" thinking), has provided a very useful account of jiao (he spells it chiu) performances in two communities in the far south of China, well separated in space as

Robert Hymes

well as time. 36 Both aimed at protecting the communities in question from plague. The later of the two was a jiao carried out in Shatin, again in the New Territories region, in I 97 5. The biggest temporary poles-and-matting structure (p'eng) constructed for the meeting was the Cantonese opera-house .... Directly opposite the operahouse was a large pavilion honoring Che Kung and his spouse. Part of the role of Che Kung was that of [plague god] ... he was credited with clearing Shatin of its infectious disease at the request of the populace. His was the main shrine, most elaborate in its construction and decoration .... Situated closer to this main temple but still outside its front steps was a large offering table on which were a lantern and three modest sized incense burners for worshippers to insert sticks of incense into at all times. A cushioned platform about one foot high was in front, for worshippers to kneel in prayer or to kowtow after they made the offering of incense. Draping this table was a red skirt with the embroidered words, "Gratitude to Gods for Favors" .... In the inner sanctum [of the main temple] sat images of Che Kung and his spouse, under a red and yellow canopied enclosure. In front of them were two more tables. The smaller one, closest to them, was laden with the usual incense burner, candlesticks, and food offerings. Next to it was a much larger table (about twelve feet by twelve feet) which filled the rest of the hall all the way to the front entrance.-17

Further on Hsu explains the function of the large table: Instead of scripture recitation and chanting, the offerings here [i.e., in the Che Kung temple], aside from the inevitable and constant lighted incense sticks, were mainly food. On [the table] were displayed a variety of polychromatic animals: roosters, fishes, horses, pigs, goats, tigers, ducks, and so forth .... All of these animals were made of flour. On the last day of the chiu meeting, each contributor who gave $5o or more and who wished to do so could take one home. In addition to these lively animal figures, also displayed on this table were heaps of oranges and stacks of pastry and many-layered cakes, red and white roses, and a variety of figurines of gods, angels, and other immortals, also made of flour. All of these items were blessing-givers in anyone's home. The edibles were blessed because they were first offered to (and presumably eaten by) the gods; the other articles because they were associated with them. 38

Thus here, as in Faure's material, the central deity, who has a longstanding direct connection to this specific community, is the focus of abundant personal offerings on the one hand and of an operatic performance, which entertains him and his worshipers together, on the other. This entertainment, Hsu tells us, was "an essential ingredient of any expression of gratitude to gods and spirits for favors. " 39

152

A ]lAO IS A ]lAO IS A ?

Again, of course, there are Daoists, although here they may have more competition than in Fan Ling: Other structures in the chiu ground were a number of side shrines: (1) one for five Daoist deities; (2) one for three Buddhist deities; (3) one for fifty Village Guardian Gods (Chun Shen) [cun shen], lorded over by the Earth God; (4) one for souls of deceased ancestors by descendants who paid to have them uplifted; and ( 5) one for Cheng Huang or District God [the City God], flanked by Earth God (Tu Ti) on his right; on his left was an assistant god named Ta Wu Chang, a figure with white hair and a pointed hat. Of these five the shrine housing the Daoist deities was the largest and the one for Buddhist deities the next largest, while the others were much smaller. 40

Again a map would help; and it would be useful to know whether the images of the more bureaucratic deities-the city god, the earth god, and the like-came to the celebration with the Daoists, as in Faure's community, or were installed by others. It is interesting in any case that their shrines, like that of the five Daoist deities, are clearly secondary in size and centrality (Hsu calls them "side shrines") to the temple of Che Kung and his spouse. But the Daoists themselves center their efforts on their own shrine: "Although Che Kung and his spouse occupied the main temple and the chiu meeting was staged in the name of gratitude for their favors, the priests' ritual activities were concentrated at the side shrine for Daoist deities. Here polychromatic pictures of various Daoist gods were hung on the main wall. " 41 In sum, we find the same separation in Hsu's account as in Faure's evidence between the devotions directed at the local god (here Che Kung and his wife), from which individual worshipers expected to gain personal benefits (not only the blessed foodstuffs to be taken home by those who had been especially generous in their donations, food that in effect was shared with the god, but also the boons that worshipers sought in their personal prayers and with their personal offerings of incense), and the rites of the professional Daoists, directed exclusively at their own deities. During most of the jiao, these latter rites are only one show among many for the lay participants who circulate on the grounds, visiting whatever building they choose and making offerings "to any image of any spirit in any shrine. " 42 Again the jiao as a whole has none of the unified ritual structure of the jiao in Taiwan described by Lagerwey and others. Hsu does not have much to tell us about the content of the Daoists' rites, but it is clear that for most of those attending the jiao the figures of Che Kung and his wife, not the gods or performances of the Daoists, are central. These Daoists, however, do in the end move to the front a bit more than

Robert Hymes

153

Faure's do: on the last night of the ceremony, they lead a ritual and procession to offer charity to the wandering ghosts without living descendants, which Hsu tells us draws an enormous attendance even though the opera is still going on. But there is nothing in his description that suggests that the Daoists' own construction of this rite, whatever it may be, attains any special status in the form of the procession. This jiao is simply not the same sort of centered, hierarchy-embodying ritual that we find in Taiwan. In the Fan Ling and Shatin jiao, then, we see a way of structuring the relation between professional Daoist and individual lay worshiper, between their respective universal and locaVpersonal gods, and between alternative vocabularies of divine-human interaction that seems to leave the elements in each of these pairs on an equal footing, or at least to leave entirely open the question of any hierarchical relation between them, and certainly to leave them looking quite separate, capable of articulating their respective domains of religious action without, at times, much interaction at all. Yet the two vocabularies are equally capable, on the Taiwan evidence, of being structured into a mutual relation that is strongly hierarchical. The differences between the organization of New Territories and Taiwan versions of the jiao suggest that my strategy in treating the two ways of conceiving and interacting with the divine as distinct and at least potentially in tension is a useful one. The New Territories cases in particular confirm that celestial bureaucracy and mediated god-human interaction are most characteristically the baggage of professional Daoists;43 thus that what we see in Taiwan is a jiao that has been powerfully shaped in specifically Daoist directions. To historians of Daoism itself, this last may seem a strange notion, since the jiao is, historically, originally a Daoist product; a jiao "shaped in specifically Daoist directions" may seem simply a jiao in its most natural, original, and thus proper sense. But the New Territories evidence suggests things are much more complicated. Still, one can go further. The physical and spatial organization of the ritual-its structuring, or not, around the performances and gods of the Daoists-is not the only medium in which differing notions or models can express themselves. Here we may turn to Hsu's other case, a jiao he observed in West Town (a fictitious name), a small community in western Yunnan province, in 1942. In this instance, although the community was much smaller and vastly different from the Taiwan cases (in having, for example, no local temple of its own in which to hold the ritual and, as far as one can tell, no strong specific cults to local deities whose images could be repositioned), the fundamental rituals in which both members of the populace and professional Daoists took part were just as firmly centered

154

A ]IAO IS A ]lAO IS A ?

on the Daoists themselves and suggested just as strongly a subordination of the community representatives. Here, after the Daoists issued public notices that represented the relation between the gods to be entreated and the people of the community as entirely a disciplinary one, a series of rites took place, including most notably for my purposes one in which, while the chief Daoist read scriptures and prayers to an image of the plague god (apparently supplied by the Daoists themselves and not a god of local connections), a leading member of the community "knelt just behind him and kowtowed at regular intervals following each chant ... during the entire performance without intermission, which lasted one and a half hours." The performance left the man with tattered trousers and bleeding knees. This is a more grueling demonstration of subordination to the god and to the Daoist himself than that the equivalent "community representatives" in Lagerwey's jiao undertake; and in this and every other public ritual performance the members of the community simply followed the lead of the Daoists. Yet Hsu finds abundant evidence that the jiao had very different meaning for these townspeople than for the Daoists they followed. Perhaps his best evidence, since it involves direct action, is a separate altar erected by the local "managers" of the jiao-again, those responsible for initiating it and for hiring the Daoists-in one part of the town, directly opposite one of the numerous jiao altars erected according to the Daoists' instructions. 44 This special altar brought together a number of Buddhist and other god images, with an incense burner in front for individual offerings and a fountain from which one could drink "angel water" that would save one from illness; it "was created in a moment's enthusiasm by individual worshipers," and no Daoist had any part in its construction or in any offerings made before it. The images were arranged in a way that certainly suggests a hierarchy of a sort: a rough, widerimmed circle of smaller figures surrounded a more robust image of the Buddha. But the hierarchy here was far vaguer than in the god-listings of Daoist liturgy, since no clear ranking among the images is suggested by distance from the center. The altar is a difficult document to read, although the presence of two images of Guanyin, goddess of mercy, and the associations of the Buddha himself with free grace and goodwill, suggest an emphasis rather different from the disciplinary tone of the Daoists' public proclamations. The most that can be said, perhaps, is that townspeopleindeed, the very initiators and managers of the jiao itself-here made a religious artifact whose meaning seems utterly independent of the Daoist

Robert Hymes

I

55

rites in which they were involving themselves, and which like the villagers' half of the rites at Shatin and Fan Ling was the setting for direct appeal, and direct receipt of benefits, by individuals. Hsu's work on West Town, particularly when reread with his and Faure's New Territories evidence fresh in one's mind, suggests that one might profitably examine Taiwan practice anew. For there, too, a great deal happens on the occasion of a jiao that is not directly connected to the formal Daoist ritual. That in the Taiwan case this ritual occupies the physical and symbolic center of the celebration and draws the participation of leading laymen should not, if West Town is any guide, discourage us from looking at all the other lay activities, entertainments, and ritual acts that take place at the same time, as it were on the periphery, and from asking just what vocabulary is being deployed and what hierarchies built or unbuilt in and through these. Yet I think that the systematic, unified, and focused character of the central rite has discouraged such inquiry in the Taiwan case. Lagerwey certainly sees the rest: But a Daoist ritual is not only what goes on inside the temple, it is also what goes on outside, for an Offeiing is a time when ordinary time does indeed stop, and the whole community engages in its own re-creation. Outside the temple there is theater, often two or three competing companies, each with the microphones turned up full blast. There are processions, of youth squadrons dressed up in military garb and making mock charges in front of the temple with staffs and halberds, of neighboring communities who have brought their gods in sedan chairs to pay a visit, of the community itself carrying its gods on an inspection tour.... So much indeed is going on outside the temple that on the rare occasions when the Daoist priests do emerge from their "retreat," no one pays much attention and if someone does it is not for very long-unless it is the Universal Salvation (p'u-tu) ritual [pudu -the same ritual as the one that brings the Daoists to the fore in Shatin] .... One could mention as well the organization of the festival: the armbanded officials selling "stock" in the community and keeping tally on a huge bulletin board of the names of every contributor and the amount contributed; the capped-and-gowned community chiefs arguing over whether to allow a foreigner in the temple and resorting to the divining blocks to assist them in making their decision; the private rituals being performed by Daoists in front of altars erected at the gates of individual homes of the well-to-do, or by barefoot priests in the side rooms of the temple, with buffalo horns for summoning their legions and snake whips for chasing away evil. ... But all this excitement we will have to ignore in order to concentrate on what is going on behind closed doors, in the magic mountain of the Tao. 45

A JIAO IS A ]IAO IS A ?

Thus Lagerwey, clearly the representative of the professional Daoists in this respect, quite deliberately chooses to ignore what happens outside the temple, just as those outside (as he tells us) may choose to ignore what happens inside. Practices like those he mentions have certainly gained anthropologists' attention, but not, I think, in the specific context of the jiao as a whole and not in terms of their specific differences from, and perhaps competitive relation to, the formal ritual proceeding inside the temple. To stop taking sides-to stop ignoring-might make it possible to show that the Taiwan jiao is not in all respects as different from the New Territories jiao as a concentration on the ritual inside the temple can make it seem; and that the hierarchy and subordination that the Daoist officiants seem to achieve there, the redirection of lay worshipers' efforts toward a focus on their own celestial bureaucracy, does not even in Taiwan reach much beyond the bounds of the temple itself, or beyond the circle of "community leaders" who fund the temple ritual. Equally interesting in Hsu's work are the answers he received to his direct questions about the jiao's meaning. Most notably, all professional Daoists and only professional Daoists reported a bureaucratic-mediative interpretation: the plague god was punishing the people with illness at the orders of his superiors, by sending out his own subordinate spirits to do the infecting; to end the plague, one had to appeal to the superiors, who would then order the plague god to end the plague; "the petition papers burned during the chiao meetings were like petitions to our officials in this world. " 46 From the managers and other participants among the local population, on the other hand, Hsu received an extraordinarily wide variety of answers, some having nothing to do with spirits or gods (it should be remembered, again, that the town seems to have had few specific deity-worshiping traditions of its own), some simply invoking custom, some focusing on the charitable character of the personal act of taking part in the ritual for the community's benefit, but none suggesting any notion of a celestial officialdom or of mediation by one god with others. Most interesting among Hsu's list of common categories of answers, however, is this: "What the priests did in these chiao meetings were their professional secrets. They were only paid to do a job. " 47 We have already seen much testimony to the lack of understanding or ignorance of lay people (including the wealthy and powerful "community representatives" in Tainan) before the rituals of the Daoists. It seems to me, however, that Hsu's "professional secrets" class of answers suggests this is ignorance of a special kind. It would be nice to have the original texts of the answers his informants gave, and to know the phrases Hsu is

Robert Hymes

157

translating as "professional secrets" and "do a job." But even without this, I would argue that when a Chinese man of some standing in a community (and these were a major part of Hsu's informants), certainly in dynastic times but even in this century, refers to a class of people as possessing "professional secrets" and as "paid to do a job," he is not simply pleading ignorance. Rather, he is classifying the people he refers to as, essentially, artisans or hired men and thus as people whose knowledge, however extensive, is of a kind not much worth other people's bothering about, especially not much worth the time of a leading man of the community. It is because one does not wish to know much about such things that one hires other people to take care of them; and it is because one hires other people to take care of such things that one does not wish to know much about them. That is, whereas the Daoists wear their special knowledge and its secrecy, the exclusion of others from access, as a badge ofrank, and whereas their texts often treat it as precisely akin to the knowledge that a high scholar-official needs to govern men, those whom they serve always have available (and I am arg~ing, to be sure, without all the evidence that I would like) the equally culturally characteristic move of treating some knowledge as the knowledge of hirelings, which therefore it is not only natural, but right and proper, not to possess oneself. To speak of a form of knowledge in this way is to treat its possessor as one's subordinate-thus, in the case at hand, to reverse the subordination symbolized by the kneeling and kowtowing in the rite itself. Knowing, that is, is in China not always a positive value, and not knowing is not always shameful: everything turns on the kind of knowledge at issue and on the sort of person who normally would have such knowledge. What may be going on here, if I am right, is a sort of mutually coopting agreement by the Daoist and his customer: an agreement that the Daoist knows and will continue to know what the customer does not know and will continue not to know. But it is crucial to the mutual cooptation that the two do not agree on the significance of not knowing, for each by his interpretation of this asserts the subordination of the other. To confuse the customer's side of this bargain with mere passive "lack of understanding," I think, is a mistake and probably seriously underestimates the opportunities open to all lay Chinese, especially but not only the literate, to learn in a general way what the significance of Daoist rites was to the Daoists themselves. (Consider the public disciplinary notices the Daoists issued on this occasion.) The Daoists' customers, perhaps, were not simply ignorant: they ignored.

A JIAO IS A JIAO IS A ?

An alert reader of an earlier version of this chapter has suggested that what I am describing is a sort of "zero degree of resistance," in which power not otherwise deniable or resistable is resisted by being redescribed or simply ignored, and finds this a phenomenon of great potential interest to students of the ways in which, say, peasants or intellectuals may have oriented themselves to the power of both imperial and more recent Chinese states. 48 I am far from hostile to the drawing of such connections, but in conceiving "resistance" one must exercise a certain care. It is easy when talking of power and resistance to imagine that one knows where the power is, that it is all in one place, and that all resistance must therefore be on the other side. But the sorts of cases I have been considering largely comprise situations in which different participants bring different sorts of power to the field, and in which "resistance" is in a sense mutual. It is worth elaborating this point. A gentleman-patron of the Huagai Immortals in the Southern Song who described and dreamed his own relations with the immortals in terms that were direct and personal rather than mediated and bureaucratic may indeed have been engaged in symbolic resistance of the bureaucratic state and its attempts to make bureaucratic relations supreme among political and social relations. Such symbolic redescriptions of ideal power relations may have held great appeal for men living in a world in which significant bureaucratic careers had grown difficult to achieve. This forms a large part of my argument elsewhere about the significance of the rise of the Huagai cult in the Southern Song. But at the same time this gentleman as often as not exercised power that was quite real in his own locality-all the more as projections of state power into the localities were in important ways less vigorous and insistent than they had been earlier in the Song-and was not necessarily at a disadvantage in all confrontations with representatives of the state in that arena. Indeed his accounts of how the Huagai Immortals acted provided a positive model for a sort of power (inherent not delegated, personal not official, direct not mediated, informal not law-governed) that he may have seen as much like his own. When gentlemen and the state acting at the local level upheld different images of the divine, who was resisting whose power? The question is even more acute when we consider relations between gentleman and professional Daoist: without the backing of an organized and centralized clerical hierarchy or a religiously committed state, the Daoist practitioner must often have stood in the relation of mere hireling to the gentleman who paid him to perform a jiao. In such a context, a gentleman's claim that the Daoist was "merely hired to do a job" and his

Robert Hymes

I

59

disregard of the Daoist's own "knowledge" of what the job meant would look not like resistance of Daoist power but like the power to state facts, and one might plausibly read the Daoist's ritual projection of a bureaucratic model as an assertion of a higher, formal, merit-based power able to subordinate or redefine as inferior the merely personal power of the man who was, in fact, his boss. Yet we have also seen in the ethnographic evidence that Daoists are necessary to the performance of the jiao, and this certainly seems to have been true in the Song as well; their necessity, in effect, supplies them with power of their own, power that perhaps expresses itself most clearly in the Taiwan jiao, and that might need in turn to be denied by being redescribed or ignored. My argument from the evidence of both the Taiwan and New Territories jiao is, at bottom, that the lay bearers of a ritual vocabulary of direct human-divine relations and the clerical bearers of a vocabulary of bureaucratic relations are each "resisting" the power of the other. The forms in which the two vocabularies end up coexisting in a single ritual setting grow, I am arguing, precisely out of the different workings-out of this mutual resistance in different contexts. In this view it is not that Chinese do not care what their rituals mean, but that while caring, they must deal with others who judge the meaning, and work to give it expression, differently. In any case, what I would like to retain from Hsu's evidence regarding what informants could tell him about a jiao is this: that the participants in the relation that joins Daoists and lay worshipers may reconstruct the meaning of their mutual relations and of the rituals in which they take part just as surely, at least for themselves, by what they say or do not say about each other and about the rites as through the physical organization or liturgical scripting of the rites themselves. This may hold all the more (given the more public and formal character of writing) for what they write or do not write, again regardless (to some degree) of how the ritual they write about was structured and performed. One way to give meaning to a rite was by changing (or refusing to change) its performance, in effect by renegotiating its meaning with (or imposing its meaning on) the other participants. But another way (and this may have been especially important where different sorts of participants brought interests to the ritual that were difficult to negotiate) was simply to read, or better still to write, the meaning of the ritual in a particular way. I began with texts, the miracle stories in Verities of the Three Lords, which presented jiao simply as objects of exchange, or gifts, between humans and gods, or as mere tokens of thanks or apology from one to the other, without any exploration of what, internally, a jiao was or said or

!60

A ]lAO IS A ]lAO IS A ?

did. I have already complained that there is no other route of entry into what actually was happening in the jiao to which the texts refer. But perhaps this is in any case irrelevant to understanding what the Verities stories have to say. Perhaps I can partly resolve the paradox with which I began by suggesting that when the authors of the tales in the Verities treated the jiao as a mere gift or token, as something defining and defined by the external relations it entered into but without perceptible internal content, they did so because writing in this way (and talking and thinking and dreaming in this way) allowed them implicitly to assert that the meaning of a jiao lay exactly and simply in what they used it for, and not in what somebody else might say or do in performing it: that the jiao belonged to them and to the Three Immortals, and certainly not to the hireling Daoists who merely knew how to perform it.

CHAPTER

6

At the Margin of Public Authority: The Ming State and Buddhism Timothy Brook China is a realm of culture and ritual where people are compassionate and amenable to instruction. A monk who can attain the style of the Patriarchs need only explain Mahayana and the people listen and understand; he need only speak of karma and the ignorant are transformed. -the Hongwu emperor, 13 71 Those you find walking the Buddhist path in recent years are.not wise men but corrupt, unregistered characters who have changed their names and gone into Buddhism to evade trouble and save their own necks. -the Hongwu emperor, 1394

i/'omething happened between 1371 and 1394· The Hongwu emperor (r. 1368-98) had launched the Ming dynasty (1368-1644) striking the pose of the sage-ruler. His goodness and compassion, he had believed, would transform his subjects, and his reign would be buoyed by all faiths, including Buddhism. Twenty-three years later, that pose had fallen apart. Benevolence had been superseded by disgust, declarations of compassion replaced by threats and denunciations. The difference between 1371 and 1394 had something to do with changes in the composition of the clergy, but it had far more to do with a change in the norms that the Hongwu emperor believed should govern his realm. With the change in norms came also a change in state regulations governing the institutional life of Buddhism. Between 1381 and 1394, the emperor put together an edifice of laws designed to subordinate monks and monasteries to the complete authority of the state. Buddhism would spend the rest of the Ming dynasty adjusting to, and growing out from under, the burden of these regulations.

J

THE MJNG STATE AND BUDDHISM

The relationship between Buddhism and the state would shift again early in the fifteenth century, and yet again in the last quarter of the sixteenth century, as we shall see. These shifts were less violent than the mid-Hongwu assault on Buddhism and are only weakly reflected in state texts, but they are equally revealing of the instability in the state's relationship to organized religion. Seen together, these three shifts provide an opportunity to explore the evolution of the relationship between social institutions and public authority over the course of the Ming dynasty. More specifically, this chapter tracks monastic Buddhism's responses to shifts in state norms in order to probe changes in the variable constitution of public authority between the founding and late phases of the dynasty. I choose to examine the relationship between Buddhism and the Ming state in terms of public authority, rather than within the narrower terms of state control. The distinction between public authority and the state I find useful for detecting change. Whereas the Chinese state remained relatively constant during the Ming, public authority underwent redefinition. That redefinition can be used to narrate significant aspects of sociopolitical change between the late fourteenth and late sixteenth centuries. 1 Consider the difference. The state is represented or "made public" as the main source of formal authority in China in the local arena by local officials. The concept of public authority, by contrast, expresses the reception of the state's authority in that arena: public authority exists to the extent that people are aware of and respond to the presence of the state. 2 Not defined by the state alone, public authority takes form through the interaction of state and society. The inevitable instability of this interaction means that public authority is vulnerable to redefinition, notably when powerful elements within society emerge to participate in, or alternatively to challenge, state dominance in the local arena. Public authority emanates from the state, but as it emanates into the public realm it can suffer deflection or partial expropriation as new political possibilities take form. The power of elites to deflect state hegemony is not an issue in the early Ming, when the state could and did intervene to regulate social institutions largely at will. But as local gentry sought a localist reorientation of power in the late Ming, they used monastic Buddhism in their project to redefine public authority. It is this understanding of public authority as vulnerable to redefinition, and of institutional Buddhism as implicated in this process, that sets the terms within which this chapter looks at the effect of changes in state norms on Buddhism. Looking only at the regulatory texts, one might be tempted to assume that the Ming state's relationship to Buddhism was straightforwardly repressive: the mid-Hongwu regime formulating tough laws designed to

Timothy Brook

limit the size and influence of Buddhist institutions, and those institutions bending to the state's will. Indeed, the sense of an uncomplicated compliance with state regulations has been so thoroughly worked into the historiography of Ming Buddhism that few historians of the late imperial period, except for scholars of Buddhism, are even aware of Hongwu's assault on Buddhism, let alone its unprecedented scale. 3 It is worth pondering why the Hongwu suppression is relatively unknown, for it anticipates what this essay seeks to expose. The suppression's invisibility derives in the first instance from the manner in which it was announced and officially recorded. No Ming emperor declared Buddhism to be a threat to the state or to public authority; not even Hongwu said he was suppressing Buddhism. Rather, the regulations that he and subsequent emperors promulgated are suffused with the logic of bureaucratic rationalization-a matter not of shattering delusion or stamping out sedition, but more of adjusting resources to needs. If this logic passed scrutiny at the time, it was due in the second instance to Confucian complicity. Aspiring officials anxious to resecure their place in the political system after the difficult passage through Mongol occupation were content to see Buddhism moved to the margin of public authority. The sanctity of the Hongwu dispensation for the rest of the dynasty meant in the third instance that what the founder in his great wisdom had decreed could not be questioned or otherwise commented on. Accordingly, the very notion that the Hongwu regulations on Buddhism amounted to a suppression disappeared beneath the fiction of the wise ruler and the gratefully compliant subject. Chikusa Masaaki's research on similar regulations in the Northern Song dynasty has shown the consistent failures of state regulations to induce their intended effects to be more interesting, and infinitely more informative of the state-society relationship, than their limited successes.4 State regulations may be designed to impose norms, but they cannot determine how these norms will interact with the social practices to which Buddhism is tied. Subject to the supervision of the state, Buddhism as a social institution was vulnerable to severe restriction; yet by not being the official religion of the state, it remained relatively free to override the effects of state suppression, remake its social constituency, and, in the late Ming, reposition itself in relation to public authority.

Buddhism in the Opening Decade of Ming Rule The founding Ming emperor understood the presence of Buddhism in the lives of his people. Not only was Zhu Yuanzhang, who became the Hongwu emperor, abandoned to the care of a Buddhist monastery as a youth,

THE MING STATE AND BUDDHISM

but during his rise to power he was able to make effective use of Buddhist devotional networks to mobilize his followers for war. He was also aware that whereas these networks could survive wartime calamities, institutions were more vulnerable to destruction. The Yuan-Ming transition, particularly in its early phase in the r 3 5os, took a large toll on Buddhist monasteries. As a later writer from Yangzhou noted, in the areas that warring armies overran at the end of the Yuan, only 20 to 30 percent of monasteries survived the dynastic transition intact. 5 And those that did survive found it difficult to hold on to the income-producing land that had sustained the monks who once lived there. For the first decade of the Hongwu era, Buddhism was regarded as a problem only to the extent that its dilapidated monasteries stood as symbols of the destructiveness of the dynastic transition. The emperor responded during this decade by adopting the benign posture of imperial patron and sponsor of reconstruction. He funded monastic rebuilding and bestowed large tracts of tax-exempt land in the capital region, and he brought monks into political circles, sending them on diplomatic assignments (to Turfan and Japan, for example) and appointing them as advisers in the courts of his princes. 6 He called practically every eminent monk in the country to Nanjing for an imperial audience and over the first five years of his reign asked them to conduct plenary masses at monasteries in the capital to rest the souls of the war dead. The Grand Plenary Convocations of the Dharma Flower (Guangjian fahua hui) held in Nanjing in 1371 and repeated the following year were the greatest gatherings of leading Buddhist clerics since the Tang dynasty.? The emperor's Six Dynasties-style imperial patronage cast Buddhism in the role of adjunct to a state-centered structure of public authority, almost an official religion. It was charged with both regulating its own affairs and serving as a pillar of the state. The only significant legislation of the decade was promulgated early in r 3 73, 8 when the emperor required monks to obtain ordination certificates from the Central Buddhist Registry in Nanjing. Certification had been a standard practice of the Chinese state for the preceding six and a half centuries. 9 Hongwu's intention in reviving the system was to end the dislocation, vagrancy, and mendicancy of the war years when many adopted the guise of a monk to beg for sustenance or evade military service. It was also designed to halt tax evasion, since monks could claim exemption from the corvee duties incumbent on ordinary people. An innovation of the 13 73 legislation over earlier registration orders was to institute a system of universal registers (zhouzhi wence). These registers, which the emperor remandated in 1394, were to

Timothy Brook

list the names of all certified monks in the realm, and copies were to be distributed to every monastery so that monastic officials could confirm the status of visitors claiming to be monks; not surprisingly, there is no evidence that such unwieldy volumes were either made or distributed. The important feature of the I3 73 legislation, however, is that no restrictions were placed on ordination, nor did the state arrogate the power to ordain. Ordination was still to be done by senior monks at ordination platforms in leading monasteries. This was simply a registration system, not a scheme to restrict the size of the clerical population. Buddhism After 13 8o In I3So, Hongwu carried out a thorough purge of Hu Wciyong and his faction at court and reorganized his regime so that he was personally in charge of the day-to-day running of the government. 10 The transformation of the emperor's images of himself and his realm brought about by this crisis affected his attitude toward Buddhism. There was, it seems, no particular reason why the emperor abandoned the patronal relationship of the first decade in favor of surveillance and increasingly coercive measures: the new orientation was simply part of the sea-change that affected nearly every aspect of rule. After I3 So, Hongwu no longer conceived of Buddhism within the state-centered structure of public authority; it became instead for him a destabilizing, autonomous realm that, if left unchecked, could only undermine public authority. Buddhism was no longer a resource for ruling but a threat to it. Monks were no longer men of wisdom but charlatans and draft dodgers whose very existence symbolized the failure of that authority to take hold. A sequence of ambitious policies and regulations in the years immediately after I 3 So was implemented to reorganize Buddhist institutions so as to forestall that threat. The first step was to set up a bureaucratic structure under the Ministry of Rites to supervise the affairs of Buddhist monks and monasteries, rather than leave these affairs in the hands of monastic officers responsible solely to their own institutions. This bureaucratic structure consisted of Buddhist (and Daoist) registries at county, subprefectural, prefectural, and national levels staffed by local monks. A few prefectural and subprefectural officials had followed Yuan precedent and set up registries in the early I 370s to manage religious affairs. 11 The office was not universal, however, until I3 SI, when Hongwu mandated a nationwide hierarchy of registries at every level of the field administration.12 These registries were known at the prefectural level as Senggang Si

r66

THE MING STATE AND BUDDHISM

(Office of Clerical Supervision), at the subprefecturallevel as Sengzheng Si (Office of Clerical Rectification), and at the county level as Senghui Si (Office of Clerical Convocation). The registrars at the three levels went by the titles dugang (supervisor), sengzheng (rectifier), and senghui (convener). An identical structure of offices was instituted for Daoism as well. The Buddhist registries were usually housed in the most prestigious urban monastery in the local seat of government, which local officials sometimes restored when the registry was set up. They were founded throughout the country over the next four years, 13 and it was a rare county that never set one up. 14 The registrar was usually also the abbot of the registry monastery, although the prior (jianyuan), the monk in charge of a monastery's business affairs, might also hold the office.U According to a summary in a sixteenth-century county gazetteer from Zhejiang, a registrar's duties were fourfold: to supervise the Buddhist monks in his prefecture or county, to propagate the correct teachings of Buddhism, to report misdemeanors among the monks for investigation and punishment by civil officials, and to conduct public rites. 16 He could also be called upon to grade and register monastic land for taxation purposes. 17 Registrars in underdeveloped regions might bear additional responsibilities. In the frontier areas of Yunnan province, for example, registrars were ordered to see that monks were evenly distributed in small chapels along main transportation routes (a rate of one chapel per fifteen kilometers was suggested) so that they might extend a stabilizing influence over the countryside. The registry system thus served to supplement state supervision in an area where schools and academies were almost nonexistent. 18 These various tasks indicate that the registrar's function was to administer Buddhism on the state's rather than on Buddhism's behalf and, where the state's presence was weak, to embody public authority. The registry system gave Hongwu the bureaucratic means to intervene in institutional Buddhism systematically. This he did over the following years. His first step in r 3 8 2 was to have the local offices register and categorize all Buddhist institutions in the country. As part of this program, Hongwu redefined Buddhists into three sects: Chan, Doctrine (jiang), and Teaching (jiao). According to the edict of 1382, Chan monks were to concentrate on meditational exercise leading .to personal enlightenment. Doctrine monks were to study the scriptures to penetrate their meaning. The Teaching sect, created at Hongwu's initiative, consisted of monks who went out among the people to preach and conduct rites, especially funerary ritesY There appears to have been no doctrinal reason for this

Timothy Brook

bureaucratic simplification of the schools of Buddhism. It was simply a matter of imposing uniformity. By the mere presumption of defining and fixing sect categories, however, the state claimed a new prerogative of dictating doctrinal currents within Buddhism or subordinating doctrine to state uses. Buddhism was being shaped more and more into a creature of the state. Cutting even closer to the heart of institutional Buddhism than the control of sect identity was the control of property. Without sizable tracts of land, large monasteries could not hope to maintain their physical fabric or support their residents. Hongwu did not aspire to take direct control of the fiduciary property that monasteries owned, but he did desire to make sure that others did not get that control. In 1382 he decreed that Buddhist and Daoist monks had no right to mortgage or sell the land on which they lived and that the mortgagor or buyer be penalized by confiscation of his entire landholdings. 20 To prevent monastic land from being alienated, he ordered the Ministry of Rites, possibly in 13 86, to require every monastery in the country to appoint a "foundation cleric" (zhenji daoren). This monk's task was to oversee the monastery's financial affairs and be responsible for the monastery's fiscal dealings with the local magistrate. Although the emperor noted in 1393 that people were filing legal suits against foundation clerics for mismanaging monastic funds, he nonetheless reaffirmed the post in an edict the following year. 21 Hongwu's intention in ensuring that monasteries had sufficient means of support was not to control monastic resources but to ensure that clergy could afford to stay within their monastic confines and not have to seek support outside, where they might meddle in community affairs or build networks among the people. The seclusion edict of 1394, to be mentioned shortly, confirms the impression that the emperor's prime concern in regulating property was to provide the means for segregating the secular and religious realms. In an economy in which land was a commodity and traded on markets that the state could not control, this sort of regulation proved practically worthless. Nine years later, in 1391, Hongwu launched what amounted to the most severe suppression of Buddhist institutions of his reign. Through what was known as the Hundred-Day Edict (because its provisions had to be carried out within a hundred days), he imposed an amalgamation order on Buddhist monasteries. This order required the majority of smaller monasteries throughout China to close and their residents and property to be transferred to a limited number of larger institutions. These institutions were designated "abbeys" (conglin, a southern term for a large

r68

THE MING STATE AND BUDDHISM

monastery). The only places that amalgamation left alone were tiny chapels with only one or two resident monks, and convents for nuns. This amounted to a complete reorganization of the institutional life of Chinese Buddhism. It was largely completed within the hundred-day limit, although some amalgamations were still being carried out as late as I4I 5. 22 A few institutions managed to avoid amalgamation, but they were the exceptions.23 Even abandoned monasteries had to be amalgamated to forestall their revival at a later date as separate institutions. 24 Thenceforth, only abbeys had a legal right to exist. The I 3 9 r amalgamation altered institutional Buddhism in China more thoroughly than any previous suppression, and there would be nothing like it again until the I9 sos. 25 It was presented in contemporary documents of the period not as a suppression of Buddhism but as a means of achieving a more efficient use of resources. This tack was not totally unreasonable, for many monasteries had been destroyed and abandoned during the Yuan-Ming transition and the number of monks qualified to reside in them had fallen since the Mongol dynasty. On the other hand, the campaign forcibly reduced the number of Buddhist institutions to a fraction of the number that had existed before I 3 9 I. Among other effects, the amalgamation order made available buildings that the state and local authorities could put to other uses. 26 Pressure was brought to bear to have the order countermanded as soon as Hongwu was dead. Even before the Yongle emperor's enthronement in I402, he allowed recognized monasteries founded before I3 82 to revert to independent statusY The I39I amalgamation order included yet another mechanism of state control, for it made abbeys eligible to receive an official name plaque (e) from the emperor if they did not already possess one. This was a wooden plaque bearing the name of the monastery scripted by the emperor or a calligrapher he deputed. It was a legal document in that it signified official recognition of a monastery. The edict of I39I made its possession a test of legality, and monasteries founded or rebuilt subsequently were ineligible to receive one. 28 What was in effect a ban on the private founding of monasteries or chapels after I39I was incorporated as such six years later into the Ming Code. 29 The Yongle emperor reaffirmed the status of the name plaque as a legal license in his edict of r402 partially reversing amalgamation. Later emperors would take it upon themselves to overrule Hongwu's ban on the founding of new monasteries by presenting a plaque, and officials and abbots could submit memorials requesting an edict of conferral. Submitting a memorial of application could be a politically delicate maneuver. Ming emperors occasionally preferred to issue blanket

Timothy Brook

refusals, as the Chenghua emperor did in 1467 when he told his eunuchs, who were busily building monasteries around Beijing, that he would entertain no further requests for name plaques. He reminded them that it was illegal to restore or enlarge a monastery lacking such official designation. 30 The Hongwu emperor's last major intervention in the realm of Buddhist affairs came in I 3 94, when he promulgated what could be called his edict of seclusion. By imperial fiat, monks were ordered not to beg in public, enter a magistrate's office, have communications with gentry, form friendships with officials or commoners, or accept minors as acolytes. Most notably, monks had to live in large monastic communities (which amalgamation had brought about) where they could be better supervised. A monastery could house fewer than twenty monks only if it were more than twenty li (ten kilometers) from a settlement; that is, well beyond the scope of regular contact. Monks who chose to engage in religious practice in completely secluded locations could do so on condition that they did so individually or in pairs; three monks together was too many. 31 The goal of the edict was to seclude monks from secular life, although not so completely that they could gather to hatch plots to undermine state authority. The realms of religious and secular life were thus to be neatly separated, with the intention that the influence of the former on the latter be kept to a minimum. The state would no longer look to Buddhism as an aid to governing the people; instead, it viewed Buddhism as a competing source of authority. The edict of seclusion was as much a tirade as a policy statement, and I have found no evidence that any of its strictures was ever enforced. Nonetheless, it established a tone that poisoned official attitudes toward institutional Buddhism for decades. Hongwu's initial posture of patronage had expressed a belief that Buddhism was amenable to incorporation into state institutions and could supply the state with ideological and educational services. By contrast, the amalgamation and seclusion edicts indicate that he had abandoned this belief. After the late Hongwu reorganization, Buddhism was excluded from playing any role in the composition of public authority, ideological or otherwise. It was now simply an object of that rule, representing a sphere of activity subject to state regulation and excluded from the realm of the political. It had been moved to the margin of public authority.

State and Buddhism in the Fifteenth Century The editor of a r 5 66 county gazetteer from Shanxi province regarded the 1391 amalgamation order as the Ming state's last restrictive legislation

170

THE MING STATE AND BUDDHISM

against Buddhism. He noted that, after 1391, the state engaged in no further suppression of Buddhist personnel, literature, or places of residence.32 This is not precisely true. Hongwu's heirs in the fifteenth century did add restrictions in two new areas: on the number of novices, and on the amount of land monasteries could own. As I shall show, however, these restrictions lacked the force of Hongwu's edicts. The Ming state's relationship to Buddhism settled into one of tolerance rather than suppressiOn. The main intervention associated with the Yongle emperor was an edict of 1418 that capped the number of novices per prefecture (4o), subprefecture (30), and county (2o) and required them to train for five years before applying to the Central Buddhist Registry for ordination. 33 This capping was the first Ming attempt to impose a quota system on monks, analogous to the restriction on the numbers of students registered in county and prefectural Confucian schools. 34 Hongwu had imposed some restrictions regarding age and competence in order to ensure that the ranks of the clergy were not filled by imposters, but not with the precise intention of reducing the number of monksY Yongle and his heirs would gradually intensify these restrictions in order to reduce the size of the clergy. Their concern was neither security nor ideological control, but revenue. Since monks were eligible for exemptions from the service levy, the number of adult males enjoying this exemption affected state income. 36 Despite the obvious fiscal logic of this restriction, Yongle's numbers were quickly ignored. The mechanism for supervising them, the registry system, could not be relied on to enforce the new limits on novices. The notion that there should be quotas nonetheless came to be accepted. In response to a memorial submitted late in the 1450s, for instance, the Tianshun emperor ordered that "every prefecture, subprefecture, and county fix its quota of Buddhist and Daoist monasteries and set its number of [ordained] Buddhist and Daoist monks" at current levels. These quotas were to constitute ceilings on the numbers of monasteries or monks permitted within each jurisdiction. 37 The order clearly failed, for a memorialist in the 15 30s suggested specifying limits of 200 monks per prefecture, roo per subprefecture, and 6o per county. 38 Although these numbers too went by the board, the desirability of quotas was voiced as late as rs8s. 39 The other type of restriction imposed after Hongwu was on landholding, again largely for fiscal reasons. The Hongwu emperor had been concerned that monasteries have enough land to keep their monks on the grounds. His heirs tended to the opposite fear, that monks were controlling too much land, blocking peasant access to land, and evading taxes.

Timothy Brook

I7I

Their attempts to limit monastic landholdings were piecemeal, unsystematic, and short-lived. The Jianwen emperor accepted a memorial calling for a ceiling of five mu per monk in Jiangnan in qor, but the Yongle emperor overturned it the next year before it could be applied. In Fujian, a limit of 6o mu per monastery was set in I4 52 and then increased to roo mu in 1480, although neither provision appears to have been enforced. 40 A different approach to restricting monastic landholding was tried in 1448, when local officials were ordered to expropriate property acquired by monasteries after the Hongwu era and distribute it to landless civilians at a rate of 20 mu per adult male householder. 41 This expropriation, applying only to land purchased after the Hongwu reorganization, was a compromise between limiting the economic independence of monasteries and respecting the institutions established under the dynastic founder; I have found no evidence that local officials ever carried out the plan. A wave of state interventions in monastic landholding did occur in Fujian in the latter third of the sixteenth century when the state needed emergency funds to meet military costs, but this involved the expropriation only of rent, not of the land itself. By the sixteenth century, the state recognized that landowning monasteries enjoyed the same rights of private property and rent-taking as any other landlord, and it largely restricted its expropriation to rents to monastic lands that active monasteries were no longer collecting. More to the point, as we shall see, the sixteenth-century state was no longer interested in regulating Buddhism. Taking over monastic rents was a purely fiscal expedient. As for Hongwu's original regulations, while the letter of the laws remained in force, their spirit was compromised soon after his death in 1398. Only extraordinary bureaucratic vigilance could have ensured that his rules were observed, and such supervision was not available. Hongwu had hoped that the registries would serve as an internal supervisory mechanism monitoring Buddhism on the state's behalf. But the design of the registry system did not favor success. With the exception of the prefectural registrar, incumbents were below rank 6 in the bureaucratic hierarchy and hence "were not within officialdom" (wei ruliu). In lists of county officers in local gazetteers, the Buddhist registrar appears after the instructor at the county medical school: both posts were filled with local appointees rather than personnel from outside the province, as was the practice for all the posts above county medical instructorY Registrars thus were yuan (functionaries) rather than guan (officials). 43 In addition to being local people without prospects of advancement into the regular bureaucracy, registrars were monks and could be presumed to have interests

THE MING STATE AND BUDDHISM

that coincided more closely with those of the clergy rather than those of the state. From the state they could expect little more than their monthly stipend, and even that was suspended in the second quarter of the sixteenth century. 44 Their official status meant that they could gain the ear of the magistrate to argue on behalf of troubled monasteries-appealing for the return of expropriated monastic property, for instance-but it provided them with little other opportunity. 45 It had been Hongwu's intention that a Buddhist registry, a Daoist registry, and a Confucian shrine in every county of his empire would constitute a tripod of ritual support for the state's dominance of local society. 46 The system was inadequate to the task. By the mid-Ming, many registries were defunctY The rest survived largely in name only, the registrar being whoever happened to be the abbot of the registry monastery. An individual monk might exploit the post of registrar to act on behalf of the local Buddhist community, as the registrar ofNeixiang county, Henan, did in I48 I when he obtained state funds to rebuild Jianfu Monastery; 48 or he might act in the conspicuous service of the state, as the subprefectural registrar of Yingzhou, South Zhili, did in the early I5oos when he conducted an annual ceremony wishing long life to the emperor;49 or he might seek to express public anxieties, as Hangzhou Prefectural Registrar Liangjin did in I542 when he prayed to Guanyin to end a drought. 50 But in any of these capacities, the registrar's structural power was limited. The registry system did not constitute an internal organization for Buddhism, nor was it well integrated into the regular bureaucracy. Accordingly, in many parts of the country by the turn of the sixteenth century, registries had either disappeared 51 or existed only on paper, and registrars were no longer being appointed. 52 Maintaining the system depended on local initiative. On the magistrate's part, the main impulse for preserving a registry was to keep up the appearance of having all the bureaucratic offices mandated by the Hongwu emperor; on the registrar's part, it was to call upon whatever prestige still clung to the post in local eyes. Hongwu's amalgamation order was similarly compromised. As the editor of the Songjiang gazetteer of I 5 I 2 notes, "When Buddhism was reorganized in I39I, smaller chapels and cloisters were amalgamated into what were called 'abbeys.' In I4o2 all the amalgamated monasteries were ordered restored to their former statuses." Yet the editor continues to group the monasteries in his gazetteer under their I39I abbeys "so as to preserve the old system, " 53 suggesting that the reorganization had become a fact of the local Buddhist landscape. If Hongwu's amalgamation was compromised, it did have the effect of hurrying the demise of weaker

Timothy Brook

173

monasteries. The deamalgamated institutions may have had their statuses restored, but not their buildings and staff. Having been deprived of personnel and resources for over a decade, or having been converted to other uses, many proved impossible to revive. A scan of the section on monasteries in the r56r gazetteer ofWujiang county, Suzhou, reveals that over half the monasteries and chapels which had undergone amalgamation in 1391 were defunct. 54 In addition to allowing most of the late Hongwu legislation against Buddhism to lapse, fifteenth-century emperors occasionally adopted the early Hongwu pose of patron, on the assumption that Buddhist beliefs could still be invoked to buttress public authority. But the revival was guarded. Yongle spoke favorably in his preface to the widely disseminated Buddhist tract Zhufo shizun rulai pusa zunzhe mingcheng gequ (Song of the names of the world-honored Tathagata, bodhisattvas, and arhats) of the conspicuous power of Buddhism "to encourage the people to do good"; yet he rejected a memorial for an amnesty honoring the presentation of a Buddhist relic by a foreign envoy in 1406, recalling that the Buddhist excesses of Northern Wei and Yuan emperors had interfered with sound government. 55 The Chenghua emperor had a reputation for being indulgent toward Buddhism, the Tianshun and Hongzhi emperors for being strict (though even Tianshun handed out 40 name plaques for new monasteries in the first year of his reign). 56 In general, these emperors preferred the role of patron, prizing what values could be squeezed out of ideology without paying the costs of organization. The patronal pose allowed them to publicize their own virtue and their expectation of good conduct on the part of their subjects; it was inexpensive and noninterventionist. We should be cautious, however, about interpreting this patronage in terms of state norms, for imperial patronage could serve concrete political purposes. When Tianshun in 1458 conferred an edict of protection for monastic properties on Shanxi's Wutai Mountain, he did so, he declared, because "for many years the monks have prayed for the fortunes of the state above and the people below. " 57 This statement suggests an alignment between Buddhism and the state, yet other considerations were uppermost. Wutai Mountain was the Lamaist site in China most revered by the Mongols. Tianshun, in his earlier identity as the Zhengtong emperor, had been captured by the Mongols in 1449, and as of 1458 had only recently forced his way back on the throne. The bland vision of Buddhist monks buoying up the state with their prayers thus masks a diplomatic initiative directed at his recent captors. In this case the emperor was manipulating

174

THE MING STATE AND BUDDHISM

the imperial discourse of protective Buddhism and grateful state to communicate a claim of authority to the Mongol world. In contrast to their emperors, mid-Ming officials tended to focus not on ideological claims but on administrative problems that arose because of the declining regulation of Buddhist institutions. From the mid- to late fifteenth century, it became commonplace to complain about monks: their itinerancy, their nonregistration, their parasitism, their alleged licentiousness, their affront to Confucian moral values. 58 The tirades that high court officials directed against Buddhist monks through the mid-Ming suggests not only that they recognized the failure of Hongwu's seclusion order but also that they gave up on attempting to use Buddhism to gain ideological ground among the people when they feared fiscal ground was being lost to the clergy. Of course, most of the mid-Ming fiscal problems that stemmed from the clerical service-levy exemption were state-induced. The Hongwu emperor had specifically forbidden the sale of ordination certificates when he revived the certification system in I 3 7 3, but later emperors, starting with Jingtai in I45I, resorted to this hoary device to raise revenue outside the regular taxation system. The certificates offered in I4 5 I were sold to raise grain to relieve a famine in Sichuan. The precedent was repeated in I4 53 and I4 54 and within two decades had become a standard means for raising emergency funds. At the turn of the sixteenth century, the cost of an ordination certificate was between eight and ten taels; by I558 the government was selling it at a four-tael discount; in I572 the price was officially lowered to five taels. 59 Thereafter, Ming sources fall silent on the subject of certificate sales. The mechanism appears to have ceased functioning. Scholars of Buddhist history have regarded the sale of ordination certificates as having corrupted the system of clerical certification, debasing the quality of the clergy, and giving Buddhism a bad name. 60 The assumption behind this judgment-that the men who bought the certificates went on to become incompetent monks-! regard as mistaken. The polite fiction surrounding the sale of certificates was that the purchaser was a monk. In fact, most if not all purchasers were simply paying a flat fee for a permanent tax exemption. The assumption that purchasers were monks rests on a misapplied comparison with the sale of Imperial Academy studentships (jiansheng}, which also started in I45I as an emergency measure to raise funds for the defense of the northern border, where the Zhengtong emperor had fallen into Mongol hands two years previously. 61 Some jiansheng did go on to take up positions in the Imperial Academy and seek to advance

Timothy Brook

175

into the bureaucracy, whereas few, if any, who bought a monk's certificate as a lump-sum prepayment of future service levies were interested in becoming monks. Who would want to buy his way into such an unlucrative profession? If the sale of certificates impinged on real monks, it may have been to make it possible for a novice who could not get a certificate through regular channels to buy one. The sole possible effect of this practice on the Buddhist community thus was to provide novices with a legitimate, if expensive, means for acquiring certification, not to let loose a mob of imposter monks on society. Once the Ming began to sell certificates, in the words of an early Qing commentator, "there was thereafter no fixed system for issuing ordination certificates. " 62 With the certification system thrown open to other uses, this last surviving structure of control over monks had to give way. We should not overemphasize the impact of such policies on the state of the clergy, however, for as early as I4 3 6, prior to the commencement of sales, the minister of revenue regarded the certification system as no longer functioning. "3 What is more significant about the selling of certificates in the mid-Ming is that it indicates that the state's sole real concern was the fiscal affairs of monasteries. Fears about state security or clerical autonomy had faded. Not just certification, but most other aspects of state regulation, were abandoned. Thus in I 530, a circuit censor in Fujian could acknowledge that all edicts regarding Buddhist institutions and personnel issued between I373 and I45 5 were being disregarded. 64 The state regulation of Buddhism had become a fiction. State and Buddhism in the Late Ming Throughout the sixteenth century, complaints about immoral monks running riot across the countryside fade from the public discourse regarding good governance 65 and with them the desire to restrict monks and monasteries. The Ming state did not renounce its supervisory function, however. When the famous pagoda at Baoen Monastery outside Nanjing caught fire in a lightning storm in April I 5 66, for instance, the abbot and eighteen other monks were thrown in prison for negligence when the matter was reported to the court. Only the unrelenting efforts of Master Hanshan Deqing, who went back and forth between Baoen and the southern Ministry of Rites for three months to negotiate a settlement, succeeded in getting the imprisoned monks released. 66 This example does not typify a general concern with monasteries on the part of the state: the court had a particular interest in Baoen because that monastery had a history of im-

THE MING STATE AND BUDDHISM

perial patronage going back to huge grants from the Hongwu emperor and was still used as a site for state rites. The burning of a pagoda in almost any other monastery would have gone unremarked in court circles, but Baoen's destruction was an ill omen. The one instance of general state intervention in Buddhist affairs in the late Ming is the ban on mass Buddhist ordinations in I 573. With this ban we move from fiscal concerns back to the issue of the security of the state, for it was introduced as an attempt to stop sectarian organizing. 67 What could have been a major incursion into the autonomy of institutional Buddhism proved to be short-lived, however, for there is abundant evidence that public ordinations were being conducted again in the following decade. 68 This effort to restrict one form of public Buddhist activity was not sufficient to reimpose thorough state supervision of Buddhist affairs. Monasteries were by this time effectively independent of that supervision. There would be no further attempt to legislate Buddhism for the remaining seven decades of the Ming. Memories of the Hongwu restrictions, particularly the amalgamation order of I 39 I, were nonetheless still strong in the Ia tter part of the dynasty. The artist Wen Zhengming, writing in I548, could look around him in Suzhou and describe its palpable effects: In ancient times, Suzhou had many Buddhist monasteries, but since the reorganization of the Hongwu era, many have been abandoned. Of those still surviving in the city today, seven out of ten are abbeys and the rest are affiliated cloisters. There must have been a thousand chapels and halls that were amalgamated [and closed down]. The old grounds and abandoned sites were gradually taken over for private residences or made over into government offices. In some cases, the property remains intact but the name plaque has disappeared; in others, the name survives but the place itself no longer exists: in yet others, the site is overgrown and littered with trash and neither name nor building survives. 69

All was not decay, however, for Wen went on to use this gloomy portrait as a background to highlight the change he could see occurring in his own day, as derelict old monasteries were being restored and monks were gathering again for religious cultivation. The long decline was being so surely reversed that he even allowed himself to wonder aloud, "What will it be like for those who come after?" The answer, as we now know, was a widespread revival of institutional Buddhism throughout China at the hands of gentry patrons in his and the folllowing generations. However harshly the Hongwu regulations had affected Buddhism up until this time, the emergence of a new social base for monastic vitality in the late

Timothy Brook

I77

Ming would cancel out their influence. The support of the gentry rendered those regulations of no account . . The emperors of the late Ming followed suit, turning their backs on the regulations of the founder and becoming patrons of Buddhism. As of r 579, the Wanli emperor, at the behest of his mother, Cisheng, the empress dowager, was playing the role of patron in the style of the Six Dynasties, dispensing lavish gifts to dozens of favored monasteries and clergy.7° Imperial support extended to all aspects of patronage. For example, when a monk desiring to rebuild Baohua Monastery outside Nanjing submitted a memorial to the Wanli emperor in I 6o 5 seeking a new name plaque to indicate official blessing for his project, he not only received the plaque but was given a large sum of money from the Empress Dowager, as well as a copy of the imperially printed Tripitaka, a Buddhist statue, and a banner proclaiming the emperor's protection.7 1 Active as a patron of Buddhism, Wanli was not a legislator of Buddhist institutions. Imperial gifts were not acts of state, but the private dispensations of the imperial family. They signified the emperor's (or his mother's) personal piety, not state policy. The founder's laws were not revised. The formal relationship of the state to Buddhism in the late Ming-at least as expressed in the abandonment of quotas, the decay of certification, and the decline of registries-was otherwise one of relative indifference. Buddhist monasteries pursued their activities independently of the state and were left to their own devices so long as they paid their taxes and did not harbor vagrants or criminals. As far as the late Ming state was concerned, a monastery was simply another fiscal household (hu), distinct only in that it was constituted from elective rather than kinship affinities. Independence from the state was not the theme that late Ming patrons and apologists struck when they sought to define the relationship between Buddhism and the state, which their necessary commitments to Confucianism required them to do from time to time. They preferred to picture an interdependent relationship and continued to draw on the language of earlier edicts of patronage to do so. Thus Li Bangyan (jinshi I572) said of places like Guangdong's Nanhua Monastery that the teachings of the monks there "subtly support the life of the state. " 72 The noted Buddhist patron Lu Guangzu (152I-97) in a text on a Ningbo monastery could speak of "comforting the human desire to look up to heaven, fortifying the imperial plan, and settling the realm within the four seas. " 73 Li Zhi (I 52 7-1 6o2) could justify the greatness of Nanjing's Qixia Monastery by declaring it to be a place where "all pray that the present emperor will enjoy long life for tens of thousands of years. " 74 And by the time a patron

THE MING STATE AND BUDDHISM

composed a stele in honor of the restoration of Bochi Monastery in Huaian in 1677, it was possible for him to write of Buddhism: "When the Three Jewels [Buddha, Dharma, and Sangha] constantly flourish, the four quarters are forever at peace. Looking upward, [Buddhism] prays that the emperor's life may be eternally lengthened. Looking down, it showers fortune on the lives of the people. Great indeed are the advantages it provides the state! " 75 Such extravagant claims were hardly to be taken literally-except perhaps by individual officials looking to harmonize their service to the state with their enthusiasm for patronizing the sangha. They served a distinct purpose nonetheless. By tying the patronage of the monastery to a statist logic, an author could defuse the ever-lurking charge that to support a monastery was to put private concerns above public duties, to place religious needs as defined by the Buddha before the greater good as defined by Confucianism or the state. As always in the late imperial period, the state set the terms of legitimation; hence the Bochi patron's strategy of citing the state to justify patronage of an institution that did not obviously contribute anything tangible to public authority as the state might . have understood the concept.

Buddhism and the Redefinition of Public Authority The rhetoric of Buddhist underpinnings for the state conformed nicely to the norm of imperial patronage that had appealed to Hongwu in the opening years of his reign. Even though this rhetoric was being enunciated at a time when emperors like Wanli were happy to lavish attention on eminent monks and monasteries, the patrons invoking such phrases were not striving to work out a new accommodation between Buddhism and the state. Rather, they were seeking to justify their involvement, as a state-oriented elite, with an institution that was central neither to the state's agenda of rule nor to the reproduction of local society. Despite pronouncements to the contrary, Buddhism did not materially assist, or threaten, or otherwise matter to the power of the state. Nor did the state desire to mobilize the religion for ideological values or to intervene in its internal affairs. The failure of the Hongwu suppression to achieve a thorough or permanent refashioning of norms parallels the failure of earlier attempts by the Chinese state to appropriate the rituals of Daoism for its purposes. Kristofer Schipper has observed, for instance, that the state could not separate Daoism from its social background in local cults inasmuch as

Timothy Brook

179

"the necessary economical and organizational foundations were never present to make such an institutionalization into an enduring historical reality. " 76 The early Ming attempts to adopt, then subordinate, Buddhism failed for much the same reason. Hongwu provided no material underpinnings to the castle of supervision he constructed in the air around Nanjing. His limited attempts to protect the property basis on which monasteries relied were of little defense against the workings of the market, and his failure to incorporate Buddhist registries into the regular system of bureaucratic appointment hampered the creation of an organization that could be effective in realizing his goals. He lived long enough to regret his first policy of benign patronage, but not to see the futility of his second policy of suppression. Nor could he appreciate that monastic Buddhism, shunted to the margin of public authority, might there provide the gentry with means by which to pursue a quiet struggle for limited autonomy from the state-centered political realm. Having been banished to the margin of public authority, Buddhism became available as a site for alternative constructions of public authority in which the gentry could figure for themselves a prominent position. Marginality nurtured the monastery's social existence as a realm autonomous, in limited ways, from state institutions and processes. When that realm became integrated with local social structure, as it did in the widespread practice of gentry patronage throughout China in the late Ming, Buddhism provided an opportunity for local elites to create an identity for themselves in terms that were to some measure independent of the state. The Buddhist monastery was attractive to the gentry as an object of patronage by virtue of being a revered non-state forum in which the gentry could publicize their philanthropy, and hence their elite status, in local society. Unlike most other local institutions, the Buddhist monastery was hostage neither to the state's interests nor to the particular interests of corporate groups. It was not public in the sense of being a component in the structure of state authority; nor was it private in the sense of being closed to all but the members of a restricted collectivity. It existed between these arenas, constituting a limited but valuable space within which the local gentry could convert private wealth into public status and construct an identity at some distance from state mechanisms for conferring status. The patronage of Buddhist monasteries by the late Ming gentry was part of the process by which public authority was being reconstituted.77 When local gentry patronized the same monasteries that the empress dowager funded, it seemed that both adhered to the same norm of patronage, that the generosity of one was equivalent to the munificence of the

r8o

THE MING STATE AND BUDDHISM

other. But the social bases underneath this apparent congruence of norms were different. The Ming state did not appreciate that gentry patronage was a matter of undertaking the construction of autonomous space within local society. This lack of perception has something to do with the intellectual repertoire available for thinking about Buddhist institutions. Traditional discourse provided a language to complain about monks failing to obtain ordination certificates, or about laymen engaging in intercourse with nuns, or about bad characters "changing names [i.e., adopting religious names] to evade trouble and save their own necks," to quote the aging Hongwu. But there were no rhetorical resources to connect the gentry's pious giving, which looked as innocuously laudable as the emperor's activities, and their resistance to state dominance; nor between the vision of a finely restored monastery and the construction of autonomous space within the realm. The gentry were thus free to use Buddhist monasteries to deflect state control and, in so doing, to redefine public authority as an interactive field of rule and compliance within which they played a pivotal rather than a passively supportive role. The gentry's absorption in patronal relationships with Buddhist monasteries and the new norm of tolerance toward Buddhism that went with it prompted a backlash among conservative elements of the late Ming gentry. Guardians of the Hongwu dispensation liked to observe that the Ming dynastic founder had arranged Buddhist institutional life so perfectly that the monks could do nothing but "bring improvement to the people without corrupting our Confucian authority. Who but the greatest sage in the empire could have achieved this?" 78 They doubted that the involvement of their fellow gentry in the affairs of Buddhist monasteries could benefit the realm and warned them against failing to devote themselves to their Confucian duties of providing charity and promoting education.79 This argument against the Buddhicization of the gentry was a transposition of the more essential theme of the rejection of any aspiration to autonomy among the gentry, so strongly voiced in the early Qing by advocates of a reconstructed Confucian order. It did not, however, extend to a restoration of the supervisory mechanisms by the new Qing state. The Qing was content to repeat the paper regulations for monks and monasteries laid down in the Ming and take no further action. It did not revive the registry system, or impose quotas on monks, or limit monastic property. Considering the internal organizational weakness of Buddhism that the Ming zealously fostered, the Qing did not see a need to police the clergy as closely as Hongwu did. In a sense, then, the Hongwu suppression

Timothy Brook

I8I

worked, for Manchu emperors could afford to adopt the pose of patron without worrying about what the Buddhists were up to. But in the sense that the suppression failed and, in failing, helped the gentry to redefine public authority, the Qing could have had more to worry about. Qing success in binding the gentry to its statecraft initiatives, however, enabled it to recenter public authority on the state and so effectively block the possibility that monasteries could serve this function. The regime did not recognize what was afoot in local society in these terms, but it did understand that the late Ming gentry, caught up with "mad Chan" and factional literary coteries, had failed its dynasty, and that for this reason the Ming house had fallen. It was the gentry, rather than Buddhism, that bore the brunt of the condemnation. As for Buddhism, the Qing regime assumed that it remained safely and innocuously marginal to all forms of social power. And indeed, with the reassertion of state-centered public authority, it did: its contribution to social change declined as the possibility of limited local autonomy faded. The Qing denouement nonetheless does not diminish the importance of what Buddhism enabled in the Ming, or of what it reveals of the possibilities of state-religion (and state-society) relations. The dynastic sequence of state patronage, suppression, tolerance, and local patronage may not have led to a permanent refashioning of public authority, but it does show the variability and instability that characterized society's relations with the late imperial state.

CHAPTER

7

Power, Gender, and Pluralism in the Cult of the Goddess of Taishan Kenneth Pomeranz

OI"fom the mid-Ming to the early twentiethcentury, Bixia Yuanjun (the Goddess of Taishan) was among the most widely worshiped deities in China. Accounts of her cult regularly described her as the northern Chinese equivalent of Tianhou; some claimed that she, Tianhou, and Guanyin were all the same goddess. 1 The folklorist Gu Jiegang described her as the most important female deity in North China. 2 Her principal temple on Taishan attracted at least 40o,ooo visitors per year during the heyday of the cult and probably many more. Local temples were spread throughout North China and Manchuria and were apparently matched in number only by those of Guanyin, Guandi, and local tutelary deities (tudi). 3 Yet despite this popularity-and the absence of any ideas or political ambitions comparable to those of some "heterodox" sects-Bixia Yuanjun's cult never won full acceptance from the state and eventually came to be considered dangerous by much of the elite. While the only other discussion in English of this cult and its relationship to orthodoxy suggests that her incomplete co-optation was largely a result of the cult's relatively late emergence, 4 I believe that elite concerns about the goddess's dangerousness-concerns that became increasingly prominent during the Qing dynasty-made the historical trajectory of her cult not just incomplete but fundamentally different from those of cults such as Guanyin's and Tianhou's. The perceived danger of Bixia Yuanjun lay partly in familiar Confucian concerns about the excessive fervor and dubious aims of popular ritual, but these matters alone need not have been decisive obstacles to elite acceptance: many other cults had enthusiastic, participatory rituals that coexisted successfully with the more austere and controlled official and

J'

Kenneth Pomeranz

elite worship that accompanied canonical status. But in Bixia Yuanjun's case the guardians of orthodoxy were also bothered by the ambiguous message that she sent about female roles and female sexuality: a message quite different from that associated with universally accepted female deities such as Tianhou and Guanyin, and one that made her cult an important battleground in broader Qing conflicts about proper female roles. Moreover, extreme kinds of fervor (including ritual suicides) and unacceptable kinds of miracles seem to have been linked to improper female roles in the minds of the cult's critics-and, more speculatively, in the rituals, lore, and core tenets of the cult as well. Furthermore, despite this lively history, the name of Bixia Yuanjun usually evokes blank looks today, both among sinologists and Chinese from south of the Yangzi. 5 And while she does still figure in the culture of North China, it is often as a heroine of supernatural adventures rather than as an object of piety; her role seems more like those of Najia or the Monkey King than those of Guanyin or Wusheng Laomu. (These two are not mutually exclusive, of course, and her worship is very much alive as well.) Consequently, this cult provides an important contrast to the wellstudied cases in which the religious efforts of the state, local elites, and "ordinary people"-though often motivated by disparate ideas-combined to reproduce a shared set of symbols and practices. 6 Here, the fabric of what was once a growing, officially approved cult gradually unraveled into separate, often hostile, strands of state, literati, and popular activity. This is not the place to describe this unraveling in detail, but it is important to bear it in mind. Indeed, my purpose in reconstructing this cult's practices and symbols, and some of the ideas that appear to be embedded in them-particularly with regard to gender roles and to what sorts of human activities could make heaven respond-goes beyond recovering lost evidence about "popular" beliefs. I also hope to explain what aspects of this cult in particular made so many Qing literati abandon it and even single it out for public disparagement, despite the favor it had enjoyed in the Ming. This parting of the ways between literati and popular practices makes our current act of "recovery" both more necessary and far more difficult than it would be had Bixia Yuanjun retained her literati support and become a fully accepted member of the official pantheon. As we shall see, Bixia Yuanjun's rise to prominence involved support from virtually every group in Song, Yuan, and Ming society, but for various reasons, the Ministry of Rites never made her an approved object of official sacrifice. Pending such approval, she remained in an important intermediate

THE CULT OF THE GODDESS OF TAISHAN

category, which has received surprisingly little attention from modern scholars. Beginning in the late Ming, the worship of Bixia Yuanjun was generally labeled yin (licentious; also supplementary, unnecessary, surplus). This label was not nearly as damning as xie (heterodox); indeed, the existence of this third category provided an important breathing space, allowing some cults that did not measure up to the standards of orthodox (zheng) worship to escape the strict zhenglxie duality that might otherwise have marked them for suppression. Gentry could and did defend the enthusiasm of their less elite neighbors for yin cults without raising serious doubts about their own basic loyalty to the state and to Confucianism; they certainly would have raised questions about themselves if they had defended any aspect of a cult that had been publicly labeled xie. Some defenders of yin cults simply denied that the yin label properly applied to the cult in question; others conceded that the label had some validity, but only in reference to some largely harmless popular errors (usually born of excessive fervor) that were outweighed by the cult's basic harmony with orthodox morality. And during crises, even the conspicuously orthodox might publicly avail themselves of such worship/ since it was by no means a repudiation of orthodox rites: instead, yin cults appear to have often been seen as ways of calling on additional powers to supplement those invoked in ancestral worship and other "proper" rites. Ultimately, however, there was no way to paper over all the divergencies between the social and cosmological ideas implicit in zheng and yin worship. In the Qing, literati became increasingly unanimous and assertive in their written condemnations of the worship of Bixia Yuanjun and of many other yin deities-almost all of them female. 8 (Private literati attitudes and practices no doubt remained more diverse, but only the occasional nonconformist literatus-such as Yuan Mei or Chen Wenshu-continued to publicly treat the Bixia Yuanjun cult with respect.) Although most of the literati complaints about Bixia Yuanjun in the Qing and Republic pointed to the weakness of textual evidence that she had been known in antiquity, these fruits of Qing "evidential research" by themselves explain little. Certainly the textual bases for Guanyin's cult were little better, from a Confucian perspective, yet she encountered no comparable criticism. What mattered more than failing the philological tests applied by Qing literati were the various ways in which Bixia Yuanjun's cult fit poorly with the increasingly stringent standards and ideas applied to female behavior in the early and high Qing, especially among the lower gentry and better-off commoners who made up the social elite of rural North China.

Kenneth Pomeranz

r8s

The most common gender-related complaints concerned the pilgrimages to Bixia Yuanjun's shrine at Taishan; such trips, which were usually predominantly female, 9 raised the specter of immoral behavior by groups of unsupervised women. Complaints about pilgrimages were often linked to a more general elite hostility toward the leadership role that midwives, matchmakers, healers, and other such women played in the cult. These women were unusually independent, regularly breached the boundaries of family and female privacy, and thus were often seen as social irritants; moreover, their potential influence on more "settled" and enclosed women made them rivals of a sort to local notables seeking to treat (and display) "their" women as semi-cloistered models of virtue. Moreover, while the complaints about boundary-crossing women were no doubt meant literally-as opposition to the influence of these particular groups-they also stood for a still more basic sociological and ideological problem: for on another level, the very essence of the Bixia Yuanjun cult was the invocation of powers that were gendered as female, and that, like midwives and matchmakers, were imperfectly controlled but annoyingly indispensable. No matter how orthodox the ends for which Bixia Yuanjun was usually invoked, the powers she represented were in many ways dangerous and made elites nervous in a way that the worship of Tianhou and Guanyin never did. Furthermore, I will argue, Bixia Yuanjun's powers were associated with a vital but dangerous kind of power-the sexual and reproductive capacities of young wives/daughters-in-law. These were powers to which every family and lineage had to resort, but they were embodied in people consistently viewed as inadequately assimilated and potentially destabilizing foreign bodies in the family. By contrast, Tianhou and Guanyin seem to have been associated more often with the figure of the mother-in-law and with senior women in general; such women were fully absorbed into the lineage, no longer disposed of dangerous reproductive powers, and, as shared objects of their sons' affections, were generally regarded as foci for family unity. Space does not permit much discussion of the Qing elite's rejection of Bixia Yuanjun. However, I hope to convince the reader that the content of this cult differed from that of the Tianhou and Guanyin cults, which Confucian elites found it easier to approve of, and that this unusual contentwhich had earned the cult the yin label in the first place-resonated with anxieties about gender and about gendered sources of power in ways that made the cult of Bixia Yuanjun particularly vulnerable to the heightened moral and historical scrutiny that literati applied to "popular" religion in the high Qing. And finally, while the twentieth century lies still further

186

THE CULT OF THE GODDESS OF TAISHAN

beyond the scope of this chapter, I argue elsewhere that popular awareness of this elite hostility eventually created a very different Bixia Yuanjun: one who appears in contemporary tales as a trickster tweaking the noses of the powerful and is less committed to upholding order than she was in the days when she had a more substantial following among the powerful. Thus, this longer-range picture of Bixia Yuanjun's fortunes fits poorly with the "orthopraxy" model of Chinese society: the scholarly view that shared symbols and ritual practice preserve a superficial unity across classes, genders, and other groups that privately may hold different ideas about what their practices mean. Here we find instead serious conflict at the ·level of ritual and symbol; conflicts that were made meaningful by certain shared ideas, both about cosmology (e.g., the importance of Taishan in relations between heaven, earth, and humans) and social life (e.g., the nature of a daughter-in-law's duties, and the potential for young daughtersin-law to be weak links in the desired solidarity and harmony of her new family). The treatment of Bixia Yuanjun by elites in the Ming had borne a somewhat closer relationship to orthopraxy; however, this accommodation had been fatally undermined by the mid-Qing. First, the literati themselves increasingly withdrew from public participation in this cult, as they did with many other cults that failed to meet their increasingly stringent standards of orthodoxy. Second, a more casual Ming attitude toward yin practices and ideas on the part of the masses was increasingly replaced by a sense that doctrinal "errors" in popular worship could have serious consequences and could not be safely ignored by elites. The yin space was never completely squeezed out by the zhenglxie duality, but it narrowed a great deal. As a result, this case-in which co-optation never occurred-throws into sharp relief the importance of purging unacceptable aspects of a deity as part of canonization processes that were completed: a part of the process that has often been slighted by accounts that emphasize symbiosis and the more pluralistic implications of orthopraxy. 10 In this chapter, I concentrate on showing how the cult blended orthodox Neo-Confucian goals with a much more eclectic vision of what people and forces were important to the pursuit of those goals-a mix that gave the cult enormous appeal, but also provoked considerable elite hostility in the Qing and the twentieth century.

The Early Rise of Bixia Yuanjun Taishan, the "Peak of the East," has figured in official ritual since ancient times. It was, however, most important from the Qin unification through

Kenneth Pomeranz

the Northern Song, when emperors periodically renewed the Mandate of Heaven by performing the (eng and shan ceremonies there. 11 The divine entity they addressed was simultaneously the concentration of vital energy represented by Taishan itself and a still more abstract heaven; some texts explicitly equate the two, and a relic of their conflation may be found in the hundreds of temples to Taishan-and later Bixia Yuanjun-called "Tianji miao" (Temple of Ultimate Heaven). Non-official practices and beliefs about Taishan before the Song remain obscure, but a few things are clear. First, an anthropomorphized male god of Taishan developed during the early imperial period, long before he was officially recognized. 12 Eventually this unofficial deity acquired a personality, a family (including a wife and five children), and a distinct and crucial jurisdiction: the underworld. It was as Yan Wang (sometimes known as Yanluo Wang), king of the underworld, that this god of Taishan had his greatest impact. He and his staff of infernal yamen runners determined the length of human lives and then judged and examined the deceased, assigning them to one of a variety of new existences. The widespread belief that Yan Wang had his headquarters underneath Taishan fit well with yin/yang and Five Phase theory, which located the wellsprings of life (and thus the determination of its length) at the meeting point of heaven and earth (qian and kun, male and female): the top of the Eastern Peak. However, relatively few people went to Taishan to address Yan Wang; while many new rituals in which commoners sought to address the underworld directly arose during the tenth through thirteenth centuries, pilgrimage does not seem to have been prominent among them. 13 Meanwhile, a female cult figure, the Jade Maiden (Yunii), was somehow associated with Taishan. Her worship dates back at least to the Qin, but we know little more; by the time our story begins, she had become obscure, although various "jade maidens" figured in legend and ritual all over China. 14 Bixia Yuanjun appeared on the stage in roo8. The Song emperor Zhenzong (r. 998-1022), in trouble both at court and on the northern frontier, announced a number of miracles; the most important was the sudden appearance of a tablet from Heaven (tianshu). To give thanks and renew his mandate, he journeyed to Taishan to perform the (eng and shan rites at the summit of the mountain. As the climax of a series of miracles there, he discovered what he proclaimed to be a statue of the Jade Maiden, whom he claimed had been worshiped at the mountaintop in antiquity. 15 Although the accounts of this discovery were vague about Yunii's relationship to the mountain and its existing deity, the statue was treated as a

188

THE CULT OF THE GODDESS OF TAISHAN

find of great significance. Before long a variety of stories were current. A few made Yunii the wife of Taishan; most made her his daughter or granddaughter. According to others, she had come from Mount Kunlun at the behest of Huangdi (the Yellow Emperor) himself. 16 She received a succession of imperial honors throughout the rest of the Song, Jin, Yuan, and Ming; in the Jiajing reign (Ip2-66), she was given the title Tianxian Yunii Bixia Yuanjun (Heavenly Immortal Jade Maiden and Princess of the Azure Clouds). When the semi-official Daoist anthology, the Daozang, was expanded in I 607, a Bixia Yuanjun sutra was added; the goddess was mentioned in other juan as well. She was described as controlling several crucial functions, including setting human life spans, the granting of children, and the judging of the dead. The old god of Taishan was also mentioned but seems to have been deprived of most of his functions. 17 In the Ming, even literati who refused to commit themselves about whether Bixia Yuanjun really existed as a personal deity routinely noted that the top of Taishan represented the utmost extreme of kun (earth/female principle); they thus gave the spot's power and spirit Bixia Yuanjun's gender, rather than Yan Wang'sY And by this time Bixia Yuanjun was unquestionably the deity of Taishan in the popular mind. Massive pilgrimages seem to have begun in the early Ming; by the mid-14oos, her temple complex (on the site of Song Zhenzong's supposed discovery) had become enormous, while the temple of the old God of Taishan, slightly lower on the mountain, suffered both popular and official neglect. 19 By the late Ming, the worship of Bixia Yuanjun, both on Taishan and elsewhere, had taken the basic form that would last into the twentieth century. Social Organization of the Cult The temple complex at the top of Taishan was extremely rich and attracted huge numbers of pilgrims. Eight officials shared the task of collecting an "incense tax" on temple visitors. The rate for pilgrims from Shandong was 5·4 fen per head through most of the I soos; people from other provinces were supposed to pay more but seem to have avoided this. The tax regularly yielded over 2o,ooo liang per annum for the state in the rsoos, plus reputedly a large income for the collectors. Thus, at least 40o,ooo people must have paid the incense tax during the eight months each year when it was collected. Various contemporaries estimated "several hundred thousand," "eight hundred thousand," or "a million" pilgrims per year; one writer estimated that pilgrimages declined precipitously to 40o,ooo during the chaos of the r63os. 20

Kenneth Pomeranz

These pilgrims, who came from all classes, also gave to the temple itself. Court women and eunuchs often gave lavish gifts; the Wanli emperor's mother was particularly generousY However, the court did not dominate temple patronage here as it sometimes did in Beijing/2 and the huge number of smaller but still expensive gifts noted in every description of the temple-including silk, silver jewelry, jade, silver and gold figures of children (given in thanks for a birth)-make it clear that many other prosperous people were visiting the temple. The vast majority of pilgrims were, however, peasants. 23 Most pilgrims arrived either during a short period in the first lunar month or during the two weeks surrounding the goddess's "birthday" in the fourth lunar month. At these times, large numbers also came to Bixia Yuanjun temples atop Huashan, Foqiushan (in Junxian, Henan), and various lesser regional centers; cities along major routes held festivals to welcome pilgrims on their way to Taishan. 24 Virtually every county in the area had at least one "send off" festival and held sacrifices and other events for the much larger numbers who worshiped Bixia Yuanjun at local sites.ZS There were huge numbers of local temples-some in almost every county, and 39 in Heze (Shandong) alone, according to a late Qing gazetteer. 26 One county gazetteer recorded 4 7 days per year of Bixia Yuanjun festivals at various local temples/ 7 and many gazetteers noted that the volume of the commerce at these fairs significantly affected the county's overall prosperity for the year. 28 With one important exceptionthe absence of people from the court in Beijing-local festivities seem to have been organized by the same people who participated in the journeys to Taishan. Wealthy pilgrims often came as individuals or, if they were women, with a close male relative or trusted servant. 29 Most pilgrims, however, came in organized groups. Once they reached Taian, the prefectural city at the foot of Taishan, people were herded into what were essentially package tours run by innkeepers, who took care of food, lodging, and the incense tax, and supplied a guide who would take the pilgrims to key spots. 30 It is more important for our purposes, however, to see who composed these groups and who provided organization at earlier points. Pilgrimage groups were organized at the village level, or sometimes lane by lane. The large majority of pilgrims were said to be women, and many groups were all female, but often husbands and wives went together.31 Many groups were organized by women whose work allowed them to visit others: midwives, healers, wet nurses, and matchmakers. The late seventeenth-century magistrate Huang Liuhong was so upset by

THE CULT OF THE GODDESS OF TAISHAN

their influence, particularly by their promotion of pilgrimages, that he favored forbidding such women to enter any house whose residents were not immediately in need of their services. 32 The other organizers of pilgrimages and of local temple building are harder to identify. Many degree holders wrote stelae commemorating the construction or repair of a Bixia Yuanjun temple, but these literati almost never say that they participated in the task and rarely named major donors; this is in sharp contrast to stelae commemorating temple repairs for city gods, Tianhou, Guandi, and other less controversial figures. Instead, these stelae invariably emphasize that the construction was made possible by mass enthusiasm and large amounts of freely donated labor and supplies. Indeed, many cite this mass enthusiasm as the best indicator of the goddess's ling. 33 Project leaders are usually characterized as shou shi (probably village elders), junzi (gentlemen), or xiang laoye (rural elders): members of the informal rural elite but not the peers of even the lower degree holders in the county capital. 34 One late Ming stele, written by a jinshi in Yangxin county, Shandong, remarks specifically that most of the county's literati took no part in this project: the author says that they preferred the "old" division of religious labor, in which they attended to sacrifices at the county's major altars (together with the county magistrate) but did not regard popular temples as their affair.JS Degree holders and their relatives were not absent from these activities/ 6 but were far less central than in the patronage of many other temples. Their participation also declined with time: although some upper gentry appear in Bixia Yuanjun temple inscriptions during the Ming and the Shunzhi (I 644-6 I) period of the Qing, they disappear thereafter. The Yangxin stele mentions one other group that raised funds and workers for Bixia Yuanjun temples: yamen runners. 37 Their involvement is particularly significant, for while they were often quite locally influential, they were nonetheless conventionally despised. Certainly they were never thought of as appropriate moral exemplars or as people who should be encouraged to build a following beyond what their de facto power already gave them. Thus, their patronage of local temples could not be justified by the usual Confucian logic of leadership through ritual and example; indeed, it probably reinforced the aloofness of many gentry from this project. However, yamen runner participation in the Bixia Yuanjun cult makes sense from many other perspectives. Like some other female deities who stood outside the regular celestial bureaucracy, Bixia Yuanjun may have been particularly attractive to many who were considered too impure to

Kenneth Pomeranz

play a role in more orthodox cults. More specifically, as the supervisor, in both popular tales and the Daozang, of the underworld yamen, Bixia Yuanjun was the boss of the yamen runners' underworld counterparts; so, according to other traditions, was the City God (celestial counterpart to the county magistrate), but his temple would have already been colonized by other, more prestigious social groups. And from the historian's perspective, the yamen runners' involvement would help explain the large mobilizations of "ordinary" people and resources without extensive gentry or local government efforts. 38 Thus the local cults' social organization may have expressed the real structure of informal influence in poor rural areas of North China, where neither the state nor the Confucian gentry was very strong: village elders, yamen runners, healers, matchmakers, and so on. With such conventionally despised characters as matchmakers and yamen runners involved, it is hard to see this cult's local leadership as the best approximation of "proper" gentry leadership that poor villages could summon, as accounts of other village temple affairs (led exclusively by village elders) sometimes suggest;39 it bore signs of deviance, not just imperfect imitation. The local structure of the Bixia Yuanjun cult would thus have interesting, if inexact, parallels to that of important national shrines like Taishan itself, or Beijing area temples where Bixia Yuanjun was prominent. 40 At both levels, one finds influential women, often acting behind the scenes; a large base of popular enthusiasm; a conventionally despised but practically powerful group of males (eunuchs, runners) that lacks access to important mainstream channels for solidifying its social status {posterity for eunuchs, the exam system and most other honors for runners); and an ambivalent set of Confucians (the Ministry of Rites in Beijing, the local literati in the countryside), many of whom keep their distance, at least publicly. Although the actual balance and interplay of these rival groups differed at the local and national levels, the crude structural similarity is still significant. At both levels the cult was often said to be predominantly female. 41 Moreover, at both levels it was socially associated with functionally vital groups that appeared to orthodox Confucians as at best necessary evils: groups that, furthermore, were not firmly contained within the institutions (such as the family or the state) in which real moral action should occur. Many more orthodox notables at both levels provided occasional high-profile patronage in their capacities as private persons but generally refrained from sustained and official participation. In the rest of this chapter, we shall see that Bixia Yuanjun was also associated with questionable influences on a more intimate level and that

THE CULT OF THE GODDESS OF TAISHAN

these associations sprang from the content of her cults' legends and practices as well as its membership. While we know little about what kinds of women were most involved in the cult, the Bixia Yuanjun of ritual and legend had a special affinity for sexually attractive young women, particularly for young daughters-in-law, the group that was perhaps the example par excellence of a necessary but suspect addition to a vital social unit. Moreover, as we shall see, beliefs about how Bixia Yuanjun aided her devotees often connected her with powers that could be highly destructive and were so far removed from the orderly functioning of Heaven and Earth that many orthodox neo-Confucians preferred to deny their very existence. The same lore that raised these anxieties tended to blunt the cult's conflict with nco-Confucian morality by placing these powers in the service of eminently orthodox goals and providing plot twists that minimized their destructive impact. But merely placing an orthodox frame around more volatile powers, ideas, and social groups was not enough to create-either for better or worse-a thoroughly "Confucianized" 42 cult; indeed these powers were so central to the cult that any cleansing of its disquieting elements sufficient to make Bixia Yuanjun safe for the pantheon (at least by the increasingly strict standards of the Qing) would probably have destroyed or radically altered the cult's basic appeal. To see why, we now turn from the cult's social makeup to its symbolic and ideological content. Ritual, Legend, and Meaning in the Bixia Yuanjun Cult Almost all accounts of Bixia Yuanjun's pilgrims emphasize mass enthusiasm. People were especially eager to be present at the annual "opening" of her Taishan temple at dawn on her birthday. The night before, huge crowds assembled outside the temple, burning incense and money, staying up all night (in what was often extreme cold), and chanting and drumming continuously to "wake" the goddess. The same scene was repeated every night for months each year, although on a slightly smaller scale and at other locations. 43 Even hostile witnesses marveled at the crowds' devotion, often wondering what wonders such fervor could accomplish if directed at a worthier object. 44 During the day, pilgrims filed through the temple complex, stopping to prostrate themselves, pray, and make donations. The last stop was usually a rear building that was the goddess's bedchamber; although the statue in this room, lying on a raised bed behind screens and an iron grating, was difficult to see, people would toss offerings through the grating. The floor in this room was piled thick with

Kenneth Pomeranz

193

offerings (mostly copper cash); twice a year, a team of officials and the temple's resident Daoists would sweep the floor and divide the spoils. What did pilgrims expect this deity to do for them? At the peak of her cult in the late Ming, Bixia Yuanjun was an all-purpose goddess, from whom people sought good harvests, profits in trade, and general safety; but above all she brought children and good health for parents. 45 Her principal attendants were in charge of giving children (Songzi Niangniang) and protecting the vision of the elderly and the very young (Yanguang Niangniang); six other attendants handled every part of reproduction from conception to breastfeeding. In general, Bixia Yuanjun took over Yan Wang's function of deciding.life spans, creating new life, and judging the dead; various sources emphasize that she brought to these functions a far more merciful, life-affirming orientation than Yan Wang did. 46 The Daozang also attributed to her a related talent: distinguishing infallibly between sincere and insincere worshipers. 47 To this extent, Bixia Yuanjun resembled various other female, nonbureaucratic deities. They, too, listened to a wide variety of requests but were often particularly concerned with motherhood and caregiving; they, too, were supposed to be particularly merciful and said to respond directly to the sincere prayers of common people, especially women. And as noted above, many commentators claimed either that Bixia Yuanjun was an avatar of these other goddesses or that they were avatars of her. 48 (Indeed, Henri Maspero suggested that the most popular form of Guanyin-Guanyin clothed in white-is in fact historically derived from Bixia Yuanjun; he even referred to her as Bixia Yuanjun's "disciple.") 49 However, the special efficacy of female prayers, which is not generally agreed to be a feature of other goddesses-some contemporary worshipers of Tianhou, for instance, believe this, while others do not50-may have been particularly exaggerated in the case of Bixia Yuanjun. In the stories about her that I have found thus far, every individual petitioner to whom Bixia Yuanjun responds is female; the only instance of her appearing spontaneously to help males is in a post-1949 story in which she rescued a group of Shandong soldiers surrounded by Americans during the Korean War. 51 By contrast, although many worshipers of Guanyin and Tianhou seem to believe they were particularly helpful to women, the most famous miracle attributed to each of them, at least as recorded in written texts, was the rescue of a male kinsman. 52 And in some other ways, Bixia Yuanjun not only exaggerated certain "female" characteristics of less controversial female deities such as Tianhou and Guanyin and even the "heterodox" Wusheng Laomu, but also differed radically from them.

194

THE CULT OF THE GODDESS OF TAISHAN

For one thing, Bixia Yuanjun's statues were treated in an unusually anthropomorphic way: for instance, at certain times either the statue that showed her lying down behind screens in her bedchamber or the one that had her sitting up (in the front room of the temple) was removed for a "rest." 53 More important, this particularly "human" deity resembled human females (or at least social constructions of them) in unique and threatening ways. As Steven Sangren has pointed out, Tianhou, Guanyin, and Wusheng Laomu are all largely desexualized; each is, in her own way, a virgin mother, who reconciles, as mortal women cannot, the demands of Confucian filiality-which include producing an heir-and the demands of religious purity. Put in social terms, they play the role of mothers, whom the society generally sees as the family's source of unity, harmony, and security, without ever playing the role of wife or daughter-in-law (not to mention lover), whom the society generally casts as sources of discord and danger. 54 Bixia Yuanjun was also often described in maternal terms-as "the Sacred Mother," "the Old Woman," or "the Grandmother of Taishan." But while being a pure and all-merciful mother to everyone seems to capture the essence of Guanyin, Tianhou, and perhaps even Wusheng Laomu, it represents just one side of Bixia Yuanjun's personalityand probably not the dominant side. In most of her lore, Bixia Yuanjun, like these other deities, seems to have no husband, but at least two tales about her do give her a spouse (a different one in each case). 55 And unlike these other deities, Bixia Yuanjun's lore retained elements that suggested an affinity for the divisive daughterin-law role. Moreover, various aspects of her cult emphasized her physical attractiveness. In many stories, Bixia Yuanjun was the daughter of Taishan, yet she also "lived" at Taishan: that is, at her natal place. If she was assumed to be unmarried, this merely made her situation parallel to those of Tianhou and Guanyin; but whereas they both "earned" the rightto remain part of their natal lineage by rescuing males of that lineage,56 there are no comparable stories about Bixia Yuanjun. (According to one story she was the daughter her parents had wanted after five sons. 57 However, nothing is said about why such a family would retain a daughter-normally only a family with no son would do so. Moreover, Bixia Yuanjun is not credited with doing anything special for either her father or brothers, as Tianhou and Guanyin did.) If Bixia Yuanjun was married, on the other hand, other complications arose. A married woman had no business living in her natal place or choosing where she lived-yet what appear to be the most widely circu-

Kenneth Pomeranz

195

lated stories about Bixia Yuanjun, at least in the twentieth century, tell how she chose to live atop Taishan and won the right to do so. 5 8 And the oldest literary reference to a daughter ofTaishan-a text sometimes erroneously used to "prove" that Bixia Yuanjun had been known in classical times-is about a married daughter of Taishan returning home, for reasons that are not specified. 59 Bixia Yuanjun also had an association with real-life married women away from their husbands' homes. At least in early twentieth-century Shandong, one of Bixia Yuanjun's functions was to protect young wives visiting their parents. 60 These trips were culturally sanctioned, but they nonetheless emphasized the daughter-in-law's less than total integration into her husband's family and kept alive suspicions of dual loyalties; ideally, these trips were supposed to cease after a few years. Although Tianhou and Guanyin, who miraculously contributed to the survival and reproduction of their natal families, avoided any such conflict, Bixia Yuanjun seems to have been associated with it. A still more dangerous feature of daughters-in-law was their sexuality; they could seduce their husband away from his primary loyalty to his parents or disgrace the family as a whole by attracting another man. Here, too, Bixia Yuanjun was markedly different from the goddesses who were co-opted and canonized. Iconographically, Bixia Yuanjun was consistently pictured as a young and particularly beautiful woman; foreigners who got a rare chance to see her principal statue on Taishan without the grating that was usually in front of it were particularly struck by this and commented how much more attractive she was than other female icons. 61 Moreover, her main statue and many other pre-1949 images show her with very small bound feet; 62 Tianhou and Guanyin, by contrast, appear to have natural feet, when their feet are visible at all. In tales about her that circulated in the twentieth century, her embroidered slippers are her special trademarkthe item she leaves behind as proof of her presence. 63 The emphasis on this particular item would seem to call particular attention to her feet, her femininity, and perhaps also to a certain frivolity: in the most common version of the story of how she claimed Taishan, Bixia Yuanjun's embroidered slipper is paired against a rival deity's trademark muyuthe "wooden fish" emblematic of a traveling priest. 64 Other unusual features of the cult gave Bixia Yuanjun a similar cast. The most common offerings (aside from real and "spirit" money) were jewelry, shoes, and silk; at least at Taishan she was rarely given the food offerings that were more common for most deities. 65 ( Guanyin, for instance,

THE CULT OF THE GODDESS OF TAISHAN

received mostly food offerings and rather austere ones at that.) 66 According to one apocryphal story that reached the ears of a nineteenth-century missionary, the state's share of the money collected at her temple had originally been earmarked for the empress's makeupY The Bixia Yuanjun in the Beijing Dongyuemiao was also the deity who was supposed to make girls and young women beautiful. 68 All these details served to underline in Bixia Yuanjun elements of luxury, beauty, display, and sexuality that we find nowhere in Guanyin, Tianhou, or even Wusheng Laomu. Moreover, unofficial lore gave Bixia Yuanjun an additional function, absent from the texts of official Daoism, which placed her atop a reservoir of sexual danger. In various stories collected by the eighteenth-century poet Yuan Mei, Bixia Yuanjun is identified as the chief of all fox-spirits; 69 and, as countless popular stories describe, fox-spirits regularly took the form of human females, seducing young men, temporarily satisfying all their wishes but eventually leaving them worse off than ever. Whereas the Daozang, for instance, emphasized Bixia Yuanjun's role as a pillar of order who rewarded filiality and other virtues, the popular lore about fox-spirits gave her authority over a dangerous pool of sexual energy, which frequently subverted these virtues. True, the fox-spirits were supposed to be purifying themselves under her supervision-one story describes her holding a special set of examinations for them, by which they can become true immortals.7° Thus, one might even see Bixia Yuanjun's role as supervisor of the fox-spirits as a way in which she, like her human devotees, seeks to direct yin sources of power toward acceptable ends. Nonetheless, fox-spirits were famous for wreaking havoc with their magical and sexual power; being their leader hardly fits neatly into the image of a fully orthodox deity. Another story casts a fox-spirit in a role similar to that of Bixia Yuanjun herself. After an infatuated Daoist gives this fox-spirit the right to inhabit and "eat the incense" at a hilltop temple previously occupied by Guanyin (with whom Bixia Yuanjun often shared temples, and with whom she was sometimes conflated), she becomes known as a particularly efficacious but dangerous deity, whose statue is so beautiful that it must be hidden behind heavy screens (as one of Bixia Yuanjun's was) lest it cause male pilgrims to think impure thoughts. 71 The seventeenth-century novel Xingshi yinyuan zhuan (A marriage to awaken the world) refers matter-of-factly to Bixia Yuanjun as the patroness of fox-spirits, as if this is a fact the reader was bound to know. In fact, she figures prominently in the revenge exacted upon that novel's hapless "hero" after he kills a fox-spirit: a revenge that works itself out through a complete inversion of the conventional female subordination that the author

Kenneth Pomeranz

197

clearly preferred. And the symbolic climax to the domination of women over men in the novel occurs during a pilgrimage to Bixia Yuanjun's shrine on Taishan, in which only one male is present: the henpecked and repeatedly humiliated reincarnation of the man who killed the fox-spirit. 72 Thus at least in this "literati novel," 73 Bixia Yuanjun's power is associated with unacceptable gender roles and a moral decline that the author blamed for the collapse of the Ming; while she acts to avenge a wrong and so, in a sense, restore order, the female powers that she supervises can only do so by exacerbating chaos, bringing on a complete ruin that may ultimately serve as a chance for renewal. To Qing literati looking back at the Ming, it was clear that morality needed no friends of this type. Unfortunately, we have no direct statements from Bixia Yuanjun's devotees; thus we can only speculate that some of her female worshipers were drawn by these unusual parts of her "personality" and the recognition they gave to sources of female power besides maternity. Certainly the favors most often sought from her are perfectly filial and consistent with established female roles: help in having and rearing children, and longevity for aged parents. But even here some aspects of the cult made the guardians of orthodoxy uncomfortable. In most temples and pictures, Bixia Yuanjun is accompanied by Songzi Niangniang, "the woman who sends children"; quite often, as on Taishan, she is accompanied by a larger retinue of women, who oversee different aspects of pregnancy, childbirth, and child rearing.74 Many of her temples also featured other heroes associated with fertility.7 5 She seems not to have helped newlyweds obtain children but to have been the goddess who was asked for help once enough time had elapsed to raise worries about infertility, or in cases where the would-be parents had special reasons for anxiety. Although local practices varied, women throughout North China who sought special help in childbearing would go to a local Bixia Yuanjun temple and select one of several dolls that sat in front of a side altar in the temple.76 The woman would take one of the dolls home, performing various rituals as she walked through the village; she was then expected to return the doll and/or contribute a new doll (plus, in some cases, a cash contribution) after giving birth. 77 These rituals are still performed at Bixia Yuanjun temples in North China today. 78 The borrowing of the dolls (some of which were quite valuable) was often little supervised or not supervised at all.79 Thus the women seeking this favor were simultaneously marked as having sought the goddess's assistance and placed within a sphere of informal, equal reciprocity with other women who had sought or would seek the same favor. In some

THE CULT OF THE GODDESS OF TAISHAN

places, husbands were also forced to rely on the community symbolically, as well as on Bixia Yuanjun, for help in overcoming fertility problems. Their job was to procure the special red thread that had to be tied around the doll their wife chose, and although the thread cost only a few hundred cash, the custom of at least one area demanded that would-be fathers raise it by collecting one cash each from as many friends and neighbors as necessary. 80 Thus, if the couple still failed to reproduce, the public knowledge that the goddess had rejected their appeal (presumably due to insincerity) was added to their other anxieties. If their prayers succeeded, the couple (and any grandparents who joined in the rituals) were indebted to their neighbors as well as the deity: the family-and the daughter-in-law's place in it-was saved, but not without cost to its symbolic self-containment. Meanwhile, the invocation of Bixia Yuanjun may have also served to highlight the importance of the daughter-in-law, and as a reminder that the woman who bore the family's hope for continuity was herself something of an outsider: a point that Chinese ritual and ethno-biology (at least in the south, from which most of our ethnographic evidence comes) 81 sometimes seem at great pains to deny. Thus, when the normal course of yin and yang-and whatever ancestral aid had been obtained through the family's conventionally sanctioned worship at its domestic altar-had failed to produce new life, Bixia Yuanjun could bring additional powers to bear. In doing so, she affirmed the goodness of those who had previously failed to fulfill this basic filial duty and assured future worshipers for an ancestral line that had somehow failed to sustain itself on its own. Perhaps because the end being sought was so personal, and so potentially important to everyone, this invocation of non-canonical assistance was not often directly criticized, or even mentioned, in literati comments on the cult. Although many elite texts ridiculed the notion that any deity could override fate and natural forces, 82 they more often raised that point in connection with the other principal favor sought from Bixia Yuanjun-an extension of parents' lives. Both men and women appealed to Bixia Yuanjun to extend their parents' (and in the case of women, parents-in-law's) lives: a pilgrimage to Taishan for this purpose was thought particularly efficacious. 83 Bixia Yuanjun's power to grant this favor was a logical extension of her role as the merciful counterpart of the harsh Yan Wang; however, appeals on behalf of the aged raised far more complications than the equally common appeal for children and led to far more debate. First, many texts questioned whether one could justify leaving a sick parent alone to undertake a pilgrimage. 84 In general, exceptional sacrifices-

Kenneth Pomeranz

199

such as cutting off one's own flesh to succor elderly relatives-seem to have been considered less praiseworthy, and increasingly suspect, in the late Ming and Qing than before; 85 and as we shall see, getting Bixia Yuanjun to extend one life often involved risking another. It is also possible that this particular act may have again raised tensions specifically connected to the perceived roles of daughters-in-law. Requests that Bixia Yuanjun extend lives seem to have been made only when death seemed imminent without such aid; such moments of danger for the senior generation were also moments when the specter of family division loomed large, and the jealousies of wives in the younger generation were often blamed for such divisions. At least one play dating from the Yuan and one tale still popular today raise the question of base motivations for these appeals to Bixia Yuanjun. In each, two couples go to Taishan to seek longer life for a sick parent of the two husbands, but whereas the first couple is sincere, the second only wish to appear devoted; the sacrifice of the insincere couple has disastrous results. 86 Moreover, there was the question of for whom such efforts were to be made. Texts regularly describe such pilgrims as er nii, which could indicate either "sons and daughters" or "sons and daughters-in-law." 87 When women sacrificed to extend the life of their in-laws, their object was unexceptionably Confucian, even if their means were suspect. If, however, they were concerned with their own parents (as is the case in at least one tale recorded in the eighteenth century), 88 then they again entered the murky area of a woman's proper ties to her natal family. In theory, upon marriage women were expected to become members of the husband's family. In practice, however, the desire for continuing ties to one's own parents is a prominent theme in marriage laments, women's songs, and much other popular culture, including the stories of Guanyin (as the princess/marriage resister Miaoshan), and one version of the earthly life of Bixia Yuanjun herself; 89 knowledge of that longing was presumably one more reason to be suspicious of young brides. In the case of sick parents, elite hagiographers acknowledged that a daughter's continued devotion could indeed be virtuous, but also noted thatthis particular type of virtue could arouse intense opposition and be socially disruptive. 90 And since, as we shall see, sacrifices to Bixia Yuanjun to succor elders could involve great danger for descendants, any appeal on behalf of a married woman's parents raised the possibility of a daughter-in-law draining off vitality from her new family to her old one, rather than adding to its life force. (Interestingly, in the tales about such journeys that I have found, women sometimes go alone and sometimes with husbands, but men never go alone-perhaps

200

THE CULT OF THE GODDESS OF TAISHAN

because Bixia Yuanjun seems to respond more to female appeals. Thus one theoretically available way of making sure that Bixia Yuanjun's powers were deployed only along proper descent lines was apparently foreclosed by her preference for female petitioners.) Interestingly, the eighteenth-century tale cited above sidesteps these conflicts. Its heroine, a young woman who has chosen not to marry so she can care for her ailing father (and struggle to keep the family business alive), is invited to join the neighborhood women on a pilgrimage to a Bixia Yuanjun temple; she declines, saying she cannot leave his side, but asks how far the temple is. Once informed, she takes nightly walks around her courtyard once her father is asleep, holding a candle, measuring off the distance, and apologizing to the goddess for not being able to make a real pilgrimage. She finally completes the "journey" on the night before the temple "opens" for the season. Her soul is then miraculously transported to the temple, making her the lucky first petitioner to arrive. A powerful eunuch, who has gone to great lengths to be first, suspects trickery and is enraged. When the same thing happens the next night, even though he has paid the temple keeper to bar the gate, he tries to make the crowd share his disgust, arguing that this can hardly be a temple worth frequenting if ghosts can openly appear there. But the crowd does not leave; instead, someone recognizes the "ghost" that the eunuch described as the spirit of her neighbor who could not join the pilgrimage. When this is confirmed, the eunuch is so impressed by the young woman's righteousness that he showers her with gifts; her father recovers, as does the family business, and a rich merchant marries her. 91 Somewhat like the Guanyin and Tianhou stories that "square the circle" by reconciling purity with filiality, this tale allows its heroine to show a pilgrim's devotion without leaving her father or being sullied by the lewd behavior and other impieties associated with pilgrimage; it also allows her to succor her aged father without forgoing reproduction. Many other stories, as we shall see, also have plot twists that allow the devout to win Bixia Yuanjun's aid without violating the seemingly irreconcilable demands of conventional morality. Real people, of course, were rarely so lucky. All these appeals-whether for parents or in-laws-also raised another, sharper problem. Various sources refer to a widespread belief that the extra years granted a sick parent were not fully paid for by the pilgrimage and sacrifice that had sought them, but by years subtracted from the life of another family member. One notorious way to achieve this exchange was to promise that if Bixia Yuanjun cured your parent, you would return the next year and jump off Taishan's "throw body precipice." Many other

Kenneth Pomeranz

20!

sacred mountains had identical places for jumping; at Huashan an identical legend linked these suicides to vows made at its Bixia Yuanjun temple. 92 In a Yuan dynasty play widely performed in later eras, the method is different, but the idea is the same: the pious couple fulfills a vow by going to a temple atop Taishan in order to throw their son into the giant incense burner there. 93 While we have no way of knowing how many people actually died (or sacrificed their children) in payment to Bixia Yuanjun, the practice is referred to in virtually all Ming descriptions of Taishan and in many Qing and Republican accounts; 94 it apparently survived many attempts to suppress it, although it probably became much less common. 95 Such sacrifices can be interpreted in both Confucian and "folk-Daoist" terms. In Confucian terms, such suicides were an extreme example of a whole class of misguided acts that sprang from admirable filial piety, such as cutting off one's flesh to make a medicine for one's parents (or parents-in-law). The Ming and Qing governments were not completely sure whether such acts should be honored when committed by the uneducated, because they represented feelings that should be encouraged (and better guided), or condemned, lest others do the same. However, there was a strong general trend away from praising this sort of self-sacrifice. 96 The first Ming emperor firmly rejected an honor proposed for someone who had cured his mother by throwing his son off Taishan, and official efforts to honor such people seem to have ceased thereafter. 97 However, late Ming literati visiting Taishan continued to marvel at such devotion, even while bemoaning its results; they often noted that the same feelings properly expressed would rival the finest canonical examples of loyalty and filiality. 98 Such acts could also be understood in the context of non-Confucian beliefs and practices. The idea that the favors of certain spirits could enmesh the "favored" in an ever-deepening karmic debt, ultimately payable only with death, appears frequently in disapproved cults of late imperial China. It was, for instance, a commonplace in stories dealing with the fox-spirits that Bixia Yuanjun supervised. 99 It was also a recurrent theme in tales of Wutong, a mercurial god of wealth who reached his greatest popularity in the late Ming and early Qing; like the fox-spirits, Wutong was noted for his sexual appetite, and for the bad ends awaiting the mortals he seduced and (briefly) enriched. 100 It also figures in stories of Yan Wangffaishan's regulation of human life spans. 101 Thus, while these suicides may have reflected an overly zealous imitation of Confucian models, as some elite commentators suggested, they also resonated

202

THE CULT OF THE GODDESS OF TAISHAN

with a very different kind of piety: one in which the spirits themselves were far from being either Confucian or acceptably Confucianized. (By contrast, as we shall see, the guardians of the Guanyin shrine at Mount Emei explained the suicide leaps there in ways that clearly excluded from the cult any enthusiast who even contemplated such an act.) Real people who took this leap must have invariably died; here, too, however, popular tales allow the faithful a way to escape the seemingly irreconcilable demands of ordinary and extraordinary filiality. In tales of human sacrifices to Bixia Yuanjun at Taishan (including at least one still told in Shandong today), it turns out that those who are sincerely willing to obliterate themselves are miraculously lofted through the air and returned to safety; 102 this was also widely believed to occur at Huashan. 103 The gazetteer of Yangxin county, Shandong, records a less supernatural but also miraculous tale. A son, in despair over not having a father, goes to Taishan to kill himself, only to be stopped by a stranger; the stranger turns out to be the father he has never seen, sacrificing on his way home from making a fortune in Jiangnan. 104 In only one story known to me does anybody die at "throw body precipice." That victim is the unfortunate child of one of the insincere couples discussed above; they, after witnessing the safe return of a child that the husband's brother and sister-in-law had thrown in fulfillment of a vow, decide to make an equally conspicuous show of devotion to the goddess and to the husband's recently healed father. 105 The Yuan zaju about throwing a child into the incense burner also rewards the sincere couple's faith with their son's safe return. 106 Here too, then, the cult of Bixia Yuanjun sanctioned practices that Confucians saw as at least sadly misguided, and perhaps reprehensible; Confucian condemnations of many of these practices (especially pilgrimages) would become much more harsh and widespread in the Qing. 107 However, viewed through the lens of other elements of popular religion, these practices were both less and more of a deviation from Nee-Confucian principles than elites seem to have imagined. They were less so because the oral traditions used fortunate circumstances to blunt the moral conflicts that most worried elites: those over competing Confucian ends (purity versus posterity; putting one's parents first versus returning one's own body to them intact) and those between admirable feelings and the unacceptable practices (female pilgrimage, suicide vows) through which devotees expressed these feelings. However, they were also more at odds with Nee-Confucianism than at least some Ming commentators seem to have noted, because they were deeply enmeshed in different beliefs about the nature of deities, causality, magic, and cosmic justice.

Kenneth Pomeranz

203

For instance, when popular legends claim that true believers who leaped off the mountain were magically rescued, they absolve the cult of any responsibility for deaths by suicide, and so blunt an ethical conflict that worried Confucians, but they also continue to celebrate and claim for the cult's credit the fervent devotion that such leaps demonstrated. This constituted a resolution of the problem precisely because it reconciled orthodox moral ideas with popular religious ideas about the miraculous, karmic debt, a personal deity and the power of fervor per se to change the workings of nature; but while many Ming literati seem to have been content to accept this blending, later elites were less willing to do so. By contrast, the Guanyin pilgrimage center at Mount Emei, faced with the same problem-a need to absolve itself of responsibility for the vowfulfilling suicides that occurred there-found an answer that distanced itself far more from the sorts of cosmic forces to which literati and the state were hostile. According to a nineteenth-century pamphlet issued by the shrine's clergy, these suicides were the acts of hypocritical non-believers, who went insane as they neared the holy site. 108 This explanation, in many ways the opposite of the one associated with the Bixia Yuanjun shrines at Taishan and Huashan, not only upheld orthodox values but also stayed firmly within orthodox limits on what sorts of powers cults could claim to harness on behalf of those values: it took no credit for inspiring an extreme willingness to sacrifice, nor did it assert that such fervor could move deities to suspend the laws of nature, or accept the premise that a true believer might think that they had incurred an obligation to make this kind of offering. The contrast with Bixia Yuanjun is particularly striking because the sorts of miracles people sought from the two goddesses were so similar, and because partisans of both cults had sometimes claimed the two goddesses were one and the same. In sum, the Bixia Yuanjun cult was not only large enough to be important in its own right but opens important windows for us on both the extent and limits of shared religious ideas in late imperial and modern China. Not just any popular cult could be accepted into the official and elite canons, even if it had made an auspicious start in that direction, with imperial assistance; nor were the only excluded cults "heterodox" ones. While Bixia Yuanjun's followers generally called upon her for ends as orthodox as those for which people appealed to Tianhou or Guanyin, she was associated-both through the social makeup of her cult and the content of its legends-with sources and kinds of power that were necessary but that the self-appointed guardians of orthodoxy often saw as dangerous.

204

THE CULT OF THE GODDESS OF TAISJ-IAN

The most important of these associations and tensions were related to problems of gender roles. The leadership roles that the cult offered to women, the chance to participate in a religious sphere beyond the ancestral altar, and the cult's willingness to celebrate possible sources of female power besides maternity and the eventual assumption of the mother-inlaw role were probably quite important to its appeal. Conversely, the importance of midwives, matchmakers, and other such women in the cult's leadership, the phenomenon of female pilgrimage, Bixia Yuanjun's own sexuality, and her apparent affinity for the obviously necessary but still threatening figure of the young daughter-in-law made it hard for some people to accept the cult as orthodox, no matter what its worshipers may have thought. But even these conflicts over gender may be seen as one particularly important manifestation of a more general problem: the cult's potential to serve as a conduit both for social groups that had important functions but no status in the Confucian order (eunuchs, yamen runners, matchmakers, and, in some sense, daughters-in-law) and for kinds of fervor, religious experience, and magical power that had no acknowledged place in orthodox religion and cosmology. Essentially the cult offered these forces as additional resources to those who had failed to attain perfectly acceptable ends through conventional worship and so implicitly offered both its social and religious powers as auxiliaries to the existing order: the acceptance of Bixia Yuanjun into the official pantheon would have symbolically sealed the bargain. But as the standards of Nco-Confucian orthodoxy rose in the Qing-particularly, though not exclusively, in regard to female behavior-and were applied to a larger portion of the population, and as the stakes and intensity of a numerically burgeoning but often underemployed rural gentry's struggle against other rural powerholders (particularly yamen runners, but probably also healers, Daoists, and others) rose, more and more literati found any sort of association with such powers undesirable; they preferred to reject Bixia Yuanjun, interrupting and reversing her march towards "standardization" 109 and canonization. No matter what public protestations of orthodoxy the cult's devotees made-probably sincerely-it became increasingly difficult to paper over the incompatibility between the core of the cult's appeal and purpose and the ideas that elites preferred to project about social, familial, and cosmic renewal.

CHAPTER

8

Style and Suffering in Two Stories by "Langxian" Katherine Carlitz

0T he late Ming vernacular author Langxian (Free-spirited Immortal)

U

was a master of the disturbing detail. However violent the sources for his stories, Langxian's cool narrative touch made that violence weirdly plausible and hence all the more horrific. The widow in the source he used for "The Martyr," twelfth in his collection Shi dian tou (The rocks nod their heads), murders her husband's murderer and then commits suicide. To bring out the hyperbolic violence of the deed, Langxian shows us what the servants find next morning: tiny footprints left by the wife running back and forth in a sea of blood. 1 The size of the footprints reminds us that this most inharmonious of actions was committed by a woman of Confucian refinement, the daughter and sister of near-sages. The story closes with her father's exultation: "Only I could have produced her!" Are we meant to think that the literati vision requires cathartic violence? The eleventh story in The Rocks Nod Their Heads, "The Siege of Yangzhou," is so disturbing that many modern editions leave it out, although the story's source in the Old Tang History was a staple of Ming texts for women. 2 In a chilling moment near the end of the story, the heroine steps outside to declaim the sacrificial essay she has written for herself. Trapped by the late ninth-century revolt of Bi Shiduo, besieged Yangzhou has run out of food, and the inhabitants are reduced to selling, cooking, and eating each other. The story's heroine has sold herself to the butcher for her husband's traveling expenses, so he can return home to care for his mother. In her speech she dedicates herself to filial piety, chief among the canonical relationships (the gang chang) that hold society together. There is only one problem: the heroine, trapped in Yangzhou by her husband's merchant travels, reads the essay in her native dialect. No one can understand her.

208

TWO STORIES BY "LANGXIAN"

The heroine, however, was talking to her landlord and the butcher only minutes before, and they understood each other perfectly. The sudden intrusion of the dialect problem is a completely unmotivated detail. What is it doing here? Are we meant to question the transforming power of moral example? In both "Siege" and "The Martyr," Langxian took well-known tales or topoi and retold them in what had become the sophisticated written vernacular, in which a rich particularity calls into question the formulas of source materials and undermines the maxims of the storyteller-narrator who was a hallmark of the Ming dynasty vernacular story, or htfaben. The debate over the degree and kinds of irony this produced h~s been lively, but, as Keith McMahon points out, the conventional universe is inevitably destabilized when vernacular fiction brings formulas to "noisy" life. 3 Langxian, as we see from these two examples, wrote stories that asked fundamental questions of the conventional universe. Why study Langxian, about whom nothing is known beyond the association of his pen name with certain stories and songs, and hints of literary collaboration with the noted Feng Menglong ( r 57 4-r 646}? 4 The answers are that Langxian's oeuvre, as reconstructed by Patrick Hanan, displays a subtlety and skill that have yet to receive their due, and that he wrote what we cannot avoid reading as politically engaged fiction, in the broadest sense. 5 The very enterprise of literati vernacular fiction, as we will see, was by late Ming times bound up with issues of literati self-construction. In his choice of source materials, Langxian took on major state-sponsored norms and struggled with what they meant for men of his socialization and status. The late Ming literary scene abounded in responses to the paradox of power, power that seemed to be arbitrary and everywhere, but whose locus shifted in ways difficult to apprehend, much less control. Literary responses ranged from fascination with the heroic, as in the adulation and repeated transformation of Shui hu zhuan, to the indictment of hypocrisy and corruption, as in the monumental ]in ping mei. Or one might try, with Feng Menglong, to transcend conventional statements of the norms on which the state claimed to base itself: Langxian's source for both "Siege" and "The Martyr" was Feng's Qing shi lei liie (Classified compendium of the history of sentiment}, which twits the didactic tradition by ranging faithful courtesans next to gentry wives, and letting heroes and heroines prove themselves by murder. 6 When Langxian brandished his heroine's avenging sword, he was taking on Ming power in a particularly subtle way, dramatizing the literati double-bind, the feeling of many disappointed literati that their attempt to live up to the highest ideals of

Katherine Garlitz

209

their ru or "Confucian" education was more likely to see them punished than rewarded. Giving these stories their due requires that we not separate questions of genre, gender, economics, politics, and power. Langxian dramatized the experience of the literati double-bind by creating an icon of suffering virtue, namely, the sort of heroine we will meet in the stories discussed below. These heroines are purified by having all possibility of reward stripped away. They oppose an essentially idealistic or spiritual response to the evils that beset them. This sort of self-presentation is well attested in the late Ming repertoire of literati stances, in which not only the martyr to principle but principle itself, personified as the Sage, could be seen as proving its authenticity by suffering. One of the aims of this chapter, however, is to show how this purified heroine depended for her form on the expanding material culture around her. Langxian resolves tensions by creating a libidinally satisfying sufferer to identify with, but the materials for this resolution are drawn from a context of opportunity, not deprivation, a context that had an enormously transformative effect on literati self-construction in all its varieties. Langxian exposes the difficulties of true ru cultivation in his society, but far from repudiating that cultivation, he creates new ways to claim it for oneself. By enlarging the repertoire of the shi (the "scholars," who considered themselves the standard-bearers of ru culture), he ultimately contributes to the social power of his status-group. We will see below how productive it is to explore the oppositional aspect of the literati choice to write vernacular huaben. But literati vernacular authors must also have been responding to a sense of expanded opportunity, both in the material means of expression and in the very experience of variety that economic growth and diversification offered them. Studying Langxian requires us to consider the late Ming construction of vernacular authorship itself; this will involve examining the way many literati were attracted to violence and catharsis, and the longing they expressed for unvitiated emotion. Hints of intertextuality in Langxian's stories, especially his probable references to the novel ]in ping mei, will help us to understand his rhetorical method, and this intertextuality will also give us a sense of the way fiction and drama were experienced as a large and interconnected web, so that choice of genre could function as a kind of social action, a signal to others of like mind.? To understand the cultural function of Langxian's stories, we will have to conduct a sort of hermeneutics of the huaben, looking at the ways characters are coded good or evil, solutions possible or impossible. We will have to remember

210

TWO STORIES BY "LANGXIAN"

the full range of connotation of ru, which meant literati socialization as well as a tradition of Confucian texts, and of shi, which denoted cultural privilege and a flattering sense of moral, as well as social, superiority. And we will have to read these stories keeping Ming rather than modern presuppositions in mind: the fact that a gesture is not rewarded may not devalue it in Langxian's universe; rather, the reverse may be true. We will look first at Langxian's literary world and his place in it and second at the sort of huaben his colleague Feng Menglong was collecting, which must have suggested horizons of possibility for Langxian's own writing. We will then look at "Siege" and "The Martyr" in more detail. To understand the epilogues in which the heroine of "Siege" becomes a heavenly inspector of filial piety, and the heroine of "The Martyr" is vindicated by the apotheosis of her dagger, we will draw on the stories' hints of intertextuality. All of these explorations will then be used to place the two stories back into the world sketched out below.

Langxian's Literary World Langxian's voice reflects the complex interaction of many factors. Decisions of the early Ming emperors had ensured that Ming literature would be saturated with ru language up and down the social scale, in varying degrees of self-consciousness. Since Confucian categories shaped the visible language of Ming power-the schools, the shrines, the endlessly republished classics-it is no surprise that the norms and institutions of this visible language of power became newly prominent throughout fiction and drama in the Ming. The young merchant protagonist of a Yuan zaju might be recast as an examination candidate in the Ming chantefable rendition of his story. 8 By the late Ming, formulaic popular literature would include miracle stories about Confucius and his disciples. 9 In one late Chenghua era (1465-87) chantefable, the emperor's sister resolves a hair-pulling match with his concubine by quoting the Great LearningP 0 And true to the potential tension between principle and practice at the heart of almost all the Confucian classics, these genres all had ways of using "Confucian" language to interrogate the secular power relations supposed to be governed by Confucian teachings. In another of the Chenghua era chantefables, "Judge Bao Subdues the White Tiger Spirit," the spirit tries to devour an examination candidate so that she, too, can become a magistrate and have more people to eat.U In yet another, "Emperor Ren Acknowledges His Mother," the ubiquitous solver of problems Judge Baa forces the fourth Song emperor to pay a son's reverence to the filthy

Katherine Carlitz

2II

old woman whom intrigue and jealousy drove from the palace shortly after she bore him. Here the language of hierarchy is gleefully turned on the emperor himself! 12 Seen in this light, Langxian's engagement of ru norms is a culmination rather than an innovation; in the development of narrative technique he stood on the shoulders of giants, namely, the authors of the sixteenthcentury masterworks among novels. The authors of the novels were, by the nature of their works, simultaneously creators, anthologists, editors, and critics, since all these novels transformed their sources by incorporating other kinds of material: poetry, ritual texts, historical documents, drama, other fiction. This kind of integration and synthetic juxtaposition automatically introduced varying levels of diction, generic contrasts within monumental single works, ironies of many kinds, and a habit of nonreductive allegorical reading and writing that will direct the interpretation of Langxian in this chapter. The "Neo-Confucian" qualities that Andrew Plaks sees in the masterworks of fiction thus have their roots in the early emperors' enshrinement of ru texts and categories, but, as Keith Mc.Mahon points out, the very mimetic power of vernacular fiction ensured that fiction would continually interrogate and renegotiate those categories. 13 We can assume that Langxian knew the novels well; he seems almost eerily familiar with fin ping mei. 14 But why were literati authors attracted to the vernacular genres at all? This can be traced in part to new notions of the genuine that made the huaben seem a privileged form of self-expression. The earliest decades of the dynasty witnessed a startling new valorization of the interior life, 15 and by the middle of the sixteenth century the primacy of Wang Yangming's heart-and-mind was felt everywhere in literati writing. The Jiajing era (I 522-67) statesman-turned-playwright and teacher Li Kaixian (I SOI68), for example, looked askance at the excesses of overenthusiastic Taizhou preachers, but he assured his academy students that if they eschewed artifice and deforming ambition and followed their "childlike hearts" (tong xin ), they would assuredly pass the civil-service examinations. 16 For someone like Li Kaixian, a champion of vernacular drama and song, the childlike heart and associated notions of the genuine (zhen) were not neutral but oppositional: they were part of his arsenal to fight against the artificial learning and writing he associated with the corruption that cost the truly worthy (among whom he numbered himself) their rightful positionsY And because the notion of the genuine was so often used to castigate opponents, our first temptation may be to understand the literati use of the vernacular primarily as a gesture of opposition. Certainly there was

212

TWO STORIES BY "LANGXIAN"

much for literati vernacular authors to oppose. The litany of Ming literati complaints is well known: shidafu were vulnerable to quotas, factionalism, and pure bad luck on the examination route; the examination curriculum that in fact widened opportunity was despised as oversimplified and a hindrance to true literary cultivation; and the government was seen as failing in its responsibilities, even as its unprecedented reach increased the likelihood of unpleasant contact with it. Cumulatively, these factors gave rise to a sense on the part of many mid- to late Ming intellectuals that the state, by basing itself explicitly on ru tradition, had vitiated that tradition. 18 Coupled with this was the recurrent fear that literati writing was forever losing its grounding in the genuine and degenerating into ambition or artifice. The vernacular voice could be exalted as a corrective to this. If institutions were untrustworthy, one answer was to elaborate a voice that could be experienced as genuine or sincere in a way that the institutions were not. On the one hand, the vernacular genres had become sophisticated literati instruments by the late Ming; on the other, Feng Menglong's assurances that he could teach the people by using language that "strikes the common ear and affects the constant heart" reminds us how these genres were conceptualized as speaking the language of the people, who, by virtue of their exclusion from shi channels of advancement, had access to genuine emotion untainted by ambition. 19 Ming literati worked this theory out most explicitly in their discussions of song-poetry (ci), 20 but given the fact that the same men were collecting, editing, and imitating song, drama, and vernacular fiction, it is hard to imagine that the justification did not cross generic lines. This had nothing to do with social integration: Li Kaixian championed the expressive power of the poetry of the people, but the joke epitaph he wrote for a charcoal-seller shows that actual people were meant to stay firmly in their place. 21 (The people, understood as those in need of government from above, were after all an indispensable social and material medium for self-defining shi activity.) And it was Ling Mengchu's (rsSo-1644) fellow examination-sufferers, not "the people," who "slapped the table in amazement" upon reading his stories. 22 The relevant ideal, rather, was that of the people's privileged form of self-expression. One wrote in the vernacular genres as a literatus: Feng Menglong, as Hanan points out, differentiated himself from the "village schoolmaster" by holding that true literary cultivation was required in order to use the "common language" morally and effectively. 23 Nevertheless, by writing poetry and fiction in a voice conceptualized as being derived from the untainted folk, one added the magical efficacy of

Katherine Carlitz

213

their untainted self-expression to the moral and social power of oneself understood as a shi. And finally, as noted above, one thus created a new and valid way to be a shi, a new avenue of shi self-realization that was sorely needed, given the institutional constriction of access to shi power and prestige. Treating choice of genre as a gesture of opposition has a great deal of explanatory power for Langxian's stories. What we will see .in "Siege" and "The Martyr" is a struggle with the very idea of public morality, which for Langxian seems to have exemplified the vitiated, state-sponsored ru tradition. Elsewhere Langxian is savage about the opportunistic way authorities trot out the language of the canonical relationships: in his story "Wang Benli," a powerful eunuch welcomes the aged mother he has brought from afar-but the narrator tells us in an aside that the eunuch, finding his real mother too poor to be seen with him, has hired a prostitute instead! 24 True morality, in Langxian's stories, is a matter for individual self-cultivation, or for the family that closes ranks to use the institutions of the empire for its own purposes. The huaben, with its formal opportunities to interrogate convention, and its association with the idea of the genuine, obviously seemed an appropriate instrument to express these ideals of true morality. But treating choice of genre as a gesture of opposition cannot fully account for the flowering of vernacular genres among Ming literati. These authors cannot have been motivated solely by negative factors. We must remember that despite institutional and ideological bottlenecks, what set the Ming apart from preceding centuries was a complex wealth-driven information explosion _IS The hand of the state may have been heavy, but its increased reach meant vast numbers of new recruits to carry out state functions. The state-sponsored ru tradition may have been narrow, but the official enthusiasm for Confucian badges of honor, when coupled with the spread of printing and publishing, resulted in a flood of new xianlevel gazetteers, with their lists of local heroes and martyrs. The print explosion and the new commodity market in books led to a rich interplay of mixed messages, as the gospel of virtuous shi poverty was preached in exquisite, well-printed compendia of model lives/6 The material abundance of their environment transformed the idea the literati had of themselves: gazetteers and beautifully printed exempla were a new material medium for their exaltation of transcendence of the materialY In this paradoxical way, Ming economic growth fostered the legitimation of shi ideological hegemony, and-however much he suffered-it is from this position of group empowerment that Langxian wrote.

2I4

TWO STORIES BY "LANGXIAN"

Nor can the form of what he wrote be isolated from Ming economic growth and diversification. Ming vernacular literature grew in mimetic power during a period when there was simply more and more to imitate: more kinds of goods, and increasingly complex relationships revolving around the possession and exchange of those goods. Not only were Langxian's stories, like many other Ming stories, both vernacular and classical, increasingly set in the merchant milieu, but the very length and scope of the chuanqi drama and the masterwork novels suggests the confidence of a culture whose material base is expanding, affording new opportunities. This confidence is apparent even when the language of goods is explicitly presented as the language of corruption, as in ]in ping mei. (Such an exuberant indictment could only have been framed in the robust world of new and ever more refined commodities.) The correspondence of the literati huaben author Ling Mengchu, who described his stories as written for an impatient publisher, reminds us that the extant collections of vernacular stories would not have come into existence without the burgeoning Ming book market-even when these huaben were written by literati whose group self-definition depended on the ideal of disinterestedness, the productive fiction that the literati, at their purest, were above pecuniary concerns. 28 For someone like Langxian, the spread of books and printing meant both an increased opportunity to publish and a new wealth of source materials. The fat compendia of model lives written to satisfy the growing market in books eventuated in Feng Menglong's Qing shi, which in itself must have satisfied a public taste for daring novelty as well as Feng's own critical stance toward the vitiated ru tradition. And the way Langxian naturalized the brief classical-language accounts he took from Qing shi, translating them into formal-realist huaben, was part of the Ming advance in mimetic technique related above to economic change. The excess and violence that run through so much of Ming literature can be related to all these factors. The valorization of excess could certainly express opposition to ru constraints, but it could equally be a valorization of sensation itself, fostered by the enrichment of the Ming material world. And, paradoxically, the increased representation of violence can be thought of as a consequence of the very empowerment of the literati, since a look at certain aspects of Ming culture that have received little discussion shows us the extent to which emotions that we can call "violent" were part and parcel of Ming literati self-definition. The language of Confucian categories could express the most violent discord, in various genre-specific ways. The early Ming emperors had,

Katherine Carlitz

215

after all, institutionalized Song dynasty Confucian teachings by means of violence and bloodshed. For all our attention to classical Confucian paradigms of social integration, and for all the explicit emphasis on the zhang yong or ideal of the Mean in Ming texts, the fact is that a current of violence runs through all Ming writing: horrifying violence against the virtuous, satisfying vengeful violence by characters who (even when not conventionally virtuous) clearly have their authors' sympathy. In the tales of martyrs written up in the burgeoning gazetteers, the gang chang were almost always upheld by means of horrific ordeals: political martyrs have their tongues torn out, legions of women are flayed or disemboweled resisting invaders. These tales of martyrdom are the kinds of stories that made violence against the virtuous a central feature of Ming exemplary literature. But virtue could be upheld by similarly violent means. The influential Qiu Jun (1420-9 5 ), whose massive Daxue yanyi bu (Supplement to the Explication of the Great Learning) may have helped earn him a promotion to grand secretary in 1491, assembled in the Supplement a chapter of extracts from Song dynasty thinkers who give tacit permission to commit acts of filial vengeance. His Song thinkers are forced to concede that while taking vengeance into one's own hands threatens essential social hierarchies, still the desire to avenge one's father or family is a natural impulse arising from the essential goodness of human nature. Qiu Jun even described the mutual production and destruction order of the Five Phases as a cycle of acts of filial vengeance, water slaying fire for destroying water's progenitor metal, and so on. 29 In the realm of literature, the almost ungovernable Li Kui in Shui hu zhuan, among others acts of violence, rescues his leader Song Jiang by leaping about naked with a broadax (chap. 40) and beheads the kidnappers of his host's daughter (chap. 73). 30 Li Kaixian, cashiered from high office in the Ministry of Rites, rewrote an episode from Shui hu zhuan as the chuanqi drama Bao jian ji (Jeweled sword), so that his hero Lin Chong could see to the execution of his enemies (who look and sound a great deal like Li Kaixian's depictions elsewhere of his own enemies}Y Li Kaixian, one of whose mentors had been beaten at court during the great ritual controversy that initiated the Jiajing reign period,32 invoked the gang chang or canonical loyalties as the justification for having Lin Chong's enemies executed in Jeweled Sword. And Judge Bao and his subordinates, in the fifteenthcentury chantefables, boil transgressors in oil, roll them in barrels studded with spikes, make them walk on beds of knives, heat them up in iron pots-all standard punishments of the afterworld. On the one hand, as Tanaka Issei has shown, well-loved plays like Pi pa ji (The lute) were

2!6

TWO STORIES BY "LANGXIAN"

performed from scripts of increasingly refined expression as one moved up the social scale. 33 But, on the other hand, the turbulence of the Shui hu tradition was irresistible to many Ming literati, who might refine it, retell it, or reshape it (as in Li Kaixian's case) to express their own concerns. Theirs was a sort of double vision: Shui hu zhuan was recited in the same gentry households that might sponsor a ritual performance of Qiu Jun's ponderously didactic play Wu lun quan bei (The five relationships complete and perfected). 34 The chantefablcs in which Judge Bao boils rapacious imperial kinsmen in oil and in which to set out for the examinations is to invite disaster were found in the family graves of a level-five official. 35 The fact that disorder could be vicariously savored by those educated and socialized to bep order will be central to our understanding of the stories analyzed here. The fascination with the swaggering Shui hu heroes, of excess enjoyed as excess, is the most obvious manifestation of this vicarious enjoyment. More complex is what we will see in Langxian's stories, namely, a sort of glorification of violence as ordeal, the construction of heroic suffering as a means of shi self-valorization. This was the note that F. W. Mote caught several decades ago when he described the Ming veneration of the Song dynasty martyr Wen Tianxiang, who remained "loyal to the principle of loyalty when he knew his cause was hopeless." Mote cited other loyalists similarly praised, and Kang-i Sun Chang has shown in her study of the poet Chen Zilong how powerfully late Ming partisans of qing identified themselves with Wen Tianxiang. 36 All these factors-economic diversification, the publishing boom, the valorization of intense experience (including the experience of violence), the confidence evident in the scale of literary enterprises-led to the kinds of excess that Langxian both reined in and exploited. The sixteenth century produced not only the masterwork novels but also spectacular grotesques, for example, Pin shi zhuan (Biographies of indigent scholars) compiled by Huang Jishui (1509-74) during the Jiajing reign period. 37 The eccentric indigents take the ideal of disinterestedness to its extreme, positively glorying in their tattered robes, pustulant sores, and occasional meals of gruel. These renunciatory scholars depended for their meaning on a contrast with the world of goods swirling around them, a world whose prosperity made possible the vast novels, the fat compendia of model lives, and the explosion of xian-level gazetteers. This complex wealthdriven information explosion, coupled with the need for self-definition in an age of change, is the ground against which Langxian's subtle gestures are so striking.

Katherine Carlitz

217

The Construction of Langxian But who was Langxian? The most honest answer is that he is a late twentieth-century hermeneutical construct, since the name and oeuvre of the Langxian to whom works and attitudes are attributed above have stabilized only in the last two decades. While using narrative style markers to date the vernacular stories in Feng Menglong's and earlier collections, Patrick Hanan found that a group of 22 stories in Feng's third collection, Xing shi heng yan (Constant words to awaken the world), made use of style markers not shared by the rest of the extant corpus. This same pattern of narrative style markers was manifest in thirteen of the fourteen stories in The Rocks Nod Their Heads, whose author is referred to as Langxian in the preface Feng Menglong wrote for the first edition. These 22 Constant Words stories and the stories in Rocks share thematic and structural as well as stylistic features, and this is the oeuvre that Hanan assigns to Langxian, which was probably only one of the author's pen names. Hanan adduces convincing evidence to suggest that this Langxian was an editorial associate of Feng Menglong from r627 until at least some time in the r630s. 38 Deferring, briefly, issues raised by this construction of authorship, what do we find in the oeuvre thus constructed? Hanan hears a more "orthodox" voice than Feng Menglong's, a voice that "plunges us into the world of the Confucian exemplum" even though Langxian was more committed to Daoist themes than were any of the other major late Ming huaben authors, and even though Langxian shared none of Fcng Menglong's enthusiasm for the "public morality" of heroic patriotism and leadership. The world of officials and institutions in Langxian's stories is "alien, arbitrary, and even menacing," and Langxian is sometimes "stretched between two emphases" as a result: concerned to amplify a conventional didactic account, but distracted by the "social facts" of official injustice. In my view, however, these stories demand a more unified reading, despite the danger of imposing an anachronistic sort of unity. 39 We need to take into account the issues of literati self-construction raised above, and give Langxian credit for building on the achievements of the novelists whose work he obviously knew well. Langxian's engagement with social fact is a key to his method of operation, which was to naturalize the Tang tales and moral exempla that were his primary sources, by setting them relentlessly in the Ming present. A Ji in "The Old Retainer" succeeds in the world of real Ming wholesale and retail; the father and son in "Guo Tingzhi" wrestle with the minutiae of the Ming examination system; the

218

TWO STORIES BY "LANGXIAN"

eponymous protagonist of "Lu Nan" suffers the intrusion of the magistrate (and we are shown the difficulties that beset even a bad magistrate!) into the sort of garden Langxian must have known well from his years in Suzhou.40 Langxian seems to me not stretched between two emphases so much as testing the two emphases against each other. What does a woman say to her husband on the eve of filial self-sacrifice? How many sharp deals have to be cut in order to turn a widow's mite into substantial provision for her children? Is it legitimate to turn an abstract principle (for example, the loyalty of the bondservant, in "The Old Retainer") to your own needs? To a certain extent, questions like these are bound to arise in any translation from formulaic classical-language exemplum to formal-realist huaben, whose rhetoric depends on creating a convincing simulacrum of the real world. But Langxian, by insisting simultaneously on conventional normative ideals and social facts, seems to me to exploit the potential of this kind of contrast in a particularly focused way. But once again, who was Langxian? Feng's attribution of the Rocks stories to his friend the Free-spirited Immortal was a gesture that contemporaries would have understood as being both the designation of an author and, at the same time, a statement about vernacular authorship, which required "free spirits." The Ming vernacular attribution thus helped to construct the vernacular author-and it can serve as a guide to how that author may have understood himself and his enterprise. If Hanan is correct, the author of The Rocks Nod Their Heads presented himself variously as the Free-spirited Immortal, the Foolish Old Man of Nature, and the Master of the Ink-Wild Studio. 41 These names are not denotative but performative: not only is choice of genre a linguistic signal of shared attitudes, but so are the names chosen to associate oneself with a given genre. None of this gives us a biography of Langxian, but a biography might tell us less than the pseudonyms and the stories do about how language and history shaped and were shaped by literati subjects in the late Ming. (The stories themselves, representing as they do the conscious choice of genre by highly educated authors, can in this sense be considered "performative" also.) Thomas Pfau, in a recent essay on Wordsworth, concentrates on the "pragmatics" of the Wordsworthian poem, asking how Wordsworth, who actually composed with a fair degree of collaboration, engineers our assent to the Romantic idea of the individual, originary consciousness. 42 We can ask, similarly, how the huaben that Langxian wrote engineer our assent to the idea of a "free spirit" behind them, an author unconstrained by convention and thus likely to tell the truth. I have already suggested above some of the ways late Ming literati authors constructed themselves as having access to genuine emotion. When

Katherine Carlitz

219

we look below at a representative huaben collected by Feng Menglong, we will see how it too engineers our assent to the idea of an untainted inner self that, left unconstrained, would know how to align itself with Heaven. The idea of "innate knowledge" (liangzhi) elaborated by Ming thinkers originally asserted a common ground for human moral impulse, but in practice, the Ming emphasis on the inner self increasingly worked to authorize idiosyncrasy-recall the Indigent Scholars-as a means of demonstrating true moral impulse. 43 The late Ming literati-vernacular world was dominated by the idea of expressive authorship, authorship expressive of idiosyncratic motivations, as we see in the overwhelming tendency to read fiction and drama aclef. This did not lead to the evidentiary scholarship that would carefully sift texts and attributions in the Qing. Rather, motives were if anything more interesting than attributions. (The prefaces to ]in ping mei are passionate about the author's motivation and relatively unconcerned with "who" he was.) Wang Shizhen (152690) could be said, improbably, to have written fin ping mei to avenge his father's death; Li Zhi (1527-1602) was said to have championed Shui hu zhuan to express his eccentric repudiation of authority, and so on. 44 I have fallen into the trap myself by suggesting that Li Kaixian used his Jeweled Sword to behead his enemies. We could instead have done a "pragmatics" of Jeweled Sword, examining how the play constructs an ideal hero suitable for use as an ideological missile, and how the very choice of language and genre, imagined as natural and untainted, ensured that this missile would be hurled from an unassailable height. This is the sort of lens we will turn on "Siege" and "The Martyr," but first we will look at the kind of story that may have suggested Langxian's enterprise to him.

"The Apotheosis of Ren the Filial Son" Langxian's story "The Martyr" uses murder to uphold the gang chang and ends with the apotheosis of the murder weapon. At the conclusion of "Siege," the heroine is a heavenly inspector. Langxian did not invent these plot elements: his source for "The Martyr" was an account of avenging murder in Feng Menglong's Qing shi, and, even more significant for our purposes, analogues to all the plot elements mentioned above can be found in "The Apotheosis of Ren the Filial Son," a sixteenth-century story that Feng included in his first huaben collection Gu jin xiao shuo (Stories old and new). 45 This is not to suggest that "Apotheosis" was a source for either "The Martyr" or "Siege," although tantalizing hints of intertextuality will be noted below. But the plot of "Apotheosis" shows us a tradition of huaben violence and vengeance that Langxian would be able

220

TWO STORIES BY "LANGXIAN"

to exploit, and its selection by Feng Menglong shows us the fascination this sort of story held for Langxian's colleagues in the world of the literati huaben. We will see below that "Apotheosis" glorifies an act of vengeance by a hao han or stout fellow, the sort of hero who would have been at home in Shui hu zhuan. By the late Ming, some of the literati enthusiasm for the hao han carried over into the construction of authorship itself: the aura of heroic idiosyncrasy that had gathered around Li Zhi, for example, assured that his name would be attached to a number of vernacular works, and Feng Menglong himself was considered a sort of hero of the vernacular voice. 46 And one of the things that characterizes the hero, the hao han, is that he can be transgressive in the name of spontaneous truth, since he has access to the inner morality that convention and ambition would deform. (Here again we must remember the booming book market that made it possible to print-and thus to think-such unorthodoxies.) We have already seen Li Kui jumping about with a broadax, and although Feng Menglong, the author of model examination essays and eventually an official himself, was no Li Kui, still his preface to his song collection Shan'ge (Mountain songs) introduces the collection as a corrective to the "false medicine of Confucian morals," which, as Hanan points out, is an attitude shared by Li Zhi and others. 47 And although neither Feng nor his contemporaries produced a monumental sustained work like the masterwork novels, still his collecting and editing enterprise was a huge one, fitting the Ming pattern, and Qing shi, as noted above, had a daring and transgressive edge. The collection lived up to its prefatory claims of profound morality only if true ru morality was understood as passionate commitment, rather than as adherence to forms. "The Apotheosis of Ren the Filial Son" was bound to have an appreciative audience in the literati circles that enjoyed Shui hu zhuan and Qing shi. In brief, this is the story of a young man, Ren Gui, whose stolid dependability and filial piety are enough to exasperate his sensation-seeking wife. ("Every day he left early and returned late, which suited her not at all.") She soon takes up with a former lover. In the first half of the story, we see Ren Gui essentially through her scornful eyes. By accusing Ren Gui's blind and aged father of attempted rape, she compels Ren Gui to return her to her parents' home, and when he nearly surprises her there with her lover, he is mistaken for a thief and thoroughly humiliated by being beaten in the privy. But this is his nadir, and the rest of the story is seen through Ren Gui's eyes, from which the scales have fallen. Rejecting the prudent advice of his father and his employer, he murders his wife,

Katherine Carlitz

221

her lover, her parents, and the maid, and then, far from trying to escape, proclaims the deed like the hao han he is and waits for the authorities. He is the object of adulation from crowds and officials alike, and after he is taken away to be executed (with many expressions of official regret), his bonds burst and his soul rises to Heaven from the killing ground. After death he becomes a tu di gong or local tutelary deity, his filial piety (xiao) and heroic self-sacrifice (lie) having demonstrated that he can sustain the empire (baa guo). "Apotheosis" is relevant here for far more than its violence, although the beating and beheadings are graphically described. We watch Ren Gui become a hao han in the Shui hu mold: originally the dupe of an unfaithful wife, he goes from the role of a Wu Da to that of a Wu Song! Excess in the service of righteousness is made salient and explicitly valiclated-even though the authorities regretfully remind us that Ren Gui has committed "extra" murders and so cannot escape punishment, still Heaven gathers him up in a way that sweeps such a caveat aside. In a detail of the sort McMahon emphasizes, Ren Gui literally escapes containment, by jumping the wall of his employer's garden as he sets out to have his wife's head. And that which cannot be contained is identified in the closing poem as the hero's spirit or jing shen: No matter how many stones or iron chains, Spirit alone can never be effaced. Risking his life vile lewdness to repel, He rivals in courage the very King of Hell.

What the apotheosis and this poem "engineer" for us, in Thomas Ffau's sense, is the most powerful possible coding of an indwelling spirit that knows "how to conform to the cosmos, a spirit for whom the conventional formulations of morality are simply irrelevant. This did not by any means have to result in a repudiation of such norms as hierarchy, fidelity, filial piety. Rather, it validated the indwelling spirit that understood those norms better than the authorities, a spirit that could not be contained by ordinary prudence. We need to look carefully at the way this story codes virtue, excess, and containment if we want to understand its cultural function and use it as a register for calibrating Langxian's literary effects. The violence in "Apotheosis" may look subversive or carnivalesque, the vigor of the hao han overturning the rules of the magistrate. But we need to remember that as collected in Feng Menglong's Stories Old and New, "Apotheosis" was

222

TWO STORIES BY "LANGXIAN"

being offered for the delectation of those who were socialized to become magistrates, to make the rules. (Despite the piety of his prefaces, and recognizing that Feng's self-understanding as a moral teacher was a very productive fiction, Feng the aspirant to officialdom can hardly have wanted to teach "the people" that Heaven would reward bloody murder.) 48 Ru socialization required the literatus to imagine himself as occupying a position of cultural superiority, and for those who never realized that position by becoming officials, it offered another option, that of transcending mere rules in the name of ru principle. We may wonder why Ren Gui's filial piety is emphasized in the title and conclusion when it has no real plot function in this story of adultery and vengeance, and the answer is that filial piety, the Ming value par excellence, functions in "Apotheosis" to code Ren Gui as having the right to transcend the rules. And the story transcends the rules in a most libidinally satisfying way, offering sex, food, and violence and, most important, vanquishing humiliation. "Apotheosis" is thus a cathartic story, and ultimately a very uncomplicated story. We meet with surprises, but no ambiguity: no one's motives, virtue, or villainy are ever in doubt. This was the sort of ground from which "The Martyr" and "The Siege of Yangzhou" grew, but we will see that Langxian's stories offer us no such easy resolutions. "The Martyr" and "The Siege of Yangzhou" The most obvious difference between "Apotheosis" and these two Langxian stories concerns gender. "Siege" and "The Martyr" are stories of suffering heroines, rather than swaggering heroes. On the one hand, Langxian's ru heroines are a sort of quintessence of the literatus himself, cultivated, discerning, and disinterested. But even if we see these heroines as occupying what were essentially male subject-positions and as dramatizing essentially male issues, Langxian could make powerful use of the way heroines were coded as women in the ru didactic tradition. Morally powerful but socially powerless, shielded from ambition by their exclusion from the examination system, his heroines are ideal instruments to dramatize the aspirations and the setbacks of the idealized shi. We see this particularly clearly in "The Martyr," whose heroine is something of a hao han herself, improbably dispatching the same number of victims as Ren Gui. Shentu Xiguang, the ru heroine of "The Martyr," was already fully formed in Feng Menglong's Qing shiY There she is a treasured daughter and accomplished poet given in marriage to a carefully chosen young scholar, only to find herself coveted by the landowner Fang Liuyi. Fang

Katherine Carlitz

223

uses his vast wealth to have her young husband, Dong Chang, framed and executed as a rebel. Hiding a dagger in her clothing, Xiguang agrees to marry Fang if he will first build a tomb and give her husband a proper burial. She stabs him to death on the night of their wedding. Announcing that Fang is ill from overindulgence, she calls in his entire household and kills them one by one. She then decapitates Fang's corpse, carries the head in a bag to her husband's tomb, and offers it in sacrifice. The next morning she summons a crowd, announces that her vengeance is complete, and hangs herself. This account in Qing shi is brief but brilliantly constructed. After leisurely quotation of Xiguang's poems and a retardation of the action while she wheedles Fang into arranging for Dong Chang's burial come the rapid-fire efficiency of the murders and Xiguang's resolute disposal of herself. Langxian's additions to this brief account-a magical murder weapon and families and histories for all the protagonists-introduce ambiguities but make this huaben "quieter," if anything, than its source. The Xiguang of "The Martyr" is no longer the sole focus of the story. "The Martyr" begins with Dong Chang's irritation at his stepmother's lack of refinement and his despairing search for a wife (what girl of good family could possibly want to come live with them?). The heroine's father, Shentu Qian, appears as a sort of magic benefactor, who recognizes Dong Chang's brilliance, dispenses entirely with the customary gifts and dowry, and gives Dong his daughter in marriage-even though he detects an aura of doom about the young man. A whole world of male shi relations and supernatural shi discernment is thus established in the huaben before Shentu Xiguang is even introduced. Moreover, it is the hapless young husband who brings his own world down around himself and his wife. In his irascible rectitude, he shames his pleasure-seeking stepmother and her friend Old Mother Tao, and Mother Tao revenges herself by bringing Xiguang to the landowner's attention. Dong Chang cries out helplessly that he is "a man of books and learning," even as the bribed officers of justice lead him off to execution. It is the father's exultation that closes the story, as his half of the family's jeweled sword (Xiguang has used her half to murder Fang Liuyi) rises into the sky, rejoins his daughter's, and flies off as a dragon. Fantastic as this story is, it is brilliantly naturalized. Shentu Qian meets Dong Chang at the sort of ritual dinner for official students that was held in every county. We get the stepmother's furious point of view when she is left out of the wedding festivities (Dong Chang was supposed to ask her a second time!). When Xiguang's body is discovered after she hangs herself,

224

TWO STORIES BY "LANGXIAN"

the wheels of justice grind as they would in any gazetteer, reports moving up through the proper channels and family members taken away for interrogation. This modus operandi does not allow for the uncomplicated glorification of excess that we saw in "Apotheosis," where Ren Gui's five victims-three of them "extra"-were proof of his vigor. The five bodies in "The Martyr" represent a muting of the source in Qing shi, where Xiguang simply keeps on until she knows that the household has been exterminated. In "The Martyr," where we are told repeatedly that there are exactly five victims (can Langxian have had "Apotheosis" in mind?), a verbal stratagem maneuvers each victim into place, heightening the horror but slowing the pace. Xiguang cheerfully forces Old Mother Tao to drink herself into a stupor, calls the two maids in to help their sick master (she scolds them for forgetting a candle), and persuades Fang's eight-year-old son to open his bedroom door by fabricating a summons from his father. (There are no other female relatives present to swell the number of victims; Fang's reputation for lechery is such that none of his female relatives will visit him.) The overall effect produced by this naturalization is that of improbable plausibility, which arouses a sense of cognitive dissonance rather than cathartic delight. And unlike the heroine of Qing shi or the Ren Gui of "Apotheosis," the Shentu Xiguang of "The Martyr" does not publicly announce her vengeance. At her husband's tomb she gives way instead to her own grief. Heaven hears her; the dagger she has fashioned from a piece of the family's heirloom sword flies skyward to await the final apotheosis. But while admiring officials commend her just as they did Ren Gui, the shrine they build commemorates memory and absence. She does not remain an active presence in the community as did the deified Ren Gui. What are the ideals and ambiguities constructed by this careful transformation of Xiguang's story? Given the bribery that sends Dong Chang to his death, we might well be tempted to see the depiction of Fang Liuyi in Langxian's story as a standard late Ming critique of the Ming economic boom and its attendant dislocations, especially since Fang Liuyi is just the sort of tu hao (local strongman) who bedevils earnest magistrates in countless Ming gazetteers. Langxian was able to draw on centuries of stories about tu hao preying upon vulnerable women, and in the late Ming, such anecdotes were indeed being collected, reprinted, and circulated as a standard element in expanded editions of Lienu zhuan, where they served as templates for new lienu stories. But this reading of Langxian's landowner as an indictment of the new power of wealth would be too easy and reductive, and it would not help

Katherine Carlitz

225

us get at the highly nuanced relationship that existed between Ming economic change and Ming vernacular literature. Illegitimate wealth as such is not really central to "The Martyr" the way it is, say, to fin ping mei, where we see silver ingots and small change slipping in and out of chapter after chapter, reminding us constantly that everything is up for sale. In "The Martyr," Langxian is using the tu hao for what we might call his lexical value in demonizing the illegitimate acquisition and deployment of power, since the point of those centuries of stories about tu hao preying on women had always been the disparity in power between aggressor and victim. It is in the literatus Dong Chang that money and status are problematized: one of Dong Chang's complaints about his stepmother is that in her hands, the family's limited funds are put to non-shi ends (e.g., entertaining her friends), actively hindering his attempt to find a shi bride. Shentu Qian, his father-in-law, is a true shi benefactor because he divorces true shi marriage from all exchange of material goods. The doomed Dong Chang is unable to transcend the material, whereas his father-in-law is the epitome of transcendence-but the very need to define shi nature in terms of transcending the material shows us how salient the material was. Defining the authentic shi and the relation of the shi to power are the real questions in "The Martyr." Conventional examination culture, as we might expect, is no match here for the illegitimate exercise of power. ("I'm a man of books and learning!") Only true shi violence is able to bring the tu hao down, and Langxian has been at pains to make sure we see that it is shi violence indeed, by telling us at the outset that this is a tale of retribution all the more remarkable in that the heroine is "delicate, secluded, and shy, knowing nothing but needlework, reading, and writing." The tiny footprints in the sea of blood make the deed somehow even more terrifying, since they force upon us the disproportion of power involved. And by shi violence Langxian meant violence by the authentic shi, the shi who has transcended all the forces that would distort his essential ru cultivation. (The heroine's father and brother have long since gone off to collect herbs and to live the philosophical-Daoist fisherman's life, respectively.) But by foregrounding both Xiguang and her father, Langxian suggests that this power of the authentic shi is neither indivisible nor unambiguous. The joining of the sword identifies Xiguang with her father, but the literary resonances of this very sword problematize the conclusion of the story. The murder weapon, a mere "dagger" in the Qing shi source, becomes an heirloom "jeweled sword" (bao jian) in Langxian's story. Jeweled Sword is the title of Li Kaixian's Shui hu play, in which the sword in question (transformed from a casual acquisition in the Shui hu zhuan to a family

226

TWO STORIES BY "LANGXIAN"

heirloom in the play) is the symbol of the hero's integrity. Li Kaixian's play looms larger than any other in the copious allusion to drama in ]in ping mei, and given Langxian's familiarity with ]in ping mei, we can at least hazard the guess that Langxian was evoking not only the play but the play as texturing the novel, where Li Kaixian's language is appropriated to excoriate the abuse of power. 50 In ]in ping mei the optimistic certainties of Jeweled Sword (where the sword reunites husband and wife, and justice is restored) melt away. The sword in the novel symbolizes, if anything, admirable integrity that is no match for the ungovernable world, and this, of course, is the role of the Shentu family in "The Martyr." And by giving us two ru exemplars, Langxian effects a sort of division of shi labor. Shentu Qian's exultation and the apotheosis of the sword give us Heaven's validation of the shi who transcends the world. But Xiguang's death is given to us with all the slow buildup of tragedy, the action retarded by carefully placed recapitulations (her letter to her cousin, her letter to her father), and Xiguang herself preparing for her end by disposing of her property. Hao han she may be (and Langxian lets her be "excessively" cruel and cunning to Fang Liuyi's son and the maids), but she is also a lienii, and the delectation of tragedy was part of the appeal of the Ming lienii tradition. This combination of cultural coding and narrative construction ensures that we will read her experience as one of suffering. The heroine of the Qing shi account was allowed to announce her own triumphant vengeance, but Langxian divides triumph and tragedy between his hero and heroine. The effect is to make triumph and tragedy coequal, elevating the experience of suffering itself, making it a precondition for shi transcendence of the world. Shentu Qian, knowing as he did from the outset that his son-in-law was gifted but doomed, can almost be said to have arranged his daughter's life with her tragic vengeance in mind! This is very different from the celebration of vigorous violence in "Apotheosis," in which Ren Gui's excesses continue to sustain the empire after his death. Triumph in "The Martyr" is an outsider's triumph: Shentu Qian, having dispatched five victims through the instrumentality of his daughter, will disappear. Nothing will be regenerated; no one will reform. "Apotheosis" was in its improbable way a celebration of public morality; "The Martyr"-for all that the state used the ideology of martyrdom to bolster itself-is not. "The Martyr" casts a shadow on both public morality and the possibility of transcending public morality. "The Siege of Yangzhou" casts an even

Katherine Carlitz

227

darker shadow. In the weirdly naturalized setting of besieged Yangzhou, leadership by moral example seems to have no force at all. And what the story offers by way of transcendence is undercut by hints of outcomes too idyllic to believe. "Siege" is based on the following account in the Old Tang History: Zhou Di's wife, surname unknown. Di was a skilled merchant who plied his trade in Guangling. During the period of Bi Shiduo's revolt, people were kidnapping and selling each other for food. Di was near starvation. His wife said: "If we want to return home now, we will never both manage it alive. Your parents [qin] are still living, so we cannot both die. I desire to be sold, so as to enable you to continue your family line." Di could not bear to have it thus, but his wife convinced him, and sold herself for several thousand cash. When Zhou Di reached the city gates, the guard was suspicious and went back to the market with him to check his story. There they saw his wife's head already displayed upon a pillar. Di wrapped up what remained of her body and took it home for burial. 5!

Once again Langxian chose a source with great narrative power, narrative power that ensured its inclusion in Ming exemplary literature. 52 Langxian begins his own treatment of the tale with deceptive urbanity, chatting with his readers about filial piety (scholars know all about it; unlettered folk need to be taught by example), and promising to provide the examples that women, who sit gossiping all day about food, clothing, and visits to temples, particularly need. He moves lazily toward his main story, offering first the tale of the filial daughter Cao E, who drowned rescuing her father's corpse from the waves, and then the tale of Tang furen, who revived her mother-in-law by breastfeeding her, gaining countless children and grandchildren as a reward. Like the story of Zhou Di's wife, the story of Tang furen was also included in almost every Ming expansion of Lienu zhuan, and even after repeated imperial edicts prohibiting the sacrifice of children, the story was still current enough in r 67 3 to serve as the basis for commending a woman who left her son by the roadside, so as to save her milk for her aged mother-in-law. 53 But Tang furen, Langxian's narrator reminds us, was able to live on in comfort. The heroine he is about to show us will give not her milk but her very body to save her mother-in-law. And with this, Langxian begins a story that turns the Tang dynasty account into an interrogation of the structures it was meant to uphold. The heroine of "The Siege of Yangzhou" now has a name and ru attributes: the wife of the merchant Zhou Di, surnamed Zong, is introduced as "the daughter of a ru family, who in her youth had studied the

228

TWO STORIES BY "LANGXIAN"

Rites and Odes." Childless and in their forties, Zhou Di and his wife have been impoverished by the rebellions that make merchant travel impossible. Zhou Di's aged mother insists that he go on the road again and take his wife with him in hopes of producing a child. When they demur, she races to the kitchen for a knife to kill herself! They relent and depart. The trip is a disaster. Zhou Di neglects basic rules of caution and loses their entire stake (the money his mother had saved for her coffin). Only the kindness of a Huizhou merchant who takes them to Yangzhou keeps them from starving. But in Yangzhou they are trapped by Bi Shiduo's revolt, and the rest of the story follows the Tang account, with one striking difference. In Langxian's huaben, Zhou Di's wife does not selflessly sacrifice herself. Rather, she repeatedly offers her husband the chance to sacrifice himself and send her home to care for his mother. Not only does her offer aid in delineating the story's characters, pointing up as it does Zhou Di's cowardice, but this construction of filial piety is one of the many touches (like the Huizhou merchant} that anchor the story firmly in the Ming. Only in the Ming did the intense filial bond between mother-in-law and daughter-in-law come to dominate the literature of women's filial piety. 54 All this is set against a background of commerce in human flesh, narrated with Swiftian coolness. Shops spring up specializing in children or adults, or in particular cuts of the human body. When Zhou Di's wife sells herself, she haggles over the price with the same acumen she has always shown in keeping her unfortunate family afloat. A crowd gathers when she enters the butcher's shop, but quickly disperses: "It's only someone selling his wife again." The butcher jokes with her when she wants to bathe before going under the knife ("My, aren't you a pure little piece of goods!"), and in an exquisitely placed detail that concentrates much of what is disturbing in this story, Zong writes the sacrificial essay that no one will understand with the brush the butcher uses to write his bills. Langxian concludes the story by setting it in a cosmic and mythological framework. Heaven is on the wife's side: Yangzhou is hit by a storm after she dies, and everyone who consumes her flesh perishes. Spirits see to Zhou Di's safe return: the deified Qu Yuan, disguised as the sexton of a tumbledown temple, sends him home on a magic steed. Like Ren Gui and Shentu Xiguang, Zhou Di's wife is officially commemorated: having reached Zhou Di's courtyard, the horse refuses to budge until a local "lover of virtue" begins the process of petitioning for a shrine. Zhou Di comforts his mother, who tells him of a waking dream in which her daughter-inlaw appeared to her with a \Vith a cryptic message (Zhou has "taken a hardworking concubine" wh~-has borne a son who is "both real and

Katherine Carlitz

229

unreal"). Zhou Di sets up a household altar to San Iii dafu, the deified Qu Yuan, never remarries, and cares for his mother until her death at the age of 130. The supreme god Shang Di, moved by Zong's self-sacrifice, sets her up as a heavenly inspector of the filial piety of daughters-in-law. This naturalization of the Tang account and Langxian's new ethicocosmic settingfor it are his instruments for raising questions about power and public morality. The merchant locus typical of so much Ming fiction lets Langxian use apparent conflicts between money and morality to highlight deeper conflicts more difficult to name. "The Siege of Yangzhou" is a story about shi vulnerability-a story all the more frightening in that we are given no identifiable villains and hence no opportunity for cathartic revenge on them. The selling and eating of bodies in his Tang source gave Langxian a powerful language in which to describe the experience of vulnerability. Zhou Di's terror springs to life by contrast with the concrete details of the marketplace, when he goes trembling forth to seek his wife's body by "counting the stall numbers as he went, to make sure he found the right one." Yangzhou carries on its unspeakable commerce in the most matterof-fact way: bodies are repeatedly referred to as goods, huo; the butcher has just sold out his stock and is "in need of goods" when Zhou Di's wife arrives, and although he is almost unable to bear bringing the knife down on her body, he reflects that since he has already paid four strings of cash, he has no choice. But, as in "The Martyr," the point of this naturalization is to produce the experience of cognitive dissonance, to make the unthinkable weirdly plausible. It is not to demonize commerce as such. Many of Langxian's moral heroes are merchants, and the sensible commercial advice she gives is one of the things coding Zhou Di's wife as positive in this story. Small-scale commerce is presented as being just as powerless as any other sector of society against the kinds of arbitrary forces (natural, institutional) that menace much ofLangxian's oeuvre. Floods and rebellions keep Zhou Di from exercising his metier to provide for his mother, just as he and his wife are unable to coerce their bodies to provide her with descendants. The sad banter of this aging, vulnerable family underscores their helplessness: "Mother-in-law," cries Zong, "if we had a child tonight, he still couldn't grow up in time to take care of us in our old age!" Zhou Di's wife, the embodiment of ruin this story, is made vulnerable by association with this vulnerable family, and they will be jointly powerless against the nightmarish experience of confinement in Yangzhou. As in "The Martyr," a probable reference to fin ping mei intensifies the

230

TWO STORIES BY "LANGXIAN"

indeterminacy of their situation. Zong's apparition to her mother-in-law dangles before her a half-hideous vision of the posterity that had been promised: a son "neither large nor small, neither true nor false," with the unkempt hair and dirty face of a demon, and moreover the very image of his father. This image of a sort of virtual son who may or may not be his father recalls the powerful indeterminacies that close ]in ping mei and helps render it difficult to take the supernatural resolution of "The Siege of Yangzhou" straight. 55 And just as Dong Chang's examination learning was no match for the forces that menaced him, so the limited notion of filial piety that sends Zhou Di's wife on her way is no match for the forces to which she ultimately submits. But as noted above, the nightmarish quality of the story results from the fact that we never do see a true villain. On the one hand, Zong has known from the beginning that the demands of filial piety, a cornerstone of ru self-definition, are being misunderstood when she and her husband are sent off on their fruitless travels. She has made the commonsense suggestion that she care for her mother-in-law at home and her husband take a concubine on his travels to get a child. But her mother-in-law will have none of this: "Why should a concubine get all the glory when you've done all the work?" The mother-in-law's misguided demands have put Zhou Di's wife in an intolerable double-bind. The competence she must display as an ideal ru daughter-in-law requires her to silence her husband's complaints about his mother, even as she knows that the mother-in-law (with the best intentions in the world!) is sending her to her death. On the other hand, Zhou Di, his mother, and the butcher are all victims. The story has constructed a world in which one unnecessary stupidity leads to another, with no possibility of escape. The fact that the junzi or superior moral figure here cannot get an appropriate response from the "small men" about her is what elicits the story's moments of rage. Zhou Di's wife has been unfailingly constructed as the reliable voice in this huaben (her careful management in faminestricken Yangzhou has even ensured that she will still be plump enough to fetch a good price), but the conversations in which she offers Zhou Di the opportunity to sacrifice himself show her competence shading over into exasperated cruelty: Zong laughed coldly. [Zhou Di has once again begun to weep and blame his mother for their misfortunes.] "You can cry all day today, tomorrow, and the day after tomorrow, but it won't get us both home. I remember that in ancient days Zuo Botao and Yang Jiao'ai were on the point of freezing and starving to death, and in the end they sacrificed one to save the other.

Katherine Garlitz People are now being butchered and sold in the market-that at least would bring us a little cash. Whether I sell you or you sell me, it'll get one of us home to care for your mother. You decide who it's to be." When Zhou Di heard her speak of selling bodies for cash, all his own flesh began to quiver and twitch. He waved his arms and cried: "It's out of the question!" Zong laughed again. "All right, we'll both starve, and someone else will get a full belly eating us!"

Zhou Di attempts to rise to the occasion and offer himself, but is unable to bring himself to go to the butcher. His wife, under the guise of checking out prices, takes the innkeeper with her and completes the transaction for herself, requiring only that she be allowed to return home briefly to give the money to her husband. Here she plays with his emotions in a way that can only be described as vengeful: Zong took the four strings of cash home, placed them on the table, pointed to them and said to her husband: "This is your mother's money from the sale of her son. Quick, off with you to the market. I'll take the money for my traveling expenses and go home to care for her." Zhou Di's soul wouldn't stick to his body; his face was the color of ashes; he wanted to answer but the breath seemed stuck in his throat; three or four times he stretched forward his neck, but could not utter a sound. Tears the size of yellow beans pulsed from his eyes. Zong watched all this and started laughing: "Well, it looks as though we can't make this business deal. I'll go back and call the whole thing off."

Thereupon she returns to the butcher, bathes, declaims her essay, and submits to the literal dismemberment of her ru attributes. As the basically genial and accommodating butcher (reflecting on what the four strings of cash require of him) brings down the knife, "in a moment he had the filial [xiao] and heroic [lie] Zong laid out in pieces on the table." To what end has she died? The nominal successes of the story run counter to the normative ideals of ru success. Zhou Di's wife makes a public announcement of her self-sacrifice-but no one reforms. Her motherin-law is kept alive-but with no posterity (and with none of the joyous rejuvenation of the Tang furen story, where the mother-in-law's teeth grow back, her eyes grow bright, and her hair turns black again). Heaven rewards Zhou Di's wife, but with none of the community-consolidating fanfare of the Ren Gui story. As a heavenly inspector of the filial piety of daughters-in-law, she supports the very institution that failed her. And Langxian pointed to the literati analogues of this experience by the way he opened and closed "The Siege of Yangzhou." The prologue tale of Cao E may seem at first merely to offer what was promised, namely, a

232

TWO STORIES BY "LANGXIAN"

standard anecdote typical of those used to instruct giddy women, but Langxian actually supplied more than the typical Lienii zhuan account focused on Cao E hersel£. 56 Langxian preceded it with a Song dynasty account focused on her shaman father, describing official attempts to stop the inhabitants of Wu from "riding the waves" to welcome the spirit of the wronged hero Wu Zixu. 57 And the supernatural conclusion has Zhou Di saved by arguably the most prominent of China's long line of political martyrs. Icons of literati resentment thus frame the story. The speech that Zhou Di's wife makes to an uncomprehending crowd is the moment in the story that epitomizes the late Ming ru predicament and suggests the way Langxian proposed to channel that literati resentment. The shi, to be true to his own self-definition (the" Rites and Odes," a venerable metonym for the whole of ru tradition), had to proclaim and act out his fidelity to ru principle. He did so, however, in a world of partial understandings, a world in which power could be experienced as dismembering ru attributes while still demanding adherence to them. But the explicit ru claims used to undergird that power were the same that empowered the shi himself, and so our story cannot name a villain without undermining the very foundation on which the shi stood. The shi had to find some avenue of transcendence. "Siege" accomplishes this by going "The Martyr" one better: Shentu Xiguang simply avoided addressing the public; Zhou Di's wife does so but cannot be understood. Her words, at this point in the story, matter to no one but herself. Is this solipsism? No, because "pragmatically," Langxian's is such a masterful performance. On the one hand, he created a powerful and self-consistent fiction. But on the other, the fact that his source was so well known ensured that he would draw attention to himself as performing this transformation from Qing shi entry to vernacular huaben. What he performed at this point in the story is, paradoxically, the pure inwardness of a spirit from which the possibility of impressing others has been stripped away. He was signaling, to a ru public of like mind, ru integrity, integrity completely divorced from any question of benefit. His glory lay in the impossibility of his miSSIOn.

Opposition and Affirmation In other huaben as well, Langxian problematized ideals and conventions and refused to provide the solutions that an easy reading of tradition might lead us to expect. Shi reclusion is not invariably held up as the right way to deal with a troubled world: his recluses in "Lu Nan" and the

Katherine Carlitz

233

prologue to "The Old Retainer" are troubled and intemperate themselves, even as they epitomize ru connoisseurship and culture. The father of Wang Benli, Langxian's most famous filial son, tries to retire to a monastery but is forced to return to the world. The magistrate of "Metamorphosis into a Fish" becomes an immortal, in a standard "deliverance" conclusionbut here, too, the conventional conclusion is inadequate to the horrors of his nightmarish state, in which he is nearly eaten by his colleagues. 5 8 The fact that Langxian repeatedly juxtaposed these conventional ideals to his mimetically dense and disturbing stories suggests to me that his point is the inadequacy of conventional rewards, and that in itself is a way of demanding the style of integrity evinced in our two stories above. This shows us that we have to beware of reading into these stories a specifically modern bleakness, a questioning of the idea of value itself. The prime Ming virtue of filial piety suffers a terrible beating in Langxian's stories, as his filial children suffer for the sake of their undeserving parents. Wang Benli goes on an arduous quest for a father not at all anxious to return home. The Constant Words heroine "Rainbow" Cai must submit to repeated rape in her struggle to avenge her drunken parents. 59 Since filial piety was explicitly understood in ru tradition as an analogue to and training for loyalty, there is an unmistakable suggestion here that the government, the "parent of the people," is similarly irresponsible. Langxian's filial tragedies can thus be read as dramatizing his lack of confidence in public morality. But the insistent coding of the filial children and faithful wives as paragons sets up an alternative icon by valorizing their suffering, and even, as I suggest above, "purifying" it. What may look to us like a devaluation of that suffering-the fact that it changes nothing, that it does not "work"-may represent the absolute authentication of that suffering in Langxian's eyes. The suffering is untainted by ambition, as it were. This may be an unfamiliar gesture to us, but it was a familiar gesture in the late Ming, as we can see from statements in which literati identified themselves with the suffering transhistorical Sage, or offered themselves for martyrdom, or indeed gave heartfelt adherence to the Ming cult of martyrdom generally. The venerated teacher and essayist Gui Youguang (1507-7I) wrote of the junzi or Sage, in language that prefigures Langxian's heroines, as "contending uneasily against calamity with his weak and tender body," and standing "isolated and defenseless" in the midst of "stratagems and deceptions. " 60 And two decades after Langxian published "Siege" and "The Martyr," the poet Huang Chunyue committed a loyalist's suicide that fused the language of self-sacrifice with the language of passion, as he stated that only the "indestructible heart" of this "orphaned

2 34

TWO STORIES BY "LANGXIAN"

servant of the Ming" survivedY The evocation of heroic suffering could in itselfbe a satisfying way to end a tale of trouble, injustice, or mismanagement, since in late Ming literati culture, self-identification with that suffering was in itself a means of self-valorization. (Gui Youguang, writing of a martyred young woman in the I 54os, exclaimed that a lifetime of perusing history has turned up no one else who so "made his feelings manifest." }62 This sort of self-valorization has direct consequences for the way we read the conclusion of "The Siege of Yangzhou," in that the inadequacy of Zong shi's reward is a way of making her suffering seem all the more poignant and admirable, all the more attractive as a representation of the shi. To return to the terms used earlier in this chapter, such a reading of the way these two stories conclude understands them as gestures of opposition, as dramatizing the situation of the shi assaulted by the slings and arrows of a difficult world. But what of the affirmative aspects of late Ming literati culture? Langxian was, after all, writing in one of the literary and luxury capitals of the late Ming world. How do his stories articulate with that context? We have already seen that the book trade itself offered a wealth of new venues for publication, and that the connoisseurship evident in the collection of songs and huaben brought together a wealth of narrative options for Langxian to evaluate. Langxian's valorization of pure suffering shows him to be as much a man of qing as the hero of that connoisseurship, Feng Menglong, and the way he distanced his heroines from the common plane in itself evinces the connoisseur's way of thinking. Hanan observes that Langxian "plunges us into the world of the Confucian exemplum," but we can feel fairly sure that these particular exempla came to him not from Lienii zhuan but from Feng's racy and transgressive Qing shi. And might not even the courtesan ideal vaunted there (Feng was a close friend to the courtesans of the city} have subtly contributed to his characterization of Shentu Xiguang? That courtesan ideal in its Ming incarnation flowered along with the Ming ideal of the highly educated and companionate gentry wife. Both, as Dorothy Ko has pointed out, were trained to meet the needs of governing-class men, and while their training was not identical, still "the same ultimate goal, the cultivation of talented beauties, lay at the heart of both kinds of education. " 63 At the very least, the icon of the courtesan was so prominent that she cannot be ignored in considering the late Ming literati construction of ideal womanhood. (Young faithful courtesans figured in a number of tragic stories of

Katherine Carlitz

2

35

their own, emblematizing the dream of literati freedom and the libidinal satisfaction of that dream in tragedy if not in triumph.) 64 Nor, as noted above, can Langxian's narrative method itself be dissociated from the world around him. The rhetorical instrument developed by the sixteenth-century novelists mirrored the expansive Ming environment in scope, but deflated its pretensions: even the attractive, exuberant violence of Shui hu zhuan is questioned, muted, and finally dissipated to a certain extent in the fullest literati recensions of the novel. 65 We have seen Langxian similarly rein in the violence of his Qing shi source and deflate the grandiose claims of a story like "Apotheosis." But the quietly grotesque situations that he created in their place mirror the exuberant Ming in a more subtle way. His insistent naturalization valorized the particular, as, increasingly, did the world around him. To be master of the telling detail is to be a representative of a ramified and diversified world. The exquisite placement of detail so as to afford maximum incongruity reflects a sense of the rich possibilities of that ramified world. To problematize excess is to dramatize the experience of living with excess. The ru culture that produced this pair of stories was in a continual process of mutual, dialectical self-production with the world around it, and these stories show us how analytically inseparable are the difficulties and opportunities presented by that culture and that world. Reading these stories as instances of that dialectical self-production has, I hope, enriched them rather than reduced them. These are powerful stories, and they can only gain in power when the reader knows the materials from which Langxian fashioned them, and the world to which he offered them.

CHAPTER

9

Ming-Qing Women Poets and the Notions of "Talent" and "Morality" Kang-i Sun Chang

aze_w modern readers of Chinese literature realize that the works of

J'

as many as 3,500 female poets were published during the MingQing dynasties alone. Most people are, however, familiar with the famous dispute between two eminent Qing scholars over the issue of women's position in culture-a debate in which the liberal-minded Yuan Mei (r7r697) promoted women's role as creative writers and the conservative Zhang Xuecheng (r738-r8or) championed their role as upholders of morals and ritual. Modern opinion, as David Nivison says, "inevitably sides with Yuan," because "his encouraging women to write and publish wins our sympathy at once." 1 Conversely, Zhang Xuecheng's outmoded proposal "for the cloistering of women in ritually correct relationships," says Susan Mann, does not seem convincing. 2 But certainly the core issue of the Yuan-Zhang debate-one that centers on the. basic dispute between cai (talent) and de (virtue)-reflects a longstanding issue in the history of Chinese literature. Since ancient China, the cai!de discussion has not inherently been a gendered issue, because it applies to both men and women. 3 However, gender puts a particular spin on the debate. The Ming and Qing periods, when an unprecedented number of women poets appeared on the scene, were bound to see the opposition of cai and de in a new light. In fact, two hundred years before Yuan and Zhang's time, the cai/de issue had already become the main controversy in female education and divided late Ming critics 4-although the debate then never became as personal and intense as did that following the publication of Zhang Xuecheng's "Fuxue" (Women's learning) in I797· Zhang's immediate target was Yuan Mei and Yuan's circle of female disciples, whose

Kang-i Sun Chang

237

anthology of poems, Suiyuan nudizi shi xuan, was published a few months before Zhang's "Fuxue." 5 In "Fuxue," Zhang alluded to Yuan Mei (though not by name) three times as "a literatus of misconduct" (wuxing wenren), who was responsible for leading women poets to ignore morality, ritual, and women's heritage of classicallearning. 6 In Zhang's view, it was such "literati of misconduct" who led the so-called talented women (cai nu) to depart from the ancient model of the "quiet woman" (jing nu) celebrated in the classical canon Shijing (Classic of poetry). Zhang explained that he did not categorically condemn all women's poetry, since he did approve of the Six Dynasties women poets, who, in his view, "not only knew how to compose five- and seven-character line poems, but also took pride in their 'three obediences and four virtues' [san cong side]." And he even admired the courtesan-poets of the Tang, who "not only excelled in poetrywriting but were also well-versed in ancient classics," so that their love poems could easily be taken as allegories for their own virtue of "meekness and gentleness." But the problem with contemporary "talented women," Zhang argued, was that they lacked true learning-the kind of learning (xue) that would have led them to the true understanding of moral principles. The "talented women" of the present, added Zhang, were eager to "show off their talent," not knowing that the wuxingwenren who promoted them were actually "admiring their beauty" (!ian qi se) rather than their talent. Zhang's severe criticism may have been directed against the general "moral corruption" of the contemporary world of letters, male and female alike.? But there is no doubt that Yuan Mei and his female disciples (totaling 28) were outraged by Zhang's accusations. The widow Luo Qilan, foremost "feminist" among Yuan Mei's disciples, apparently publishedher new anthology of poetry (Ting qiu guan gui zhang tong ren ji, dated 1797) in rebuttal. In the preface to this anthology, Luo Qilan argued that women who distinguish themselves as poets are in fact superior to men, for they need yet to conquer the various restrictions imposed upon them. In other words, Luo Qilan did not believe that women poets should function in a separate domestic sphere. Eventually it was the success of women as published authors that counted against Zhang Xuecheng's criticisms. To the women disciples of Yuan Mei, their anthologies of poems (as well as their separate collections) became the strongest argument for what they believed to be the true voice in poetry, the kind of xingling (spontaneous self-expression) that Yuan Mei claimed to be the primary consideration in poetry writing. As Arthur Waley said, it was the idea of xingling that led Yuan Mei to claim that

MING-QING WOMEN POETS

poetry "has domains of its own and need not necessarily be a vehicle for moral edification. " 8 In Yuan Mei's view, not only should poetry transcend morality, but it should also transcend learning 9-a notion that Zhang would certainly object to. In his defense of women's right to create poetry, Yuan Mei reminded the public that women were important contributors to the ancient canon, the Shijing: "As for the three hundred odes' 'Ge Tan' and 'Juan er,' which of them was not written by women? The narrow-minded view of those pedants is wrong indeed. " 10 Of course, Yuan Mei was not the first person to point out that the authors in many of the poems in the Shijing were women-for at least r 50 years before Yuan Mei's time, scholars had used this claim as one of the main strategic constructs by which they tried to elevate and legitimatize the literary status of women's poetry. 11 Zhang Xuecheng, for one, would not agree to this claim. 12 But Yuan Mei went beyond his predecessors in at least one respect. For him, it was not the fact of authorship that counted so much as the presence, in the Classic of Poetry, of an authentic female voice. Yuan Mei's defense of women apparently made many contemporary female poets more self-conscious in their efforts to define themselves and their own place in literature and society. What made Yuan Mei's approach particularly attractive to many Qing women poets was his unqualified respect for "talent. " 13 Yuan claimed that talent is of paramount importance in poetry writing, because the power of talent far exceeds that of learning (xue) and knowledge (shi). In particular, he believed that a woman of poetic talent, regardless of social status, 14 should be encouraged to write and publish so that her natural abilities would be developed to the fullest. To a certain extent, in promoting women poets, Yuan was merely continuing the liberal trends of the late Ming, as I shall explain more fully below. Zhang's attack on Yuan as leader and inciter of women poets was based on a misapprehension that a survey of the relevant bibliography will correct. In fact, women poets had already been engaged in compiling anthologies and publishing their own collections of poems for more than 200 years. What Zhang found most offensive was precisely these women's growing desire for recognition through publication. In his essay entitled "Shi hua" (Poetry talks), Zhang especially singled out this aspect of women's activity for criticism. In his view, women's publishing their own works (along with the congratulatory commentaries and notes they wrote for one another) was a means of showing off or "fishing for fame" (si li mingzi). 15 In fact, even for men, "showing off is considered a major flaw in a gentleman's character" (lu cai shi shi junzi dabing) according to the

Kang-i Sun Chang

239

orthodox Confucianists, or so said the Ming moralist Lii Kun ( r 53 6r6r8).16 Thus, Zhang Xuecheng harshly criticized those women poets who were, in his view, "fishing for fame," for women since ancient times were expected to remember that "words were not to pass from the women's quarters to the outside world" (nei yan buchu kun wai). 17 All this explains why Zhang was so deeply troubled by the appearance of Yuan Mei's anthology of poems by his 28 female disciples. Perhaps it was in this context that Zhang felt it necessary to redefine the (by then old-fashioned) saying, "A woman without talent is a virtuous woman." This phrase first became popular in the late Ming as part of the saying, "A man with virtue is a man of talent. A woman without talent is a women of virtue" (nanzi you de bian shi cai, niizi wu cai bian shi de). 18 The saying, of course, belongs to a long tradition of gender inequality, one result of which was to discourage women from demonstrating their talents. But Zhang Xuecheng, in his rereading of the whole idea, explained that those who invented the saying did not really abhor female "talent" itself, but simply wished to caution those women who were eager to show off their talent against making fools of themselves (ZXCY, p. 48 ). 19 Thus, Zhang says, the phrase should be construed as "A woman who does not show off her talent is a virtuous woman." It may come as a surprise to the modern reader that perhaps more literate women than men during the Qing would have agreed with Zhang Xuecheng. A case in point is the gentrywoman Wang Zhenyi who, in her letter to a certain Lady Bai, indignantly condemned contemporary women for being unduly anxious to publish their works as an easy way of fishing for fame. 20 This example reveals that under the influence of orthodox Confucianism some Qing women indeed believed that publishing one's own poems was a transgressive act. Here we can see some interesting parallels between some literate Qing women and the eighteenth-century English women, for whom (as Vivien Jones says) women's publishing was "figured at its most extreme as a loss of chastity," for it "exposed an essentially private activity to the public gaze, blurring the ... delineation of separate spheres." 21 Like Samuel Richardson's Pamela, who says that "to rob a person of her virtue is worse than cutting her throat, " 22 many Qing women felt compelled to protect their moral reputation as virtuous wives and mothers. 23 Behind this attitude of these Qing women lay an increasingly popular notion on the part of some Confucian moralists-that is, the idea that "talent itself can indeed hinder one's morality" (cai ke fang de). 24 The concept may be traced back to the Song Neo-Confucianists' notion that

MING-QING WOMEN POETS

men's natural talent can lead to vice (as well as goodness)-which was originally meant to be a critical revision of Mencius' idea that talent is beyond good and eviJ.25 Under the influence of this Nco-Confucian idea, it is not surprising that many Song women began to feel that literary talent for women could be harmful. A case in point is a certain Lady Sun who showed unusual potential for poetry writing but refused to take lessons from the then-famous poet Li Qingzhao, even after the poet herself offered to tutor her. 26 Lady Sun was only in her teens when she said to Li Qingzhao, "Literary embellishment is not a woman's affair." 27 The view was shared by many gentrywomen during the Qing. For example, Wang Zhenyi claimed that "cultivating one's talent is not a woman's primary duty, for it can hurt her morality. " 28 Thus, faced with the "danger" of publishing, many literate Qing women simply burned, or attempted to burn, their poems. 29 One of the most notable examples was Ye Baolin (1609-76, wife of the famous Huang Zongxi, r6r0-95), who burned all her poems in protest after she heard that many local gentrywomen, "in a breach of morality," had begun to organize poetry clubs in which they exchanged poems with men while drinking. 30 Another example was Zhong Yun, who burned all her poems before her death; nonetheless her son Zha Shenxing (r6 50-1727) managed to publish a collection of her "poems" based on his own memory of them. 31 In a similar situation, Yu Xiusun started to burn her poems a few days before her death at the young age of 3 3. But her father, Yu Quyuan (I 82rr9o6), who was an exceptionally active promoter and sponsor of women poets, managed to publish those of her poems saved from burning. 32 All these examples show that the lives of a substantial portion of Qing women were in fact the ultimate mirror for Zhang Xuecheng's theory concerning women's decorum and morality. As Nivison observes, the unusual success of Zhang's "Fuxue" at the time (in sharp contrast to the rest of Zhang's work) is evidence at least that Zhang's "prejudices were widely shared. " 33 In fact, as the modern historian Qian Mu points out, deeply entrenched in Chinese history is a long-standing tradition of "prizing morality over talent. " 34 For example, in judging the relative importance of de versus cai, Confucius said: "A man of virtue is sure to be a man of words, but a man of words is not necessarily virtuous. " 35 Later, the Song scholar Sima Guang (ror9-86) further dramatized the cailde contrast by claiming that "those whose virtues are greater than their talents are called 'gentlemen' [junzi], whereas those whose talents are greater than their virtues are called 'small men' [xiaoren]." 36

Kang-i Sun Chang

All these examples show what I would call a "cultural priority" for a Confucian moralist like Zhang Xuechang. They explain why Zhang was so deeply troubled by Yuan Mei's unqualified respect for talent. The issue was essentially bound up with traditional male concerns such as the selection of officials and the role of moral teaching in a rapidly changing society. To endorse "talent" over "morality" was to embrace the changes brought by widespread literacy, printing, urbanization, and the growth of commerce; to cling to "morality" was to resist those new developments. Yuan Mei was, of course, not alone in defending the supreme significance of talent, in reacting to the common biases on the part of orthodox Confucian moralists. Dai Zhen (I724-77), for one, claimed that "because man's nature is good, his talent is also good" (xing shan ze cai yi mei)-in his famous attempt to revive the original philosophy of MenciusY Significantly, it was the Qianlong emperor (Hong Li; r. I736-95) who, in his "Cai de shuo" (Essay on cai and de), proclaimed the equal importance of talent and morality-a gesture that had the effect of promoting the long subordinate notion of talent. The famous official Zhu Gui ( I73 I-I 8o6) summed up Emperor Qianlong's main purpose in writing his essay: "[Thanks to this essay, the emperor's subjects] would no longer dare to deceive themselves and others by being 'talented without being virtuous.' Nor can they any longer feel self-contented in their own uselessness by being 'virtuous without being talented."' 38 These examples make it clear that the talent/morality debate was not exclusively, or even primarily, a gendered one for Yuan's and Zhang's contemporaries. However, I believe it is important to examine how Qing women, and women before them, helped to produce so fierce and divisive a controversy as the debate between Yuan Mei and Zhang Xuecheng. The issue of cai versus de is an essential subject to consider if we want to explain how women themselves identified and solved the paradox that shaped their literary culture. But the cai!de issue has a long history linked to many other issues and attitudes; only when it is viewed in a historical context can it gain in clarity. To describe the history of the cai!de issue regarding women, we must start with Ban Zhao's (ca. 49-ca. 120) "Nii jie" (Precepts for women). 39 When Zhang Xuecheng began his essay "Fuxue" with the four classical female ideals of "virtue, words, demeanor, and meritorious deeds" (de yan rong gong), he was probably thinking of Ban Zhao's "Nii jie" in a section called "Fuxing" (Women's conduct). For by reminding women of their original mission as upholders of morality, Zhang was actually ac-

242

MING-QING WOMEN POETS

knowledging women's tremendous power of moral suasion, of which the transmitted text of Ban Zhao's "Ni.i jie" is a symbol. How did women like Ban Zhao, as writers of moral texts, acquire the power to ensure the immortality of their words? Interestingly, in the case of Ban Zhao, the tactic she adopted would seem at first to be one of insisting on woman's powerlessness-for she began the preface to "Ni.i jie" with the words, "I am stupid and ignorant, and the character with which I have been endowed is not clever. " 40 Humble words; but it is precisely this tone of genuine humility that gives Ban Zhao's essay great moral suasion. For despite her learning-she is, of course, known as the eminent scholar-historian who completed the History of the Han after the death of her brother Ban Gu-Ban Zhao possessed the very virtue of modesty that would have graced a Confucian gentleman. Who more than the learned and virtuous Ban Zhao is better qualified to give women advice? Indeed, the moral power associated with proper feminine language is celebrated in many historical anecdotes 41 -although it is said that Ban Zhao's sister-in-law, the liberal-minded Cao Fengseng, wrote an essay to challenge Ban's moralistic viewsY In any case, it was Ban Zhao's "Ni.i jie," rather than Cao Fengseng's critique (which has not been preserved in any form), that has won immortality in Chinese literature. Significantly, Ban Zhao makes very definite claims about the importance of "Ni.i jie" by emphasizing the urgency of her concern that women be trained in proper female conduct: But I am pained that my daughters are of marriageable age, yet they do not steep themselves in instruction-they do not hear the wifely rituals. I fear that they will lose face in another household and shame their ancestors. Thus in my spare time, I have written these "Precepts for Women" in seven chapters. I suggest that you each write them out completely, so that you will gain great advantage, which will assist you in your life. 43

Ban Zhao has faith in the power that comes from adherence to ethical codes. Clearly it was the same sense of moral authority that prompted Song Ruohua and Song Ruozhao (fl. ca. 790) later to say in the preface to their "Female Analects": "If you follow these words, you will be a worthy wife whose virtue surpasses those of ancient people, and your name will stand out in history. " 44 Works like the "Ni.i jie" spoke not only of rules but of rewards; and doubtless it was because they could promise rewards that such writings could triumph, largely through women's own consent and promotion. Ban Zhao's "Precepts" are not merely an ethics of the female sphere. The names of the four female virtues listed in "Ni.i jie"-de, yan, rang,

Kang-i Sun Chang

243

gong-immediately remind us of the section on women in the ancient Zhou li. 45 But Ban Zhao's interpretation of these virtues is reminiscent more of the Zuo zhuan, another ancient canonical text. According to the Zuo zhuan, a man could become immortal in one of three forms: "virtue" (de), "meritorious deeds" (gong}, or "words" (yan). This enumeration of the "three ways to immortality" (san bu xiu) had, by the Han, become a motto for the entire scholar-official class. Thus, in a way, Ban Zhao attempted to link moral suasion to the ethical significance of the male san bu xiu idea, even though her idea of gong refers more to women's needlework than to their social and political deeds. 46 But what is important is that, like traditional male scholars who valued de much more highly than either yan or gong, 47 Ban Zhao also singled out de as the most important form of immortality. In her poem "Dong zheng fu," written around A.D. 9 5, she contemplated the affective potential of de while following her son "on an eastward journey":

ITft 11 j,'il.\ ~ ::f t'J 15~ret~£ ffiJ ;3 f¥

't1U!!!!I! ZJ5JT ~ 15-

Only good virtue cannot perish, The body dies but reputation survives. This is what the classics and canons have extolled; They honor the Way and virtue, benevolence and wisdom. 48

Thus later, in "Nii jie," Ban Zhao was careful to spell out the attributes of what she meant by "virtue," to ensure that her daughters would carry out their moral responsibilities properly. In defining "virtue" for her daughters, Ban Zhao was apparently making a choice between de and cai: to her the female de means that women should first of all be virtuous, but "need not be particularly talented or brilliant. " 49 Ban Zhao's emphasis on virtue, however, represents only the general trends of the time. For it was not until the third century A.D., at the end of the Han, that the concept of talent became a main concern in Chinese literature-partly due to the influence of the Cao family members who promoted creative writing and made the Jian'an period (r96-2r9) the first period in China known for the excellence and abundance of literary talents. Serving as the "helmsman" of the Jian'an literati, Cao Cao (rs 5220) "judged people by their ability, regardless of morality. " 50 And his son Cao Pi (187-226), in his famous essay "On Literature," confidently proclaimed literature "a way to immortality"-thus giving literary talent an unprecedentedly exalted status. 51 A famous anecdote concerning Xie Daoyun (fl. fourth c.) illustrates the new way of literary evaluation, by which poets were judged two hundred years after Ban Zhao's day, at the

244

MING-QING WOMEN POETS

beginning of the Six Dynasties: "Once when they were assembled inside, a violent snowstorm suddenly arose. Xie An said, 'What does the snow resemble?' His nephew Lang said, 'It can best be compared to scattered salt in the empty sky.' But Daoyun said, 'Is it not like willow floss rising in the wind?' And so An was greatly delighted." 52 Xie An (320-85), the famous scholar-official and general, "was greatly delighted" because of his niece Xie Daoyun's extraordinary poetic talents-which indeed surpassed those of her male cousin Lang. Significantly, in evaluating Daoyun against Lang, Xie An relied on aesthetic criteria rather than moral considerations-thus leading him to assign value to the mere poetic power of Daoyun's line. In other words, Xie An was proud to have such a gifted woman poet in his already prominent Xie clan, whose combined talents dominated the literary scene of the time. Xie Daoyun went on to distinguish herself as a writer. 5 3 Indeed, Xie Daoyun's case is, while exceptional, not isolated. There emerged a widespread interest in and respect for women's poetic talents during the Six Dynasties. People began to adopt the principle of pin (evaluation and ranking) in judging and comparing female poets as they would male poets. Thus when Emperor Xiaowu (r. 454-64) of the Liu-Song asked the poet Bao Zhao to say something about his famous sister Bao Linghui, Bao Zhao responded in typical Six Dynasties fashion-by comparing his sister with another woman writer Zuo Fen, the sister of the poet Zuo Si: "My younger sister's talents [cai] are of course not as great as Zuo Fen's, just as my talents [cai] are inferior to Zuo Si's. " 54 Similarly, Xie Daoyun was compared to another of her female contemporaries, Madame Gu: Xie Xuan held his elder sister, Xie Daoyun, in very high regard, whereas Zhang Xuan constantly sang the praises of his younger sister and wanted to match her against the other. A certain nun named Ji went to visit both the Zhang and the Xie families. When people asked her which was superior and which inferior, she replied, "Lady Wang's [Daoyun's] spirit and emotions are untrammeled and bright; for this reason, she has the style and air of the Bamboo Grove literati. Madame Gu has a pure heart that shines like jade, and this is the elegance of the inner chambers. " 55

Interestingly, this last statement, contrasting the distinct styles of women poets, touches on an important distinction between the new emphasis on "talent" and the conventional "moral" considerations. Whereas the more individualistic and "untrammeled" Xie Daoyun is compared to "the Bamboo Grove Literati" (linxia (eng qi), Madame Gu is said to have "a pure heart that shines like jade" (qing xin yu ying) because of her role as a

Kang-i Sun Chang

2 45

virtuous and dutiful wife. That Xie Daoyun is being treated as a talented equal of the most famous contemporary male poets, known collectively as the "Seven Sages of the Bamboo Grove," is most interesting. 56 The Bamboo Grove literati were known of course for their lifestyle of free creativity and nonconformity. And in many ways, Xie Daoyun resembled them, for she was known as a woman of unrestrained spirit, capable of engaging in philosophical "pure talk." 57 Most important, she was a woman of independent judgment. It is said that once Xie Daoyun even passed judgment upon her own husband, Wang Ningzhi, suggesting that he was inferior to her male relatives-hardly the act of a submissive and "virtuous" wife. 58 It is obvious that her self-conscious judgment and independence came largely from pride in her own talents. Later during the Tang, the concept of female talent became even more common. The Quan Tang shi records works by more than 100 female poets of the Tang, totaling roughly 6oo poems. One immediately striking feature of this poetry is its variety: it includes works of court poets, palace ladies, gentrywomen, concubines, abandoned wives, courtesans, and Daoist nuns. 59 It is interesting to observe that those Tang women who managed to achieve the status of major poets and attain enormous success as published authors were mostly courtesans and Daoist nuns. This is not surprising if we consider the fact that the lifestyles, or the professional requirements, of these women made it easier for them to break away from conventional limits. Indeed, there were few gender restrictions on them, for they mingled freely with the contemporary male poets and scholar-officials. It was in this special context that the courtesan Xue Tao (ca. 768-ca. 83 1) and the Daoist nuns Li Ye (8th c.) and Yu Xuanji (84 s-68) 60 became the most famous women poets of the Tang-certainly the only female writers at the time who issued their own collected works. These women poets socialized and exchanged poems regularly with famous male literati-for example, Xue Tao with Yuan Zhen (779-83 1), Yu Xuanji with Wen Tingyun (ca. 812-70)-and were praised generally for their literary talents. In particular, the male poet Wang Jian (ca. 75 1-ca. 830) lauded Xue Tao as a "talented scholar with painted eyebrows" (saomei caizi), and the Tang emperor Xuanzong (r. 713-56) summoned Li Ye to the court when "he heard about Li Ye's extraordinary talents" (wen qi cai). 61 Most important, these women also began to manipulate a special image of themselves as the talented equals of male poets-an equality symbolized by their reciprocal exchanges of poems. But because the recognition of equality based on talent was not borne out by equality of opportunity-

MING-QING WOMEN POETS

for women were simply not allowed to take the imperial examinationsYu Xuanji complained: How I hate this silk dress That conceals a poet! I lift my head and read the promotions list In powerless envy.6 2

Yu Xuanji's ambition to be men's equal was praised by Xin Wenfang (in his Biographies of Talents During the Tang) as a sign of true talent in a woman. 63 And it was respect for talent that caused many men to sympathize with the tragic death of Yu Xuanji. 64 Although many male literati admired the way courtesan poets and Daoist nuns broke away from the traditional "four virtues," 65 there were many others who condemned these women for their "immoral conduct. " 66 Indeed, there were so many moralists who insisted upon the bad consequences of transgressing the rules of the "four virtues" that some people began to equate talent with immoral conduct. Legend has it that when Li Ye was still a young child of five or six, her father had already noticed how poetically talented she was, and so he commented: "but she will grow up to be a bad woman! " 67 Similarly, Xue Tao's father was "sorrowful for a long time" after seeing Xue Tao's "early display of talent in poetic composition. " 68 In fact, some courtesans were ashamed of being "talented women" themselves-for example, Xu Yueying lamented in a poem that she did not have a chance to fulfill women's virtue of the "three obediences" (san cong). 69 Xu Yueying's self-pity is of course understandable, for the price of being a courtesan was high. It often meant a lifetime of loneliness and drifting, as suggested by the image (or self-image) of the flying willow catkins courtesans were so fond of invoking in their own poetry. Indeed, the Six Dynasties female talent Xie Daoyun seems to have inadvertently forecast the fate of the Tang courtesans when she first conceived of her famous line: "Is it not like willow floss rising in the wind?" Compared with courtesans, Tang and Song gentrywomen felt more secure in their domestic world, but at the price of accepting certain limitations on women's poetry writing. In sharp contrast to courtesans like Xue Tao and Xu Yueying who were full-time poets and singers, most gentrywomen had to write in intervals snatched from domestic responsibilitieschief among them the womanly duty of weaving. Since traditional conduct books gave primary importance to weaving-an activity Ban Zhao defined in her "Nii jie" as the way to fulfilling one's "meritorious deeds" (gong)many women simply accepted it as their principal duty in a properly female

Kang-i Sun Chang

247

sphere of activity. To them weaving was "meritorious," whereas poetry writing was insignificant, if not a direct peril to one's virtue. This highly popular view was no doubt instrumental in defining an ideological identity for the Tang and Song gentrywomen. But the poet Zhu Shuzhen (ca. 1079-ca. 1131) apparently did not agree with this common view and wrote in complaint that, in the eyes of her contemporaries, "It's not my business to wear through an iron ink slab;/Only if I embroider until my golden needle snaps do I gain merit. " 70 Zhu's complaint is reminiscent of a similar comment made by the American woman poet Anne Bradstreet (ca. 1612-72): "I am obnoxious to each carping tongue/Who says my hand a needle better fits? " 71 Like Bradstreet, Zhu Shuzhen insisted on her right to compose poetry, and she eventually became the most prolific female poet of the Song. Zhu Shuzhen was born and brought up in an extremely literate family. But she was married to a husband who did not share her literary interests. Worse still, after her death, her parents burned most of her poems-perhaps afraid that her love poems would provide evidence for allegations of an extramarital affair. 72 But half a century later the male scholar Wei Zhonggong gathered all of Zhu Shuzhen's surviving poems and made Zhu's Collected Works, along with commentaries by Zheng Yuanzuo, available to the general public. 73 Thus, finally, Zhu Shuzhen became known as a major poet. Compared with Zhu Shuzhen, another famous woman poet, Li Qingzhao (1084-ca. I I 51), was extremely fortunate to have male relatives who supported her literary endeavors. Li Qingzhao and her husband, Zhao Mingcheng, were the ideal match of the "talented husband and gifted wife" for later generations. But among contemporary gentrywomen, Li Qingzhao had no peer. Indeed, she never considered herself a "woman" poet and never felt pressured to prove a relationship between morality and the legitimacy of women's poetry writing. Rather, she saw her proper place among the other male literati and felt confident enough to evaluate the works of male poets-as she did in her "Ci lun" (Essay on ci poetry). In her unique style, Li Qingzhao seems to remind us of the "untrammeled and bright" Xie Daoyun, who was capable of critiquing men and, thus, of seeing herself as a talented equal in the literary tradition. But Li Qingzhao was truly exceptional among the gentrywomen in her time. It was not until the Ming that the ideal "talented woman" (cai nii) emerged as a popular central figure, when she became a cultural signpost for many of the ideas developed during that period. In the beginning, the Ming courtesan was the prototype of this "talented woman," a reincarnation of the Tang courtesan indulging in the reciprocal exchanges of poems

MING-QING WOMEN POETS

with male literati-though with a special emphasis on the new "cult of love" current at the time/4 First, we see in the early Ming the talented courtesan-poet Zhang Hongqiao (fl. 14th c.) and her lover Lin Hongthe equally gifted male poet known as one of the "ten talented scholars of Fujian"-swearing undying love to each other through numerous exchanges of poems, which are remembered even to this day. 75 Then there was in the mid-Ming the distinguished courtesan Wulingchun (Qi Huizhen), whose devotion to her lover until death was praised by the male scholar Xu Lin (1462-I538).?6 Finally, in the late Ming, there emerged an unprecedentedly great number of talented courtesans who were published authors, poets, painters, calligraphers, or musicians-and who were able to form a relationship of true compatibility and mutual respect with contemporary male literati. These were also courtesans who managed to get married to important poets and scholars-indeed a striking contrast to the Tang courtesans who often complained in their poems that their male friends, mostly famous poets and high officials, would never marry them.7 7 Thus, although the Tang courtesan Yu Xuanji lamented that "it is easier to get priceless jewelsffhan to find a lover with a true heart" (QTS, r r: 9047), many late Ming courtesans did succeed in finding their ideal man and husband. These lucky courtesans were no doubt a real-life reflection of the heroines in the then popular caizi jiaren (talented man and gifted beauty) fiction and drama, who not only possess talent but are inevitably led to a talented man. 78 Moreover, the idea of "lovers coming in pairs" promoted by the caizi jiaren fiction 79 has a special relevance for the matching "literatus-courtesan couples"-couples such as Hou Fangyu (r6r8-ss) and Li Xiangjun, Mao Xiang (r6rr-93) and Dong Bai, Gong Dingzi (r6rr-73) and Gu Mei, Yang Wencong (1596-r646) and MaJiao, Ge Zhengqi and Li Yin, Qian Qianyi and Liu Shi (r6r8-64). Just like the talented man and the gifted beauty in the caizi jiaren fiction, these late Ming "literatus-courtesan" couples had a romantic idea that not only were they destined to marry each other but their life together would surely constitute an exemplary "wonderful tale that will last for a thousand autumns. " 80 Indeed, their "wonderful tales" have continued to fascinate Chinese poets, critics, and readers, who repeatedly return to the details of these love stories. Some later people even commented on the compatibility of certain "literatus-courtesan" pairs of the late Ming-for example, Xi Peilan, the famous disciple of Yuan Mei, regretted that Liu Shi did not get to marry the late Ming loyalist poet Chen Zilong (r6o8-47), for in her view the Liu-Chen pair would have been more perfect than the LiuQian union. 81

Kang-i Sun Chang

249

A more important question is: Why was the late Ming courtesan so insistently a figure of fascination in critical texts of the period? First, the late Ming was the period when people developed a romantic idealization of the concept of cai itself, and it is only natural that courtesans specializing in poetry, painting, and dramatic arts would at once become the object of adoration. In turn, many courtesans were inspired, and encouraged, to develop their own poetic styles and to become published authors. Their works were generally included in contemporary anthologies, and the most successful courtesan poets among them-like Wang Wei (ca. I 6oo--ca. I 64 7) and Liu Shi-were frequently hailed as the greatest women poets of their times. In particular, Wang Wei was singled out for praise by the famous scholar Zhong Xing (I574-I624), who said that Wang was a most unique and talented poet, whose accomplishment was as great as those of Li Qingzhao and Zhu Shuzhen. 82 Another scholar, Chen Jiru (I558-I639), even said that Wang Wei put men to shame. 83 Similarly, later during the I65os, Zou Siyi ranked Liu Shi first among the contemporary women poets and called her "a doctor in poetry" (shi boshi). 84 Indeed the late Ming and early Qing literati so respected contemporary courtesans that they began to rehabilitate the "infamous" Tang courtesan Yu Xuanji by elevating her to the position of "sage poet among talented females" (cai yuan zhang shi sheng).SS In promoting their own position in literature, however, the late Ming courtesan-poets emphasized their moral qualities rather than their talenta strategy that was not used by the Tang courtesans. Perhaps it was because of their ultimate confidence in their own talent that these late Ming courtesans knew how to claim the opposite, and more difficult, power of de. Liu Shi, one of the most famous courtesan-poets, who lent considerable stature to women of this period, certainly understood the power of de. In her anthology of women poets, juan 4 of the "Run ji" in Qian Qianyi's Lei chao shi ji, Liu Shi evinced a special concern with the relationship between talent and virtue and evaluated poets accordingly. For example, she praises the two late Ming courtesan-poets Wang Wei and Yang Wan both for their unusual talent, 86 but her final verdict about the two poets was based on moral considerations. For although Wang Wei and Yang Wan were friends to each other and both managed to marry famous scholar-officials (Xu Yuching and Mao Yuanyi, respectively), they differed in terms of their moral conduct. Whereas the brave and virtuous Wang Wei was absolutely devoted to her husband, Yang Wan was flirtatious and had many extramarital involvements; indeed she was the very antithesis of Wang Wei. 87 As Liu Shi was careful to inform us, these two

MING-QING WOMEN POETS

courtesan-poets were judged differently simply because they had such widely different dispositions: "The Daoist [Wang Wei] was as bright and clean as the blue lotus blossom, gracefully rising above the mud. But Yang Wan ended up being defiled by mud and became a laughingstock to all. How sad! " 88 In her biographical notes about Wang Wei, Liu Shi further presented us with other evidence of Wang Wei's morality: Wang's role as a Ming loyalist. According to Liu Shi, Wang Wei and her husband Xu Yuching participated in long and intense resistance movements against the Manchu invaders after the fall of the Ming, vowing that they would rather die than yield. 89 In dwelling on Wang Wei's virtue as a Ming loyalist, Liu Shi hit upon the salient feature of the new image of courtesans that developed during the Ming-Qing transition. To the mid-seventeenth-century literati, a talented courtesan was no longer a femme fatale whom a moral man should avoid. In fact, many late Ming courtesans shared the loyalist concerns of their male friends (and later their husbands), and a great number of them sacrificed their lives for their country. 90 Some, like Liu Shi herself (who was known for her courage and chivalry), continued to participate in underground resistance activities several years after the fall of the Ming. In the meantime, the literate classes in China suddenly developed an unprecedented taste for stories about virtuous and upright courtesans. A case in point is the historical play Taohua shan (Peach blossom fan), in which the wise and courageous courtesan Li Xiangjun consistently resists the political pressures exerted on her and eventually persuades her lover, Hou Fangyu, to reject the patronage of the notoriously corrupt official Ruan Dacheng. Indeed, in the late Ming courtesans we see most vividly the happy union of talent and morality, which offers the greatest possible contrast to the previous stereotype of Tang courtesans, who were generally viewed as "talented women of misconduct." While the late Ming courtesans began to see themselves as virtuous women, the domestic gentrywomen were trying to create a new self-image as the "talented woman" (cai nii). The sudden increase of published authors among gentrywomen and the unprecedented proliferation of female anthologies in the early seventeenth century testify to this new trend. 91 As Ellen Widmer has informed us, there was a new combination of literary skill and domesticity on the part of seventeenth-century gentrywomen.92 These women-among them the famous Shang Jinglan and Wang Duanshu (r62r-ca. 1706)-were able to "create a lively literary atmosphere at home" while actively involved in publishing and establishing

Kang-i Sun Chang

poetry clubs, as well as other intellectual pursuits. 93 Interestingly, it was women's interest in (and need for) reading conduct books that initially promoted gentrywomen's literacy, which in turn came to play an important part in liberating women's creativity. For those who read conduct bookssuch as Ban Zhao's "Nti jie" and Lli Kun's "Gui fan"-were also, as often as not, readers of romantic fiction in which a talented man is seen as unfulfilled until an equally talented woman becomes his match. The role type of the "talented woman" made such an impression on the readers of the time that some conservative scholars such as Lti Kun feared their "pernicious influence on female readers of vernacular fiction. " 94 It was the same fear of the power of fiction that later led Zhang Xuecheng to denounce caizi jiaren literature. 95 Role types and fictional idealizations, however, do not account for the prolonged and self-conscious process whereby seventeenth-century gentrywomen became serious creative writers. Major female poets such as Lu Qingzi and Wu Xiao (mid-seventeenth c.) were simply proud of being talented and publicly acknowledged poets. For example, Lu Qingzi challenged the traditional view that "poetry was the job of male literati," claiming that poetry "was indeed women's due task." 96 And Wu Xiao had no false modesty when she said: From the time I was quite young, I have a had an inclination for poetry writing .... If you were to ask me to compose the Book of Rites or to complete the History of the Han [like Ban Zhao], I am afraid I would be unable to do so. But I do have a little expertise in matters of writing and chanting poetry. I do not necessarily pale in comparison with the ancients. 97

In the meantime, these liberal-minded women poets received constant praise from contemporary male literati. They were almost always compared to the Six Dynasties talent Xie Daoyun, who, as I said above, exhibited the style and air of the Bamboo Grove literati. For example, the famous scholar and anthologist Zou Siyi (fl. r 6 55) said that both Wu Xiao and her talented sister Wu Qi were "untrammeled and elegant," possessing the style of the Bamboo Grove. 98 Similarly, in his Fur en ji (Works on women), the poet Chen Weisong (r626-82) used the term "the Bamboo Grove style" to describe the unique brilliance of a woman writer called Yan Sucao. 99 And Zhou Ming (mid-seventeenth c.) simply named his anthology of women poets Selected Song Lyrics from the Bamboo Grove (Lin xia ci xuan, dated r67r)-an anthology intended to glorify the long-standing female voice established by Xie Daoyun, which, as You Tong (r6r8-r704) says, differs greatly from the didactic and "pedantic"

MING-QING WOMEN POETS

tradition associated with the conduct books, the Female Analects and the Female Book of Filial Piety (Hu Wenkai, Lidai funu zhuzuo kao, p. 896). Significantly, it was in this context of male support of female poets that Ye Shaoyuan (r589-I648), father of the famous female genius Ye Xiaoluan ( r 6 I 6-3 2 ), felt the need to redefine the "three ways to immortality" (san bu xiu) for women. In his preface to Wumeng tang quan ji (I636), which includes mostly works by his wife and three daughters, Ye Shaoyuan said that the three commonly recognized ways to female immortality have been virtue (de), talent (cai), and beauty (se). But he laments that although many women have been known for their virtue, few have had the opportunity to develop their natural talent or establish a name through writing-indeed quite unlike those ancient female Shijingwriters whose works were so honored by Confucius. 100 Ye's idea was widely shared by the contemporary male literati, and some of them-chief among them Shen Quan-even argued that women are naturally endowed with a talent for poetry writing, "like mountains being naturally firm and compact, and water being naturally clear and deep. " 101 Some scholars, such as Ge Zhengqi and Zhao Shijie, proposed that poetry is nothing but an expression of the female quality, a quality they called lingxiu zhiqi (the element of spiritual beauty)Y12 In the preface to his anthology of women writers, Gu jin nu shi (I 628), Zhao Shijie wrote: "Indeed, within the four seas, spiritual beauty has been known to occur in females and not in males. Why are they spoken of as having 'spiritual beauty'? In that one phrase both their writings and their persons are applauded. " 103 In passing, I should mention that this idea of women monopolizing the cosmic element of spiritual beauty later came to serve as a major theme in the famous eighteenthcentury novel, the Hong lou meng, where beautiful and talented women are praised as being endowed with this lingxiu zhiqi (which David Hawkes translates as "the pure, quintessential humor" ). 104 There is no doubt that this new view of women, when it was first developed in the early seventeenth century, did encourage many women to affirm their right to write poetry and to challenge existing social and cultural forces. First of all, they bitterly attacked the by-then-popular saying that "a woman without talent is a virtuous woman." Many gentrywomen had already begun to reject the submissive silences of domesticity by taking up poetry writing and publishing as legitimate pursuits. In actively pursuing her self-definition as a poet, for instance, Xu Yuan announced in her poetry that she would abandon her sewing for the sake of her literary interest: "silk from the loom, broken off; it should be cast aside. " 105 Similarly, the ambitious female anthologist and critic Wang

Kang-i Sun Chang

253

Duanshu blamed traditional women poets' failure on their conservative notion that "words should not be passed from the women's quarters to the outside world" (nei yan bu chu),1° 6 indeed a direct challenge to the traditional prescription of wifely virtues canonized since the time of the ancient Zhou li. Most important, many commentators (both male and female) on women writers began to develop the idea that women's literary talent, rather than hurt their morality, could in fact help nourish their virtues-an idea no doubt aimed at correcting the traditional distrust, or fear, of talent. 107 In sharp contrast to traditional gentrywomen whose activities revolved around domestic concerns, poets like Xu Yuan and Wang Duanshu had a wide circle of male acquaintances and were frequently involved in literary and social exchanges with men. Moreover, it happened that these "liberated" women also grew up in a generation when it became possible for gentrywomen, especially widows or estranged wives, to make an independent living by selling their paintings and poetry-the most notable examples being Huang Yuanjie (mid-seventeenth c.) and Wu Shan (midseventeenth c.). To many gentrywomen, professional artists such as Huang Yuanjie and Wu Shan must have served as perfect models of creative freedom, for they were free from the conventional domestic rituals and could pursue a career as though they were men. Admired for her "air and style of the Bamboo Grove," 108 Huang Yuanjie especially became a sort of mediator between gentrywomen and noted courtesans-for she befriended both kinds of people, for example, Wang Duanshu and Liu Shiand consequently bridged the social gaps between the two kinds of women. The eventual merging of the tradition of gentrywomen and that of courtesans is extremely important, as an indication that previously antithetical concepts of "talent" and "virtue" could indeed be reconciled and brought into a harmonious relationship generative of new traditions. At last, in the eighteenth century Yuan Mei's famous circle of female disciples became the spokespersons for such liberated gentrywomen poets. They were the legitimate successors to the various types of late Ming and early Qing women of talent. Under the influence (and encouragement) of people like Yuan Mei, many eighteenth-century gentrywomen, more than ever before, liked to meet outside the home and mingle socially and intellectually with male literati. To a large extent, such progressive women had taken over the cultural role of the late Ming courtesans-with the consequence that contemporary courtesans gradually became marginal figures whose literary works were largely excluded from the respectable literary world. Moreover, these gentrywomen felt morally superior to the

254

MING-QING WOMEN POETS

courtesans. Significantly, one finds in the eighteenth century an amazingly great number of the ideal matches of "talented husband and gifted wife." Among the most celebrated "gifted wives" were some of Yuan Mei's prize disciples, such as Xi Peilan, Jin Yi ( r77o--94) and Wu Qiongxian. 109 These matches of "talented husband and gifted wife" centering on the nurturing of poetic and artistic creativity are vividly reminiscent of the popular late Ming model of the literatus-courtesan pair. Indeed, as I have said elsewhere, "this was an age when gentry women had the confidence to claim both sides of the female poetic tradition for their own" 110-both to preserve the late Ming legacy and to value their own innovations. It was this confidence on the part of Yuan Mei's female disciples that triggered Zhang Xuecheng's bitter attack on the contemporary talented women and to further suggest that women return to their properly defined separate spheres. Zhang Xuecheng's attitude represents the moralistic views of many orthodox Confucianists in late eighteenth-century China, but such a view was influential even before Zhang's time-as may be demonstrated by the figure of Xue Baochai in the novel Hong lou meng, who claims that poetry writing is "not really [women's] proper business. " 111 As mentioned above, some women writers, in an attempt to conform to the orthodox Confucian ideal of female virtue, even burned their poems. And there were many others who constantly felt pressured to prove that they did not neglect their household duties in spite of their poetry-writing activities. It was this kind of concern that made many women give their collected works titles like Xiu yu cao (Poems composed after embroidery), Gong yu cao (Poems composed after needlework), or Zhi yu cao (Poems composed after weaving). In his Lidai funii zhuzuo kao (A study of women's writings through the ages), Hu Wenkai records more than r7o such titles published by Ming and Qing women poets. 112 Of course, women authors might have adopted such titles for their works due to social codes of modesty, pretending that they were not really "poets" because their poems were done only during leisure hours after they had completed their domestic duties. This kind of playing down of poetic achievement, however, often works to promote, rather than belittle, women's literary position-for the fact that these women poets could achieve so much during their part-time activity proves their irrefutable talent for poetry. 113 In any case, the truth remains that these "modest" women were afraid of being judged as transgressing the bounds of moral propriety. However, under the influence of orthodox Confucianism, many Qing women writers (like many male scholars) sincerely believed that the value

Kang-i Sun Chang

255

of a poetic text lies in its moral power rather than the strength of the author's creativity. Significantly, some women scholars further claimed that moral power is a function of the female sex-a concept of morality poignantly recalling Zhang Xuecheng's idea about women as upholders of ethical education ..Perhaps the most representative Qing female scholar to hold such views was Wanyan Yunzhu (r77I-I833), known as the "female Confucianist" (nii zhong zhi ru). 114 In her anthology Guo chao guixiu zhengshi ji (Correct beginnings: collected women's poetry of our dynasty), published in r83r, she explicitly made moral edification the principle of her selection of poems. 115 In so doing, she seemed to be challenging Yuan Mei's liberal-minded view of literature, according to which spontaneous self-expression came before the didactic function of poetry. Although her claim that the Shijing "did not eliminate writings from the women's quarters" 110-hardly a new idea by now-seems to recall Yuan Mei's famous view, her insistence on women poets' moral function instead of their literary innovation echoes orthodox Confucian concerns. The fact that the title of her anthology takes the term zhengshi ("correct beginnings") directly from the Preface to the first two books of the Shijing is especially worth noting. Zhengshi is of course the epithet for the first poem in that collection, "Guanju," which gives prominence to "the virtues of the queen." Thus, whereas Yuan Mei took the love songs in the Shijing at their word, Wanyuan Yunzhu (like Zhang Xuecheng) would be inclined to read them as allegorical poems, promoting the moderated emotions of the virtuous wife. A consequence of this is the fact that no anthologist before Wanyan Yunzhu had been so antagonistic to courtesans. To her, courtesans represented the worst offense to wifely virtues. She called them "women who had lost their virtues" (shijie furen), and prided herself on excluding their works fromher anthology ("Li yan," sa). The only exceptions were a few earlier "reformed" women like Liu Shi, Wang Wei, and Bian Yujing (early seventeenth c.) who managed to cultivate their "virtues in their late years" (wanjie). Even so, Wanyan Yunzhu relegated their poems to the appendix (not the main text)-and reprinted only two poems by Liu Shi, two poems by Wang Wei, and one by Bian Yujing. She seems to have completely ignored the fact that although in her own day courtesans were no longer prominent in the world of refined letters, early Qing courtesans had provided the popular model of the "talented woman" (cai nii). In any case, Wanyan Yunzhu's categorical judgment of courtesans as women of misconduct gives us a glimpse into the moral views of many gentrywomen after the mid-Qing-which in many ways were even more

MING-QING WOMEN POETS

"conservative" than those of the male scholar Zhang Xuecheng, who at least respected the Tang courtesans for preserving the ancient tradition of women's culture (ZXCY, 48). In the meantime, many Qing women poets objected to the moralists' position. For example, Xia Yilan, an early nineteenth-century poet, complained in a poem about the constant pressures on women to value virtue over talent: A.~f-~W:::t ~~11l1Jff.Hf

m:¥WliOO:::t

M~ifB·llif;fEI&: ~~:::tPJ~ ~ 1!! N :t.!'[

:::t

iZ~J]JFf~ Jlt~ifli>'*~~

It's said a man's virtue and talent Are good provided he has both. Yet when judging a talented woman, The standard of evaluation is reversed: To a virtuous female, added talent is allowed, But a talented woman's morality will be in question ... To my mind, this is simply wrongA biased and shallow view.... 11 7

Xia Yilan, who died at the young age of fourteen, represents a familiar figure in the fiction and history of late imperial China-the figure of the female genius. An ambivalent figure, the female genius is regarded with a mixture of admiration and unease. She is the frail beauty of talent, a celestial being not intended for the mortal world and thus not destined to live long. Starting with the real person Ye Xiaoluan (I 6 I 6-3 2) in the late Ming and culminating in the fictional character of Lin Daiyu in the Hong lou meng-who both died in their teens 118-the figure of the female genius has become the sublime model of female creativity that harbors at once a peculiar power and a threat that is both compelling and mysterious. She is almost always a gentrywoman, replacing the figure of the talented courtesan. Her persistent appearance in the Ming and Qing marks the extent of enthusiasm and anxieties prompted by the unprecedented increase of gentrywomen poets. Hence, the frequency of Ye Xiaoluan-like figures in novels, memoirs, and critical works has a special relevance in the context of the debate between Yuan Mei and Zhang Xuecheng. A case in point is a female disciple of Yuan Mei named Jin Yi who died at the tender age of 24. Yuan Mei had a particular liking and sympathy for her. Frail, gifted, and beautiful, Jin Yi strongly recalls the character of the fictional Lin Daiyu. In fact, Jin Yi identified herself with the talented and sentimental Lin Daiyu, as may be seen in her poem entitled, "On a wintry night I read Hong lou meng while waiting for Zhushi. " 119 It is thus not surprising that in writing his epitaph for Jin Yi, Yuan Mei drew on the popular image of the female genius:

Kang-i Sun Chang

257

There was a lady from Suzhou called Jin Xianxian. Her given name was Yi. From the time she was born, she was delicate and fragile and had a countenance of surpassing beauty. At a very early age she was already reading books and distinguishing the four tones. She loved to compose rhymed verse, and each time she let fall her brush, it was like a fleet horse prancing along unable to stop.120 Greatly saddened by Jin Yi's untimely death, Yuan Mei mourned: I've been around for a long time, and each time I encounter a woman with talent I find her to be unfortunate. Those who have beauty as well are still more unfortunate, and those who have talent, beauty, and a good match are the most unfortunate of all. In Xianxian were combined these three harbingers of misfortune, and yet I wish she could have lived a long life. Is it not also hard to bear that my three younger sisters all were talented and all died an early death? ... Now Xianxian has also died, and now I know that the heavens are more loathe to part with an auspicious pair living in seemingly everlasting harmony than with generals, ministers, or men of high position. This indeed is the unalterable principle of the creator. What more can be said of it? Like Ye Xiaoluan's father, Ye Shaoyuan, who comforted himself with visions of his daughter turning into an immortal fairy after her death 121 (another detail that recalls the career of Lin Daiyu in Hong lou meng), Yuan Mei imagined the talented Jin Yi to have merely "gone to the Nine Doubts Mountain range to visit Goddesses Ying and Huang" near where her "bones of poetry" are preserved. In all this, the figure of the historical female genius is insistently allegorized and mythified as the very spirit of poetry. One may say that the female genius has managed to acquire an enormous power in spite of the limitations of the roles assigned to her in life or in the critical theories of her time. In a sense, she has transcended the alternative between morality and talent. But the cailde debate was too fundamental a cultural concern to disappear from late imperial writings about women authors and women's education. Especially during the Qing there were controversies regarding women's function in literature and society. As we survey the history of Chinese women poets, we can see Ming and Qing views of women adopt the language of.earlier times while subtly altering their import: the alternative of cai!de meant something different for an audience that had the Tang courtesans, the Ming "women of talent," or the Qing gentrywomen poets in mind. The problem with many modern scholars is that they tend to take that notorious and much distorted slogan "A woman without talent is a virtuous woman" as representing the whole and unchanging

MING-QING WOMEN POETS

view of traditional men toward women. As a result, today's scholars are inclined to believe that they themselves are the first to counter the traditional discourses of "male oppression." In truth, if there were discourses of oppression at all, they were instituted and maintained by both men and women, as has been shown in this chapter. "Oppression" implies a contrasting standard of freedom or autonomy; and the gentrywomen of the Qing who insisted on their ability to combine cai with de were laying claim to the freedom implicit in writing and publishing. There is, as I see it, a lesson here for the contemporary critique of the past. Positions and demarcations that seem of paramount importance to contemporary scholars do not necessarily translate into the terms most relevant to Ming and Qing China; least of all do ideological factions of the time always break down along gender lines. The question that needs asking is rather: What predisposed individuals to opt for the defense, revision, or rejection of an idea such as the opposition of cai and de? It is for this reason that I have attempted to uncover the alternatives available within traditional discourse regarding the question of women's talent and virtue, so as to warrant a more historically specific inquiry into the issues as they were discussed in their original contexts. 122

CHAPTER IO

The Scorpion in the Scholar's Cap: Ritual, Memory, and Desire in

Rulin waishi Marston Anderson [Editors' note: Although Marston Anderson attended the Laguna Conference in 1992, this was not the paper he presented there. This paper was read first in March 1988 at the Columbia University Traditional China Seminar, and since that time has garnered considerable praise as it circulated in photocopied form. Since we consider it not only an outstanding paper but also one that resonates remarkably well with the themes brought up at the conference, we decided to include it in this volume. Although Marston never had the opportunity before his untimely death to revise the paper for publication, the typescript as he left it required very little editing of any sort, and the version published here is thus virtually identical to his original text.]

The writers and critics of the late Qing and May Fourth periods firmly established the reputation of Wu Jingzi's (r70I-54) Rulin waishi as one of the great masterpieces of Chinese literature. At the same time, however, they imposed an interpretation of the novel that has come to seem limiting, if not frankly misleading. Hu Shi was the first to single out Wang Mian's denunciation of the Ming examination system in the opening chapter as the "theme of the whole book," 1 and Lu Xun was similarly impressed with the novel's critical social message: the book was a "candle" in whose light was exposed the pretensions of scholars who "studied nothing but what was required for the examination essays. " 2 These views marked a divergence from those of eighteenth- and nineteenth-century commentators, who had found the work's theme not in Wang Mian's

260

RITUAL, MEMORY, AND DESIRE IN RULIN \\f11ISHI

censure of the examination system but in the opening poem of the novel, in which the author laments the vanity of human ambition (specifically the pursuit of gong ming fu gui, "career, fame, riches, and rank"). 3 For these early critics, the novel constituted a broad attack-made from the standpoint of traditional Confucian ethics-on all manner of human striving and pretension. Of the novel's huge cast of characters, the many examples of mingshi, scholars with literary affectations who disdain conventional bureaucratic ambitions, interested them far more than the aspiring examination candidates (such as Zhou Jin and Fan Jin), who figure prominently only in the opening chapters. 4 With the New Culture movement of the early twentieth century, however, this didactic view of Rulin waishi gave way to a new sense of the novel as a systematic critique of the Ming-Qing social system, an interpretation that clearly served the needs of May Fourth intellectuals in their search for an indigenous example of fiction as Kulturkritih. Understood as such, Rulin waishi exerted an unparalleled influence on twentieth-century Chinese fiction, serving as a model not only for the "novels of exposure" of the late Qing but also for much of the satirical fiction produced in the 1920s and 1930s. Unfortunately, May Fourth reinterpretations of the novel also gave rise to the sadly reductive, but now oft-repeated characterization of it as a "satire of the examination system." Although the critical social content that May Fourth intellectuals found in Rutin waishi elevated the novel in their eyes, their reinterpretation paradoxically involved a devaluation of Wu Jingzi's formal and stylistic achievement. Earlier commentators had without exception praised Rutin waishi for its structural integrity. The author of the chapter-end comments in the I So 5 Woxian caotang edition, for example, used an architectural metaphor to describe the form of the novel: the complete work was a "palace," in which the sacrifice to Tai Bo in chapter 37 was the "central hall." 5 The sacrifice scene was the book's "great culmination" (da jieshu) or climax, for which the reader was prepared by a series of "small culminations," specifically the gatherings at Oriole-Throat Lake in chapter r2 and at West Lake in chapter r 8. This same commentator wrote glowingly of the flexibility of Wu Jingzi's style, which "with its many transformations is too marvelous to describe through a single example. " 6 In I 8 69 Jin He wrote in similar terms: "The book's style is meticulous, superior to Ji Yun's Notes of the Yuewei Hermitage. The novel progresses through a series of neat transitions, and suffers not at all from digressiveness-thus facilitating the reader's memory. " 7 Hu Shi and Lu Xun, however, sought in the novel the unifying effects that they had learned to expect from

Marston Anderson

Western fiction and were clearly disturbed by its episodic quality. Hu Shi complained that "the whole work seems thrown together,'' 8 so that it appears to be entirely "without structure. " 9 Lu Xun similarly observed that the novel "lacks one great design" and was "like a group of short stories or a patchwork quilt of silk. " 10 The May Fourth view of Rutin waishi has established itself as the standard interpretation i!). the People's Republic and continues to be influential in the West, particularly among historians. 11 Recent assessments by literary scholars, however, have generally involved, to one degree or another, an attempt to revise the May Fourth critical legacy. C. T. Hsia has emphasized that the novel was "consciously written from the Confucian point of view,'' adding, however, that its Confucianism is "tinged with melancholy over the futility of government action or social reform." 12 Other critics in the West have singled out specific elements of Confucian thought as dominant: Timothy Wong has joined Qing critics in identifying the vanity of gong ming fu gui as the work's central theme, 13 and Shuenfu Lin has focused on the principle of "ritual" (li) as both a formal and thematic element within the novel. 14 Even mainland critics, who otherwise have much invested in a notion of Rutin waishi as social critique, sometimes hark back to Qing didactic interpretations in the effort to defend the novel's formal symmetry: Huang Bingze, for example, divides the book into four sections, each of which he argues is devoted to characters possessed of a common moral failing. 15 I offer this brief review of the critical reception accorded Rulin waishi since its appearance both to remind the reader of the fundamental problems of interpretation that have been raised in connection with it and to demonstrate the unique intertwining of formal and thematic issues in all discussions of the novel. It appears that eighteenth- and nineteenth-century interpreters, with their traditional Confucian perspective, were able to perceive a unity in the book that escaped May Fourth readers, and conversely that recent critics, in defending its formal integrity, have felt the need to reassert the fundamentally Confucian nature of its message. One thing contemporary critics do appear to share with their May Fourth predecessors, however, is a powerful appetite for ideological consistency. Whereas traditional critics took the irregularities of the novel's ideological terrain in stride, twentieth-century readers have eagerly sought to identify an underlying doctrinal stance in the book, discovering everything from "social realism" to "Confucian idealism." Any such identification is accomplished with some effort, however, because of the novel's notorious rhetorical complexity-a complexity that moved one reader to complain

262

RITUAL, MEMORY, AND DESIRE IN RULIN WAISHI

that Wu Jingzi "certainly cannot be said to favor his readers with a lucid or consistent point of view. " 16 In the end any didactic reading of the novel must contend with a plot line that repeatedly, and one might think perversely, rewards depravity and punishes virtue. What positive values are discoverable in the novel must be deduced from often subtle rhetorical clues and are demonstrated more often than not by their absence. 17 Recent critics are certainly right to remind us that the intellectual horizon of the novel is Confucianism, but that is not quite the same thing as saying that the book advocates any particular strain of Confucianism, or even Confucianism itself. While clearly a meditation on Confucian themes, the text of Rulin waishi subjects those themes to rhetorical and narratological operations that have yet to be adequately described. Only when we have accomplished such a rhetorical analysis will it be possible to determine more clearly Rulin waishi's relation to Confucianism. Perhaps the primary reason Rulin waishi has proved so troublesome for critics is its resistance to allegorical interpretation. Of course, this resistance in itself suggests an affinity with Confucianism, given the fact that traditional Chinese novelists conventionally organized their materials within allegorical frameworks that relied heavily on Daoist and Buddhist mythology. At one point in the opening chapter, the text of Rulin waishi seems on the verge of assuming for itself such a framework: the reclusive Wang Mian, observing a "hundred small stars" falling toward the southeast horizon, declares that Heaven has sent them out of pity "to maintain the literary tradition. " 18 This scene is reminiscent of the opening chapter of Shui hu zhuan, which narrates the emancipation of 108 devil spirits who are later equated with the bandit-heroes of the saga, and some critics have suggested that Wu Jingzi intended the reader to recognize a similar correlation between his Heaven-sent stars and the virtuous scholars who gather to revive Confucian rituals in Chapter 37· This parallel, however, is never sufficiently developed in the novel, and Wang Mian's cryptic hint in fact serves only to highlight the absence of an operative allegorical structure in the work as a whole. In the same way, other devices traditionally employed by Chinese novelists to introduce an allegorical dimension into their fiction, such as dreams and fortune-telling, also surface occasionally in the novel, usually to be subverted in the interests of a satirical joust. For example, early in the novel the fortune-teller Zhen Hefu uses the planchette to read the aspiring scholar Wang Hui's fortune. Zhen then interprets one line of the augury, "The horses leading the way turn out to be ministers of the celestial court," as a metaphor for Wang's eventual promotion to prime minister. 19

Marston Anderson

In the following chapter, this line from the augury in indeed borne out, but not in the way we were led to expect: while staying in a public hostel, Wang Hui discovers a wall-hanging inscribed with the first half of the line ("The horses lead the way"), which during the night is blown away to uncover a placard on which is printed the rest of the line ("Ministers in the celestial court"). The hapless Wang Hui takes this as proof of universal predestination. We are clearly to learn from this vignette, as from the more celebrated instance of Ma Chunshang's victimization at the hands of a self-styled "immortal" in chapters 14 and 15, that greed and ambition render one gullible to ostensible manifestations of the supernatural. 20 But more interesting for our concerns here is the way in which the narrative asserts this point. The text pointedly blocks Zhen Hefu's metaphorical interpretation of the augury, substituting for it an arbitrary verbal contiguity (exemplified by the chance proximity of the two placards); as a result the prophecy through its very realization is rendered irrelevant to the deep-seated ambitions that motivated Wang Hui to consult the fortuneteller in the first place. Although not all passages in the novel that involve dreams and the supernatural conform to this pattern, 21 many must be viewed as intentional parody of the allegorizing tendencies of earlier Chinese fiction. Rutin waishi would appear, then, not only to lack an explicit allegorical framework but actively to discourage metaphorical modes of reading. At least on its surface, the novel proceeds not through vertical reference to a larger mythological or ideational structure but through a series of horizontal linkages that sometimes seem as arbitrary as the association of the placards in the episode just recounted. In broader rhetorical terms, we may say that the text favors metonymy, or relations of contiguity, over metaphor, or relations of resemblance. 22 As Paul de Man has observed, metaphor carries a suggestion of necessity, metonymy of chance/3 and one of the consequences of Wu Jingzi's preference for metonymy is the critical perception that his book is little more than a random assembly of "short stories." Metonymy, however, embraces not only chance affiliations but also relations of cause and effect, and at least within individual episodes of the novel little happens that strikes the reader as unmotivated or that violates one's sense of causality. Indeed, Rutin waishi has been credited with achieving a new level of "realism" in Chinese literature largely because it observes a relatively strict verisimilitude. The many chance meetings the novel employs in transitional passages bear little resemblance to the kind of forced coincidental occurrences that riddle much of traditional Chinese fiction, primarily because in Rutin waishi coincidence is

264

RITUAL, MEMORY, AND DESIRE IN RULIN W11ISHI

rarely represented as a disguised agent of fate. It is therefore not the chance encounters themselves that strike the reader as improbable or arbitrary in Rulin waishi, but rather the shifting focus of the narrator as he directs his attention from one character's story to another's. One effect of these shifts in narrative focus, and of the text's general affinity for metonymy, is to force the reader to supply the missing metaphorical associations that would make of the work a "closed and legible whole. " 24 Much of the interpretive effort accorded Rulin waishi to date has amounted to just this kind of systematizing labor and has primarily involved recognition of thematic correspondences among the various episodes. Viewed this way, the text appears to be a kind of anatomy, offering through a series of examples a Confucian taxonomy of human behavior. Alvin B. Kernan has observed that the anatomy is a natural form for satirical works, which are predisposed toward an underlying plot line in which one variety of human pretension after another is subjected to ironic deflation. This drives satirical works toward a "dismemberment of form," which necessitates the use of such omnibus structures as anatomy. 25 But as C. T. Hsia and others have observed, thematically similar units are often juxtaposed in Rulin waishi, suggesting that each episode is in part chosen for its contextual thematic resonance: Fan Jin and Zhou Jin are an obvious case of such a pairing, as are the young charlatans Niu Pulang and Kuang Chaoren. 26 Once the correspondences between adjoining episodes are recognized, individual segments of the narrative reveal themselves not as simple exempla of abstract moral categories but as legs of a larger thematic argument. For example, Shuen-fu Lin, in his "spatial" reading of the novel, follows C. T. Hsia in dividing the interior chapters of the text into three large units. These are then viewed as constructing a broad chain of ethical reasoning: the first part of the book (chaps. 2-30) with its many satirical portraits argues the need for a revival of Confucian ritual, the climactic middle section (3 1-37) describes the preparation for and culmination of such a ritual in the sacrifices to Tai Bo, and the last third of the book demonstrates the failure of those sacrifices to have any long-term effect on society. 27 Such an outline seems to me convincing, although in the absence of an explicit allegorical structure in the novel any thematic blueprint must be recognized as extrinsic to the text, the product of a reader's reconstruction; given the richness of incident in Rulin waishi, any reader's interpretation will be selective and partial. In the discussion that follows, however, I would like to take Shuen-fu Lin's general outline of the novel's thematic progression and overlay a somewhat more detailed scheme of my own,

Marston Anderson

for which I make no claim of conclusivity. As noted above, the several chapters that immediately follow the prefatory story of Wang Mian offer the most convincing rationale for the May Fourth reading of the novel as systematic social critique: they take pointed aim at aspiring examination candidates and venal officials, characters whose primary objective is to play a role in the public arena (by which I mean the imperially recognized world of the bureaucracy). Through the stories of Zhou Jin, Fan Jin, Senior Licentiate Yan, and Wang Hui, the vanity of such ambitions is rapidly established, leaving the narrator free to scrutinize at greater length another element of the social environment, the private cliques of mingshi. We are shown in detail three such cliques (four if one includes Du Shaoqing's circle), and each group portrait is capped by a "small culmination," consisting of a social gathering organized and attended by the clique. In each case, the participants themselves intend their gathering to serve as a demonstration of the group's high-minded detachment from public affairs. But in fact this pretense only thinly disguises a more fundamental hope that the occasion will somehow augment the personal fame of the participants and thus help them translate their private alienation into a kind of public influence. As we soon learn, lust for fame has contaminated even the proudest expression of the mingshi's disengagement from society, their poetry: repeatedly the language of their compositions is exposed as derivative not only of early poets but even of the examination essays that they deride. 28 A fundamental irony suggests itself: the more loudly the mingshi insist on their detachment, the more they reveal their resemblance to bureaucrats. Poetic and bureaucratic ambitions are frankly equated in chapter 17, where a group of mingshi discuss the relative accomplishments of two acquaintances, one a famous poet blessed with a happy family life but no official position, the other a magistrate whose wife has died young, leaving him childless. The scholars agree that the poet has the better lot not because of his domestic circumstances, but because his poetry has made him "much more famous than most scholars who have passed the metropolitan examination." 29 The pedagogical consequences of the fractious social environment forged jointly by the pseudo-scholars and bureaucrats are explored in the adjoining tales ofKuang Chaoren and Niu Pulang in chapters 15 through 24. Though different in character, Kuang and Niu both learn to manipulate, through self-betrayal and subterfuge, the gap between public and private life depicted in the early sections of the novel. As we first meet Kuang Chaoren, he is a model filial son; his naivete, however, makes him receptive to the counsel of the well-intentioned but foolish editor Ma

266

RITUAL, MEMORY, AND DESIRE IN RULIN WAISHI

Chunshang, who plants in him the fatally corrupting idea that "passing the official examinations is the most important way of pleasing your parents. " 30 Ma's advice conjoins the private virtue of filiality with ambition for public office, two things that Kuang's parents, simple villagers, instinctively know to be incompatible. Parental instinct soon proves correct: Kuang's studies draw him away from his family into the hands of a series of patrons, some of them bureaucrats, some of them mingshi. With his quick wit, Kuang rapidly absorbs from each new benefactor first the nature of his ambitions and then the social and practical skills required to satisfy them. But devoid of an interior moral faculty, Kuang becomes guilty of increasingly sinister acts of betrayal, not only of the patrons he outgrows but also of his youthful identity and values. This has consequences to both his public and private personae: he first undermines the very system that elevated him by serving as an illegal substitute at an examination, and by the end of the story he is dissembling his marital status in order to impress his superiors. In rationalizing the bigamy that follows from this latter instance of bad faith, he remembers a dramatic prototype: "There is an opera about Cai the Number One Scholar who had two wives, and it was considered a great romance; what can it matter?" 31 Here Kuang has indeed chosen an appropriate literary antecedent, but he has failed to get its message right: the early chuanqi play he cites here, Pipa ji, is not a paean to bigamy but, like Kuang Chaoren's own story, an extended examination of the incompatibility of filial piety and bureaucratic ambitions. Niu Pulang, for his part, is a fraud from the start, having entered society with a false name and identity (that of an elderly poet whose name by chance resembles his own). What is of interest in his story is not the internal development of his character but the way in which the surrounding society abets his duplicity at every turn. 32 The monk in his local village encourages his childhood thefts of books, thinking them the sign of budding scholarly aspirations; later the pseudo-scholars Niu meets are so eager for the benefits their acquaintance with a famous poet will bring that they overlook the many incongruities in his story; as a final irony, the magistrate to whom the widow of the real Niu Pulang appeals for redress dismisses her case. In both Kuang's and Niu's stories, we view the process by which falsely assumed identities are rcified and eventually assume the status of unalterable social facts. Having established self-betrayal and impersonation as the underlying realities of social intercourse, the text's attention turns to characters who make their living by impersonation, that is, members of the theatrical

Marston Anderson

profession. The introduction of this theme is neatly woven into Niu Pulang's story: on the occasion of one of his temporary setbacks, when he is stripped, beaten, and left to die by another impostor, it is a theatrical costumer who rescues Niu and supplies him with the scholar's gown he requires to resume his hoax. Costuming as a necessary accessory for the social impostor is evoked again in a later incident involving the actor Bao Wenqing. On encountering an old colleague dressed in a scholar's blue gown, Bao asks him, "If you wear clothes like that, what can the scholars wear?" In reply the old man boasts that actors are now more welcome in society than even some high officials. 33 But it is not the arrogance of actors who fail to recognize their place that is the author's primary target in this section of the book: indeed, Bao Wenqing's fastidiously principled behavior elsewhere in the novel suggests that the acting profession may actually instill in its practitioners a more exacting sense of the difference between true and false behavior than is available to the average man. 34 Wu Jingzi's moral indignation is provoked rather by the literati, whose interest in theater and actors arises from an unexamined fascination with artifice and impersonation. The beauty contest for actors that Du Shenqing organizes in chapter 30-significantly the last major public gathering before the sacrifice to Tai So-represents the ironic culmination of this theme: it is a full-dress dramatization of the unhealthy intercourse between two kinds of impersonators, the more properly professional ones, the actors, and a more insidious variety, the literati. The stage is then set for an ethical countermeasure in the climactic central chapters of the book, which describe the preparation for and observance of the sacrifice to Tai Bo by Du Shaoqing and his circle. These sacrifices are in part undertaken out of a desire to mend the private-public gap depicted in the first section of the novel. They have a specific pedagogical and political objective: in the words of Chi Hengshan, who proposes them, they are to "produce some talents [rencai] who will be able to serve the government well. " 35 But the rites serve a local aim as well, in that they are dedicated to ''the worthiest man Nanjing has produced," Tai Bo. They are also of great personal significance for the autobiographical character Du Shaoqing: through the rites he hopes to recover a full sense of personal authenticity and salvage his declining reputation. 36 At least in their public mission, the rites must be judged a failure, for reasons discussed below, although their influence continues to be felt throughout the latter third of the novel. In this final section, we witness many other individual efforts to restore moral order to the world: through a personal quest to revive lost familial attachments (Guo Xiaozi), through military

268

RITUAL, MEMORY, AND DESIRE IN RULIN WAISHI

and civil campaigns (Xiao Yunxian, Tang Zhentai), through righteous defiance of social convention (Shen Qiongzhi), and through an over-literal observance of Confucian codes of filiality and etiquette (Wang Yuhui). When each of these efforts fails, the character who initiated it becomes consumed with nostalgia for the sacrifice to Tai Bo, which has come to represent a collective effort at reform that is beyond an individual's capacity to duplicate. On seeking out the rite's original participants, however, one after another of these latecomers discovers them tragically disbanded, confirming the aura of thwarted moral purpose that surrounds all recollection of the rites. Having developed such a thematic reading of the novel, I must now admit that I find it wanting, because it fails to take into account tqe feature of the novel that H. C. Chang called its greatest technical achievement, greater "than even its satire and parody," namely, the novel's sense of "subjective time. " 37 Indeed, to ignore Rulin waishi's unique manipulation of time is to fail to give full credence to what we have identified as its most basic rhetorical attribute, its affinity for metonymy. It is, after all, through the series of syntagmatic linkages governed by the trope of metonymy that the text establishes its temporal dynamic. This is in addition, of course, to the obvious markers of historical chronology in the novel, such as the narrator's periodic announcement of the reign year, from which we learn that the events in chapters 2-55 occur between 1487 and 1595. Since the episodic structure of Rulin waishi denies us the naturalized sense of time passing that would emerge from watching a single character age, these "objective" markers provide an important temporal framework for the events of the story. But at a more organic level, we are made aware of temporal progression by the succession of generations in the novel's shifting cast of characters (by the appearance, for example, of Wang Hui's son in chapter 37, or of Zhen Hefu's son in the penultimate chapter). It is this generational succession to which H. C. Chang is referring when using the term "subjective time." But the text conditions our sense of time at an even more elemental level than this, by manipulating the forward-driving impulse of reading itself as we pass from episode to episode. The object of the rest of this chapter is to examine the rhetoric of temporality in a critical passage of the novel, and then to consider the consequences of a time-centered reading of the text on the thematic outline offered above. The description of the sacrifices to Tai Bo in chapter 3 7 has long been recognized as the novel's climax and symbolic nexus. The passage itself, however, has seemed to many readers singularly anticlimactic. In C. T.

Marston Anderson

Hsia's words, it is "unduly disappointing in merely giving us a bald summary of the proceedings in a solemn fashion. " 38 Indeed, much of the depiction of the ritual consists simply of intoned ceremonial injunctions with unembellished reports of the corresponding ceremonial gestures: "Take your places!" cried Chi. "Kneel!" Dr. Yu knelt before the table. "Offer wine!" Ji Jiao knelt to pass the wine to Dr. Yu, who placed it on the table. "Offer jade!" Ju Laixun knelt to pass the jade to Dr. Yu, who placed it on the table. 39

And so forth. The rite itself emerges clearly in a way that may interest students of religious anthropology, but the written account seems devoid of novelistic interest: what we miss above all is any hint of the psychological response of the participants. In this the passage stands in marked contrast to other scenes in the book that recount rituals, such as the marriage of Ju Gongsun and Miss Lu in chapter ro or the enshrining of the virtuous women in chapter 4 7, both of which are saturated with rhetorical clues alerting us to the secret motives of the participants. The sacrifices to Tai Bo, however, are psychologically neutral, or perhaps one should say neutralizing, given their numbing effects on many readers. To understand why Wu Jingzi chose this particularly flavorless mode of narration for his climactic passage, it is necessary to consider the nature and intention of a Confucian ritual such as this one. Etymologically li means simply "treading" or "following," suggesting that ideally the rites represent a kind of perfect re-enactment of the movements and gestures of the sages who formulated them, performed not simply in the interests of pantomime, of course, but to attain a moral and psychological identification of the celebrants with the ancient sages. 40 This psychological effect was of particular interest to Xunzi, the classical Confucian thinker who offers the fullest explanation of li. Writing at a time when the ancient rituals had already declined sufficiently to require a defense, Xunzi wrote that the rites were designed above all to mediate the natural human emotions of desire and memory: "The ancient kings ... established ritual principles ... to train men's desires and to provide for their satisfaction. They saw to it that desires did not overextend the means for their satisfaction, and material goods did not fall short of what was desired. " 41 And again: "The sacrificial rites originate in the emotions of remembrance and longing for the dead. Everyone is at times visited with depression and melancholy longing .... Therefore, the

270

RITUAL, MEMORY, AND DESIRE IN RULIN WAISHI

former kings established certain forms to be observed so that men could fulfill their duty to honor those who deserve honor and show affection for those who command affection. " 42 The emotions Xunzi specifies are, significantly, those that stir in the individual thoughts of a future or past generation, emotions that implicate him in the temporal flux. To successfully moderate these emotions through ritual is to come to terms in a profound way with human temporal limitations. Having tamed desire and memory, one will have achieved, in Xunzi's words, an "ordering" of birth and death. In two senses, then, the successful performance of the rites may be understood to involve a mastering of time: in their mediating psychological effect on the individual participant; and in their power (through exact imitation of the ancient words and gestures) to make the sages of antiquity coterminous with the contemporary celebrants. In this light we can, I think, begin to see that the unembellished rhetoric of chapter 37 represents an attempt by the narrator to find a stylistic equivalent of the tranquil yet sober psychological state that the ritual's enactment was intended to foster. The passage's litany of injunctions and ceremonial gestures suggests a perfect consonance between the text of the ritual (whose utterance by the Masters of Ceremony revives the voices of the ancient sages) and its contemporary performance. This unity of word and action creates a privileged temporal precinct, in which the participants, freed of disturbing time-bound emotions, may give themselves fully to the present tense of the ritual. As we have observed, however, the ironic consequence of this for the reader is an obliteration of narrative interest, and this fact suggests by way of contrast a fundamental truth about the depiction of human behavior in the novel: conventional fictional representation assumes precisely that contamination with temporally delimiting emotions that the rituals are designed to counter. It is memory of, or desire for, an absent glory that defines and characterizes the social identities of all the characters whose adventures we have followed up to this point in the novel. Under the influence of memory or desire, one character after another has labored to recast his social persona by creatively equating his present circumstances or credentials with those of an imagined future or of an envied rival; in other words, one character after another, through varied acts of plagiarism and impersonation, has attempted a metaphorical reconstruction of the self. When this kind of self-invention becomes endemic in society, it results in the general theatricalization of social behavior that Wu Jingzi repeatedly illustrates. Of course, rituals themselves offer a perfect occasion for

Marston Anderson

271

such theatricalization, a fact that is insinuated frequently in the novel by the prominent place given dramatic performances at ceremonial observances. At the wedding and enshrining mentioned above, for example, the presence of the actors not only distracts the attention of the celebrants but eventually lures them away from the ceremony proper. The de~ication of the temple to Tai Bo is similarly threatened, a fact that is suggested obliquely, if in no other way, by its juxtaposition to the strictly theatrical "small consummation" that preceded it (Du Shenqing's beauty contest for female impersonators). Wu Jingzi's careful rhetorical treatment of the dedication scene is thus necessary to indicate its seriousness of purpose and to distinguish it from all other such observances. It may not be too much to suggest that the passage is intended as the text's rhetorical degree zero: through it the narrator, by draining his style of all contaminating rhetorical flourishes, strives to attain an authenticity of voice corresponding to the psychological state that the ritual ideally produces in its participants. But a full understanding of chapter 37 cannot rely entirely on an isolated rhetorical analysis: we must consider its place in the larger story. Contextually, the ritual itself becomes a primary object of both desire (on the part of its organizers in the preceding chapters) and of memory (on the part of the many characters in the latter part of the novel who long for it). This is entirely in line with Xunzi's view of the mediating power of ritual: desire and memory are, in his view, unavoidable natural impulses, which can never be eradicated but may be beneficially displaced onto the rites themselves. It is thus entirely fitting that in Rutin waishi the rites should become the eventual focus for the multifarious desires that propel the plot line of the first half of the novel; through ritual, desire is transformed into a kind of utopianism, or at least a kind of moral reformism. So, too, the acute sense of loss that afflicts so many characters in the concluding section of the book ought ideally to be converted by the rites into a generalized reverence for China's shared ancestors, the sages of antiquity who originally formulated the rites. For the reader, too, the first half of the book proceeds under the sign of desire, the latter half under the sign of memory. That this second half has proved less hospitable to critics is in part due to the textual problems associated with it; 43 it is also, I think, the fault of our critical terminology, which is better equipped to explicate a fiction of desire than of memory. Xunzi's observation that memory itself is a kind of longing may be of help here: the local momentum of each individual episode in the final section of the book is generated by longing, but at the closure of each episode, whereas earlier incidents directed our interest forward to a long-delayed "culmination," later epi-

272

RITUAL, MEMORY, AND DESIRE IN RULIN WAISHI

sodes instead fold back, returning us again and again to the still central point of the sacrifices to Tai Bo. Beyond these psychological effects, however, there remains the question of the pragmatic influence of the ritual on the narrated world of the novel. If the ritual is to accomplish its political and pedagogical objective, it must extend its authority beyond the present tense of the ceremony, outward into the broader social world and forward into the future. The immediate response to the ritual is encouraging: as Dr. Yu and the others depart from the temple, the local people line the road, crying for joy, "We have never seen a ceremony like this, or heard music like this before! The old folk told us the gentleman who was the master of sacrifice is a sage who has come back to the earth, so we all wanted to see him. " 44 The ceremony has brought not only the celebrants but the entire community into a renewed relationship with the ancient sages (though the reader begins to notice here, in the exaggerated description of Dr. Yu as a sage incarnate, the re-emergence of a satirical tone). But this level of enthusiasm is restricted to the immediate neighborhood and is short-lived. After the scholars disperse, the initial effect of the ritual fades into a purely symbolic influence; it increasingly becomes only the object of a distant longing. The many discontented idealists who people the last half of the book, however, repeatedly endeavor to reduce this distance by personally seeking out Dr. Yu and his circle: just as the local people in Nanjing express a powerful desire to make personal contact with the "sage," these latecomers seem driven by a need to turn their symbolic relation with the ritual into a metonymic one. Wang Yuhui, whose story is a revealing variation on the theme of filial piety, is one of the most poignant examples of such a character. Having devoted his life to the thankless task of compiling ethical primers, Wang Yuhui descends into such poverty that his daughter, in order to unburden the family and demonstrate her filiality, starves herself to death. Although recognizing his daughter has acted in conformity with the most exalted ancient models (and is therefore now an "immortal" whose story he might include in one of his collections), Wang cannot accept her loss. Earlier in the novel, most conspicuously in the stories of Wang Mian and Kuang Chaoren, filial piety has been presented as one of the bedrock Confucian virtues, and the constancy of family relations has been set in striking contrast to the promiscuous and hypocritical interchange of social life. As the novel progresses, however, we observe more and more cases of filiality gone awry. The autobiographical character of Du Shaoqing is, of course, the most notable example of this: out of an excessive love for

Marston Anderson

273

his departed father, he recklessly squanders the family fortune on his father's acquaintances, thus destroying the reputation and prospects of the living family line. In a late incident, Wang Hui's son, Guo Xiaozi, undertakes a quixotic quest to find his exiled father, only to be rebuffed by the man he believes to be his father; the reunion for which he had hoped is thus exposed as a pipe dream. The message of Wang Hui's story goes further still, implying that a too fervent devotion to the ideal of filial piety may paradoxically result in a rupturing of the family unit itself. Taken together, these stories argue a highly complex view of filiality: the family, Wu Jingzi seems to imply, may provide society with an indispensable organic ethical base, but once family relations, for whatever reason, are eroded or broken, filiality pursued in the interests of nostalgia as a disembodied ideal is apt to be more damaging than beneficial. In Wang's case, the death of his daughter not only leaves him with a profound personal grief but also inspires doubts about the beliefs to which he has so rigidly clung over the course of a lifetime. He turns away from his didactic writings to seek out Dr. Yu and his circle, hoping to find a new source of faith, nourished this time not in ethical abstractions, but through personal contact with a community of believers. In what is perhaps the paradigmatic scene of the final section of the novel, he visits the Tai Bo temple in Nanjing in chapter 48 and discovers it in ruins, the sacrificial instruments locked in a cupboard, and the placard with the program for the ceremony so covered with dust as to be illegible. The reader has not had to wait until this late point in the book, however, to discover the ceremony's failure to have a more lasting or broad-based impact on society. The inevitability of that failure is suggested to the reader even before the sacrifices to Tai Bo take place, in chapter 3 5, when the Second Master of Ceremonies, Zhuang Shaoguang, visits the capital. During Zhuang's imperial audience, the emperor himself acknowledges the need for a re-establishment of "ceremony and music" and invites Zhuang to outline his plan for the re-education of the people. But at the moment Zhuang would speak, the sting of a scorpion that has been hiding in his cap silences him. He later calls the scorpion "my Zang Cang," referring to a minister of the state of Lu who in a passage in Mencius dissuades his lord, Duke Ping, from meeting with the philosopher. In that passage, Zang Cang objects to Mencius on the grounds both of his common origins and of what Zang Cang interprets as a violation of the prescribed funeral rites. Mencius, however, refuses to credit Zang Cang with the power to prevent his meeting with Duke Ping: "My not finding in the prince of Lu a ruler who would confide in me is from Heaven." 45 Zhuang Shaoguang

274

RITUAL, MEMORY, AND DESIRE IN RULIN WAISHI

understands the scorpion to be making a similar point, that Heaven does not favor his serving as the emperor's personal adviser. 46 But the scene of Zhuang's audience with the emperor at least suggests the possibility of a different outcome for the idealistic undertaking of Du Shaoqing and his clique. If Zhuang could have persuaded the emperor to reinstate the ancient rites in the central court, thus seeing to it that they were established at the very center of public life, their influence might have then irradiated the concentric circles of the bureaucracy and produced-through a metonymic chain of effects-a fundamental transformation of the social order. The failure of Zhuang Shaoguang to make his case to the emperor, though credited to a higher wisdom, cannot but impress on the reader the highly circumscribed nature of the sacrifices to Tai Boas they are actually performed in the novel. Because of the marginal social position of Du Shaoqing's clique, the ceremony is necessarily reduced to a local performance, to a kind of private protest that can exert only a symbolic influence on society at large. As a result, the temple dedication is in many ways reminiscent of the memorial service for the last Ming emperor celebrated in scene 40 of Kong Shangren's play Tao hua shan (The peach blossom fan). In that scene, a group of exiled Ming loyalists meet to perform the service as their last act before retiring from the world into Daoist reclusion. They harbor no hope of any immediate social consequences resulting from the ceremony, but intend it to demonstrate their nationalism and their continued devotion to the values associated with the defeated dynasty. The echo of Kong Shangren's famous play in Rulin waishi raises the possibility that Zhang Shaoguang's scorpion is intended as a veiled reminder of the illegitimate Manchu origins of the dynasty in which Wu Jingzi wrote. This would add a layer of topical political allegory to the text that is impossible either to authenticate or to disprove. But whether or not Wu Jingzi intends a specific anti-Manchu message here, the scorpion's bite can only signify Zhuang's realization of the historical limitations of his clique's reformist undertaking. As Zhuang himself puts it, "It does not seem as if our way [dao] will avail in this age. " 47 Here we may return to the question of cultural criticism in Rutin waishi with which we began. As we have seen, the novel unquestionably dramatizes the desirability of a rejuvenation of classical Confucian rituals and seems thereby to reinvest the most ancient ideals of Chinese culture with a new moral appeal. But a careful examination of the novel's rhetorical operation shows that it never asserts those ideals in a simple polemical way; with its stubborn preference for metonymy and its careful attention

Marston Anderson

275

to the corrosive effects of time and history on the social world, the novel even seems to exhibit a reluctance to make of the rites a transcendent symbolic ideal. As is apparent from the discussion of filiality above, the novel both reaffirms the value of a fundamental Confucian virtue and questions the social consequences of its reification as abstract ideal. For the most part, Rulin waishi is concerned not with propounding a doctrinal or didactic interpretation of the world but with exploring the gap between moral intentionality and the operability of any kind of ideals in a world fatally subject to temporal and historical limitations. This is why seekers of ideological consistency in the novel are inevitably disappointed. The text's "lucidity" is not that of expository prose; rather, it is a purely narrative logic that takes as one of its missions the deconstruction of an unexamined idealism. Only if we remember this fact will we be able to understand the exigency behind Wu Jingzi's choice of the novel form: for if Xunzi's own ruminations on ritual reflect a period of increasing secularization in Chinese intellectual history, Rulin waishi emerges from a historical phase when a new need was felt to measure the ritual ideal against the most mundane realities of social life, and even against history itself. One cannot but feel on concluding Rutin waishi that history has the upper hand, and that it has tragically jettisoned those characters still possessed of a moral sensibility. This is surely the reason that the "positive" models so frequently cited as the moral ballast of the book display such consistently negative behavior. The four exemplary figures in the final chapter are aheroic figures, remarkable primarily for their willful assertion of personal authenticity; they seem resigned to the fact that although they may defy the world, they cannot change it, and they are rewarded for their stubbornness with continued creative vitality. 48 The fabled prototype of all these characters, the artist Wang Mian from chapter r, shows a similar mix of defiance and inspiration. For a time in his youth, he enjoys a period of bliss, significantly associated with a particular kind of creative activity: profoundly moved by his natural surroundings, he takes up painting them, with such skill that "his flowers seemed to be growing in the water, as if freshly plucked from the lake and placed on a scroll. " 49 It is precisely his painting, however, that gets him into trouble, attracting to his hut a variety of corrupt officials who wish to use his paintings to impress their superiors. To escape them, he flees to Ji'nan, where a revealing incident occurs. Still harassed by wealthy customers, Wang Mian in exasperation paints a picture of a big ox and pastes it up in his shop along with some satirical verses. 5° That even the proud Wang Mian resorts

276

RITUAL, MEMORY, AND DESIRE IN RULIN WAISHI

to such a method to fend off unwelcome societal pressures may represent a kind_ of justification for the satirical method of Rutin waishi itself: pressed to the wall and unable to flee, the wise man turns to satire as a weapon with which to resist the world. If this equation is valid, we may go further, and suggest that the "white" writing of chapter 37 is similarly Wu Jingzi's attempt to discover a pure, self-effacing method of literary representa. tion to match Wang Mian's achievement in his nature paintings. But the social world does not lend itself to such impartial representation, and Wu Jingzi's rhetorical experiment eventually proves contrary to the very spirit of the narrative form in which he chose to work. The most his narrative can offer is an impassioned retrojection of the world, a stance of principled resistance to the world's impurities. So, too, all expression of virtue in the corrupt and temporalized world of Rutin waishi is exposed in the end as a kind of pure intentionality. And it is this inoperative but still vital moral instinct that is memorialized in the ritual dedication of the temple to Tai Bo, a ceremony that earns its participants a treasured but momentary sense of timelessness and authenticity.

CHAPTER II

The Shattered Mirror: Wu Jianren and the Reflection of Strange Events Theodore Huters

f'.///u Jianren's

(r866-r9ro) Ershi nian muduzhi guai xianzhuang

flf/ (Strange events eyewitnessed over twenty years), which began to appear in print in r903, is the earliest sustained first-person narrative in Chinese vernacular literature. As such, it inevitably raises questions as to why this new mode of expression materialized on the scene when it did. Was it merely a response to the discovery that this type of narrative existed in the West? Or were there reasons it spoke to new discursive needs within the realm of Chinese letters? Although there seems no way to answer this question definitively, Strange Events does fit into a continuum of Qing dynasty fiction that deals with the problem of the writing subject. The sudden implementation of an unprecedented narrative mode, however, calls attention to itself in such a way as to move to the foreground and address in a new way notions of the capacities and the range of writing that had been in play since at least the middle of the dynasty. Central to this new dynamic was an atmosphere at once of crisis and of utopian hope created by the radical disruption of the intellectual world as it had existed prior to r89 5. Since the concatenation of literary revitalization and more general attempts at a transvaluation of discourses on culture that followed the defeat by Japan that year is not coincidental, I will argue that literature and intellectual change can be most productively understood when they are read against one another. From this perspective, the novel appears to be torn between contemplating the problematics of individual perspective on the one hand and disclosing the more lurid forms of social behavior on the other, now seen as threatening the very existence of the Chinese polity itself. At the core of the renewal project embodied by the introduction of a new type of narrative voice, however, lies an

278

WU JIANREN AND THE REFLECTION OF STRANGE EVENTS

effort to gain leverage on the social process by building a platform for speaking that lies outside cultural norms as they were perceived at the time. In the end, I will argue, this effort to find a speaking position outside current practices failed. The implicit revolution in the possibilities for writing that the novel represents did not take place in a vacuum. Much has been said in twentiethcentury China about the subordination of textual production to the concerns of a narrow political and intellectual orthodoxy under the late empire. At least from the mid-Qing on, however, there was much discussion of the act of composition itself and its relationship to officially sanctioned Confucianism. 1 Within the realm of narrative fiction alone prior to r 89 5, there seems to have been considerable space within which to discuss the problematic relationship between the act of writing and the normative discourses that had surrounded and situated writing since the institution of Daoxue orthodoxy in the Yuan dynasty. The extent of the space available for the interrogation of norms can be understood from the presentation of the figure of the stone in Cao Xueqin's mid-eighteenth-century novel Hong lou meng (Dream of the red chamber). On perceiving itself to have had no place to fit during the construction of the sky, the Stone considers itself to be alienated from the cosmic process. This alienation in turn affords it the opportunity for-or, rather, forces it toward-an aesthetic contemplation of virtually all the attributes of its relationship with the world. The comprehensiveness of the Stone's theorization of its relationship to its own act of writing that results is extraordinary: It is at once author of the text, the text itself, the subject of the story the text is telling, and the guardian of the ways in which it should be read. The world of the Stone, then, is a thoroughly textual one in which no aspect of the production of writing is left unproblematized. 2 It is also true, however, that the very elaborateness of the structure of this discussion indicates how delicate this subject was to broach, and how difficult it was to find a position within the discursive norms of the time from which to mount an effective critique of cultural practices. The novel Rutin waishi (The scholars), written slightly before Dream, also has much to say about the act of writing but tends to break down the possibilities in a more schematic way. Less intent upon gauging the potential of a particular voice, it maps out instead the constitutive features of the possibility of writing from a variety of social positions. One of the interests the author, Wu Jingzi (I?OI-54), shared with Cao Xueqin is measuring the possibility of writing that distances itself from what both

Theodore Huters

279

authors see as a corrupt and corrupting state discourse. We thus meet in The Scholars a number of people who demarcate themselves from the dominant ideology. One of the most vividly depicted is Qu Shenfu, referred to throughout as Young Master Qu (Qu gongsun), a youthful poetaster from a family distinguished for both scholarship and ethical behavior. The introduction of the Qu family immediately after the depiction of a series of mean men devoted to self-advancement through the examination system makes it seem all the more praiseworthy when we first encounter it. In chapters ro and r r of The Scholars, Qu gongsun is married to a certain Miss Lu, a young woman of great beauty who superficially seems to stand as the perfect jiaren (beauty) to Qu's caizi (talent). At this point, however, the irony begins. Qu, as we knew from the very point of his introduction into the text, is not unlike Jia Baoyu, the protagonist in Dream, an adept at poetry who considers anything to do with the examination system unbearably vulgar. Miss Lu, on the other hand, lacking brothers, has been rigorously trained in bagu, the "eight-legged essay" (the complicated rhetorical form required for the imperial civil service examinations), and considers any form of writing not associated with the examination trajectory to be utterly frivolous. Like most of the book, this episode is satirical, but it is a satire so evenhanded that it seems virtually impossible to assign a partisan role to the writing voice. Although this evenhandedness is notable (I will return to it below), for my present purposes what is striking about the episode is the way in which it sets forth two characteristic and complementary discursive positions open to writers in the post-Song empire. Miss Lu is a caricature, and a vivid one, of that most enduring of models open to writers when the set of writing practices signified by Zhou Dunyi's ( IOI?73) phrase wen yi zai dao (writing as the conveyance of the dao) dominated Chinese letters in the Daoxue era (the Ming and the Qing). For his part, Master Qu is her polar opposite-a man who tries to define himself as being other than an adherent of the state orthodoxy, a position ultimately traceable to SuShi (1037-IIOI) and his notion of a wenxue meant to be dominant in letters but particularly open-ended in its function. 3 It seems evident that Wu Jingzi's evenhandedness in presenting this binary is emblematic of a feeling that both positions were unsatisfactory, but that they also represented an insoluble problematic. From the modern perspective, the critique of Miss Lu's position is the most readily apprehensible. Twentieth-century intellectuals have been so firm in their condemnation of bagu as an instrument of intellectual suffocation that it is virtually impossible for the modern reader to see it as anything other

280

WU JIANREN AND THE REFLECTION OF STRANGE EVENTS

than the vulgar business that Master Qu holds it to be. But neither does the novel's account of Master Qu leave out the considerable economic and psychic costs to those who turn their backs on this most official of genres. What strikes one as most significant about the episode, however, is the extent to which it invests in the critique of the wenren, or aesthetic writing, alternative. This may have been simply a result of the fact that further satire of bagu would have been too cheap and unworthy of Wu Jingzi's talents, since criticism of the eight-legged essay had been a staple of literati discourse since at least the early Qing. 4 The fact remains, however, that the greater amount of detail devoted to lampooning Master Qu and his relatives, the Lou brothers, imparts a notably ludicrous cast to the field of non-official writing they represent. On first consideration, this denigration of the wemen tradition seems surprising-Why should an author so obviously writing outside the realm of official sanction not create for himself a refuge by privileging the tradition that seems to lie closest to him? 5 A closer look at Master Qu's attributes, however, begins to clarify the matter. After hearing about his family's probity, one of the first and most important things we are told about him is that his climb to eminence resulted from passing off as his own editorial work a treatise on poetics by the Ming poet Gao Qi that had come into his hands in a rather shady fashion. 6 This would seem a broad hint that the writing voice considers members of the wenren or mingshi traditions not to be what they seem. This episode taints them with the suspicion that their relationship with texts is based on plagiarizing their predecessors and that the refinement (ya) they consider their defining characteristic is merely a marker for the degree to which they cover up this unsavory fact. Perhaps even more to the point, however, is the way that Qu and his friends define taste and creativity as functions of perpetual marginalization or even straightforward superficiality. From the perspective of the fine minds of these conscientious objectors to government-sanctioned prose, in other words, to aspire to anything distinguishable as an idea is already to be thoroughly penetrated by vulgarity. Given this presentation, we can readily see why Wu Jingzi would want to avoid either condition. But the question raised by the text and continually deferred by it is whether it is possible to avoid falling into one condition or the other. On the one hand, the revealing fact that Wu chose to depict Miss Lu and Master Qu as joined in a marriage frustrating to both suggests that the two options are meant to embody the sort of claustrophobic binary opposition so thoroughly analyzed in the Western critical theory

Theodore Huters

28!

of the past ten or fifteen years. On the other hand, Wu's very act of recognizing the ineluctability of the relationship between the two modes of writing and that there really are no other choices represents-minimal though it may be-a space for critical alterity in what seems at first glance to be a field barren of possibility. 7 I would argue, in other words, that the self-consciousness Wu gained through describing the impasse that writers face provided a crucial critical space. It is by its very nature aporetic, since he portrayed a literary arena in which there was no place for the very platform he was writing from. The space is there by virtue of Wu's writing it into existence, but only if writer and reader alike are willing to battle the odds to make the effort to create it. Wu, however, did not attempt to write from a position outside the norms of his time. His is an inside job, a stance that occasions the novel's extraordinary subtlety. During the course of the nineteenth century, old ways of thinking about society came to be challenged by increasing political demands upon China from the West. The long process of contestation of the unceasing demands of the foreign powers for China to open itself ever further to the rest of the world culminated in a crisis in the years after 1895· 8 With the defeat by Japan in that year, many of the pressures for reform long perceived as the inappropriate demands of a rapacious West suddenly come to be defined as part of a pressing internal agenda. Among other things, the prospect for fundamental social and intellectual reform was reconceptualized as lying well within reach. 9 The new sense of hope that opened among the elite following this redefinition of the possible was invested to a significant degree in the notion that writing-long the site of moral and political contestation-was the arena of possibility through which new and utopian discourses could be realized. Accompanyingand, in effect, enabling-this utopianism was a thoroughgoing reassessment of the customary ways in which writing was produced and received that had profound consequences across the literary spectrum. And key among the utopian ways of thinking about literature was that it would provide a platform outside prevailing norms from which to launch a critique of them. While this general recoding of discourses had so many facets that rendering a detailed account is impossible here, it is perhaps safe to say that one of the most notable reconceptualizations of writing concerned the related demands that it be at once more socially relevant and reach a wider audience. The so-called revolution in the realm of poetry (shijie geming) and its best-known advocate, Huang Zunxian, for instance, clearly attempted to meet these two demands. 10 In his determination to broach

282

WU JIANREN AND THE REFLECTION OF STRANGE EVENTS

issues of politics and technology, Huang quite self-consciously sought to broaden both the scope and reach of the lyric genre. The novel, however, was the obvious candidate for appropriation as a tool to reach a broader audience, not only because of the historical view that it was the genre that spoke to the elite about popular concerns, but because it had, in fact, enjoyed wide popularity throughout the nineteenth century. One practical consequence of the resulting pressure for relevance in fiction was that novels written after r895 (with the possible exception of Liu E's Lao Can youji [The Travels of Lao Can]) represent a coarsening of the refined argumentation that had characterized earlier "literati novels" such as Dream of the Red Chamber and The Scholars. 11 As part of a perception that the scope of civic discourse needed to be widened, the presumed simplicity of the xiaoshuo (a term roughly synonymous with "fictional narrative") vernacular was recontextualized in these years; it was no longer seen as a drawback but as the genre's great advantage-its ostensible legibility became its great virtue.U The concomitant augmentation of xiaoshuo's responsibilities was made possible by an environment in which previously stable hierarchies of literary genres were suddenly reconceived as being contingent on particular historical circumstances and thus open to wide-ranging change. One of the principal justifications of the need for a wider readership, for instance, arose out of a particular Chinese reading of Japanese and Western modernization as being the result of increased civic participation facilitated in large part by popular consumption of socially progressive novels. 13 It need hardly be added, however, that such a drastic recasting of discursive space also implied a more or less desperate search for a new intellectual order to replace the old one. But this new enthusiasm for the novel contained an important caveat, a caveat that was to be mirrored almost precisely in the May Fourth discourse on the novel some twenty years later. This was to create a new distinction within the genre of the novel itself: the traditional (i.e., Chinese) xiaoshuo was retrograde, and a new ideal form based on Western and Japanese models was explicitly called for and established as a new object of emulation. The external origins of the new novel provided it with its leverage. The only thing the traditional novel seems to be good for in Liang Qichao's influential early (r898) "Preface to the Printing of [the series of] Political Novels in Translation" ("Yi yin zhengzhi xiaoshuo xu") is to reveal that only the vulgar avenue of fiction will reach out beyond the small band of men dedicated to high seriousness: 14

Theodore Huters Xiaoshuo in China, although it is listed in the traditional bibliographic categories [of the Hanshu ], has had few good works after the Yu, chu [the work regarded as having initiated the category]. 15 When [a writer) wished to write of heroism, he took Shui hu zhuan [The water margin] as his model, when he spoke of the relationships between men and women, he followed Dream of the Red Chamber. In general there was nothing that did not fit into the two categories of inciting robbery and inciting lust. All these [works] followed one another and stuck with one another [in their perpetuation of the harmful]. Therefore, presentable writers disdained [the genre]. Although this is so, the fact is that human nature hates the serious and likes frivolity.... So there is nothing to be done about it when scholars take time from their learning and take up Dream of the Red Chamber and talk of The Water Margin. Therefore, to try and forbid [fiction] is not as good as trying to make it better. What Mr. Nanhai [Kang Youwei] said is correct: "Among those who are barely literate, there are those who do not read the Classics, but none who do not read novels. So if you cannot teach them the Six Classics, one should use novels to teach them. If the standard histories make no impact, novels should be used. If the colloquies [of the Song thinkers] cannot be used to inform people, novels should be used. If law cannot control people, then novels should be used for control. People of complete knowledge are scarce in this world, while the ignorant are many. People with thorough grounding in the humanities are few, while those with a crude knowledge are many." 16

This total condemnation of the Chinese novel as it existed up to Liang's time is striking; his choice of the two most prominent works in the genre as implicit negative influences would seem to allow for no exceptions. That he ends the piece with a ringing affirmation of the powers of a specifically political novel in the West-"It often happened that upon the appearance of book a whole nation would change its views on current affairs. The political novel has been most instrumental in making the governments of [the West and Japan] daily more progressive or enlightened"17-does nothing to close the gap between his hopes for the form in general and the low regard he has for every novel ever written in Chinese. By thus cutting off any practical appeal to native precedent, Liang's qdl puts up a huge barrier between what he advocated and the fictional languages available to writers at the time, a discursive position that was to become commonplace in the twentieth century, from May Fourth to Qu Qiubai and on to the Chinese critical arena as it existed in the forty or so years after Mao Zedong's proclamations on literature at Yan'an in 1942. 18 But Liang's steadfast vision of the novel as an instrument of reform is probably the most remarkable thing about this essay. To cite only one

284

WU JIANREN AND THE REFLECTION OF STRANGE EVENTS

instance, the contrast between Liang's positivism about the novel and the tentative relationship of Cao Xueqin's Stone to his own narrative could not be more striking. It is clearly the external origins of the new novel that allows him to indulge his extreme notions of the instrumentality of fiction. Critics have long wondered exactly where within the European or Japanese discourse on the novel Yan Fu, Liang Qichao, Xia Zengyou, and their contemporaries found the idea of the novel as the key to successful reform. 19 Perhaps the most satisfactory answer is in fact the "nowhere" that they have long suspected. Liang's influential genealogy may be based on a negation of the old view of the novel, with which, after all, he begins his essay. Liang's censure of the traditional Chinese novel is founded on a view of its capacity to produce certain undesirable sorts of human behavior. In the new era of possibility, however, narrative retains its old power, but it is now and only now capable of being transposed into a positive key. In other words, the core assumption of the old discourse on the power of the novel has not changed, merely some of its internal relationships. But there was a cost to this reformulation-the rejection of any actual Chinese novel and the theorization of fiction as a transcendent category. Di Baoxian, the influential advocate of fiction and Liang's contemporary, further complicated the issue of how fiction is to be created in a 1903 essay on fiction published in Xin xiaoshuo. Di called for an end to the frivolity that he saw as endemic to all writing in the classical language and for a general turn to the utilitarianism that he saw as inhering in vernacular styles: 20 If literature [wen] is to be taken as having a function, it must not be taken as a plaything, but as a staple [shusu]. Once an epigrapher [jinshi jia] gave a banquet and brought out his Shang yi, Xia ding, Zhou dui, and Han jue to use as containers for food and wine. The result was that both host and guests ended up with diarrhea for the week. Beauty is beauty, but what if it is not appropriate? Therefore, the succession of the vernacular style is really an unavoidable consequence of the selection of the superior.

Liang's implicit call to transcend prior practice is left intact here, much as the equally insistent instrumentality of the earlier text is also strongly reinforced. This call for stripping language of its refinement in the writing of xiaoshuo, however, must have left the would-be novel writer in a quandary. Told on the one hand in fairly withering terms that fiction was of a markedly lower order than other writing and that it would have to improve, yet on the other hand that this very lack of aesthetic affect was the very essence of the genre, how was the author to respond?

Theodore Huters

One of the ways these two demands were mediated was to establish a genealogy of fiction that stressed hidden meaning as the core of the traditional form. 21 Thus the low quality of premodern fiction and its irrelevance to today could be ascribed to the need for authors to protect themselves against persecution by the state for the political dissidence prescribed for the form by late Qing critics. But although this theory allowed the novel to fit the new shape demanded by the reformist critics, it had little to offer writers who no longer had to obscure the called-for political side to their work. This conjuring of a hidden force within the traditional novel would ultimately contribute to a heightening of anxiety about the present potential of the novel on the part of contemporary authors by mystifying the origins of the form and creating a notion of a secret power inaccessible to those who lived in a different discursive space. The utopian side of this discourse lodged itself in a series of optimistic pronouncements about the potential of fiction to bring about a brave new worldY As Leo Lee and Andrew Nathan have pointed out, however, the actual narrative work produced either in response to these calls or simultaneous with them presented a far bleaker sense of the possibilities. 23 Given the contradictory elements in the advocacy of a new fiction, it is not surprising that the gap between theory and practice worked itself out into this particular division of labor. Based on their foreign sources of inspiration, the critics found it easy to position themselves outside the social maelstrom and transcend local contradictions. The novels, on the other hand, could do little to avoid becoming caught up in those very contradictory details. For our purposes, what is most significant about this discrepancy between theory and practice is the manifold pressures it exerted on authors. They ended up writing in a genre sponsored by a group of reformers who set contradictory demands that by definition could never be satisfied. Since the only thing many of these demands had in common was a predisposition to accord to the novel enormous influence in the abstract, the tensions surrounding the composition of actual texts only became more acute. Strange Events meets these political demands by offering what appears at least on the surface to be a fairly straightforward denunciation of corruption in politics and society. Wu Jianren (also known as Wu Woyao), a Cantonese from a family with a long tradition of examination success, had been born in Beijing while his father was serving at a post there. Although he avoided an official career, he involved himself in a variety of cultural activities characteristic of the reformers of his generation. He had, for instance, a long affiliation with the literary side of the new institution

286

WU JIANREN AND THE REFLECTION OF STRANGE EVENTS

of journalism and was apparently affected by Liang Qichao's calls for a new novel. Accordingly, he began to write Strange Events in 1903 and by 1905 had serialized the first 45 chapters in Liang's fiction journal, Xin xiaoshuo (New fiction), which was published in Yokohama. Altogether, he wrote some thirty novels and short stories during his relatively short life. The novel was published in installments in a number of different forums over the period from 1903 to 1911, when the final installment (chapters 9 5-1o8) was issued posthumously. 24 Wu continued in publishing for a time, editing the fiction magazine Yueyue xiaoshuo (Monthly fiction), for instance, in 1906. During the final years of his life, he was engaged in educational work in Shanghai. Strange Events is Wu's longest and most famous work, and one of the landmarks of the late Qing novel. It presents a thoroughly dismal picture of contemporary life among the upper classes, with a particular focus on the higher rungs of the non-metropolitan bureaucracy. Many of the men and incidents depicted in the text are thinly disguised versions of actual people and the gossip about them. The twenty years alluded to in the title refer to the period beginning about r884, the year of China's war with France; the two decades that saw the crises of the war with Japan, the struggle on the part of the imperialist powers for great concessions in China in 1897-98, the abortive "Hundred Days reform" of 1898, and the Boxer Rebellion and the subsequent occupation of the capital by foreign troops in 1900. 25 From the viewpoint of the first few years of this century, then, the twenty years after r884 could be seen from almost any perspective as uniquely disastrous for China. The novel portrays a world gripped by a virtually universal corruption, in which everything turns up for sale, sooner or later. Behind this lies a terrible breakdown of Confucian values, particularly those having to do with the family. The moral squalor of this period is represented through a large number of episodes; the more elaborate ones are personally witnessed by the narrator, and the many shorter stories are told to the narrator by people he encounters in the course of his travels through a China shrunk by the network of steamships and telegraph lines, part of the material legacy of the coming of the West. Most of these stories are quite self-consciously lurid, reflecting the narrator's view that the human world that he has encountered contains only "pests, beasts and demons." 26 However, within the rather gothic tone the narrative assumes, it portrays events with a great deal of power. There is a mixture of narrative timbre, from near-burlesque to pure horror to carefully worked out thriller to heartfelt sympathy. There is even, from time to time, the presentation of

Theodore Huters

exemplary characters. The carefully observed psychological insights and fastidious attention to anecdotal detail that characterize the stories combine with this adroit manipulation of tone to make for what is on the whole a most compelling work. There has, however, been a persistent unease in the critical evaluation of this text within China. For instance, the contrast between my assessment of Strange Events and that of Lu Xun written sometime between 1920 and 1923 could hardly be more extreme. Lu Xun's judgment of this particular novel is only part of a critique of late Qing novels as a whole: Although they were intent upon reforming the age and thus seem to be in the same category as novels of satire [such as The Scholars], their expression is superficial and their biting style is without any subtlety. More than that, however, they are full of exaggeration in order tp accord with the predilections of the time. There is, then, a marked poverty to their skill and to their degree of tolerance. I have, therefore, put them in a special category that I call "novels of censure. " 27

His judgment of Strange Events is, if anything, even harsher. After initially praising the wide range of characters Wu Jianren included in his work and the author's unbending disposition toward a critical perspective, Lu Xun renders the judgment: "Unfortunately, his descriptions are too hurried and he sometimes errs by overstating the evils he is depicting. His words disregard reality and their power to affect people is thereby diminished. The result is merely a compendium of gossip, good only for providing those at loose ends with material for their idle chatter. " 28 Writing in the years after 1920, it was easy for Lu Xun to regard the late Qing novel as already part of the old order. On one level, then, his disdain demonstrates the extent to which his thoughts on the novel were of a piece with those of the post-1895 generation in regarding all extant Chinese novels as unworthy. He did, after all, participate in the discourse on fiction of that earlier time, however much he retrospectively thought himself apart from it. 29 But why did he single these texts out for condemnation as the finale to a larger project whose goal was to recuperate the tradition of Chinese fictional narrative? At this point it would be premature to assay an answer to this question; it must await discussion of the text itself. Whatever the reason for Lu Xun's harsh words, however, they have stood as canonical ever since he uttered them, and critics feel compelled to cite them approvingly even if they are in effect engaged in a project to reverse his verdicts. 30 In other words, the later critical evaluation of the late Qing novel has never really moved beyond the condemnation of the Chinese novel as a whole voiced by Liang Qichao as early as 1898.

288

WU JIANREN AND THE REFLECTION OF STRANGE EVENTS

The strange or, perhaps more accurately, horrible events that transpire in the novel are presented to the reader from the perspective of a young man, Jiusi yisheng-literally, "Nine deaths, one life," or, less portentously, "Still alive despite multiple perils." Jiusi began keeping a diary from the age of fifteen, the year his father died. He does not present the diary to us, however. That act is performed by one Sili taosheng, "Escapee from the jaws of death," who had procured it from a man on the street in Shanghai who was touting the work as being more instructive than a shanshu, or book of morality. Since Wu Jianren announced in an essay published at the beginning of 1904-just at the time, in other words, when Strange Events was beginning to appear in print-that he had read several hundred foreign novels written in English and French, the only two foreign languages he professed to know, 31 Wu's account of his reading history renders it virtually impossible to account for the origin of the motifs he deploys in his narrative. It could be, for instance, that the motif of "discovering" a diary is either a borrowing from the Western novel or a kind of secularization of the mystical process of discovering the narrativecovered stone at the beginning of Dream, or some combination of the two. Whatever its provenance, the stance the novel employs as a diary found by obscure means serves a fairly conventional purpose in boosting its claim to greater actuality. What the diary form also does, however, is place within the novel an observer who is self-consciously outside the events he is recording. The composer of the diary thereby serves as the emblem of an external point of view on domestic events that had inspired so much enthusiasm for the novel form in Liang Qichao. When we first encounter the author of the diary, we are almost immediately given signals of his immaturity and general incapacity to make the correct decision when confronted with a set of reasonable alternatives. His inexperience, his newness to the scene, is thus set up as the novel's first theme. This unfortunate circumstance is conveyed in the text at the beginning of Jiusi's self-presentation, when the young man arrives in Hangchow only to find that his father has died before the son was able to see him one final time. The death of the father immediately presents the problem of the disposition of his assets, and Jiusi must choose among three adult men who present themselves as plausible candidates. To complicate his choice, they all are quite frank in explaining to him why the others cannot be trusted. Jiusi chooses precisely the two who are unreliable, one of whom is his uncle. Jiusi's father's trusted shop assistant had warned the young man against his uncle, but he decides in proper Confucian usage on the basis of the family connection.

Theodore Huters

The journey to see the sick father, the father's death, and the distribution of the money take place in the second chapter, which is the first chapter of the diary of Jiusi yisheng. The men to whom Jiusi entrusts his money are both given reasonably large sums and within a very few pages the young man and the reader alike learn that the money has, at least for the time being, disappeared. Although this fact provides the reader with enough clues to begin to doubt the trustworthiness of the uncle, Jiusi himself admits to no inkling of a problem. Moreover, at the end of the same chapter, Jiusi refuses to believe that a man he encounters during a sea voyage dressed as an official is in fact a thief and must be shown how this is the case in considerable detail before he will credit the accusation. The focus of all the significant events of this chapter32 on the narrator's inability to adequately judge his own perceptions establishes a pattern that persists for the first segment of the book (chapters 2-18). Once the pattern is established, Jiusi continues to see events as happening in a world that he can make no sense of. He fails to reach appropriate conclusions about events that he sees and must be told the meaning of anything that happens by a more experienced person. Sometimes, as with his failure to understand his uncle's actions correctly, he seems almost to actively resist seeing even when the evidence for making a judgment is presented in so many iterations that no one inside or outside the text other than Jiusi fails to see the actual intent of the uncle's actions. Gradually, however, through a slow process of finding and learning from competent authority, Jiusi becomes able at least to make observations that help him deal with situations in a more efficacious manner. The key authority figure here is one Wu Zeng, known by his cognomen of Jizhi, who is introduced in chapter 3· Jiusi encounters Jizhi by chance in Nanking, when they recognize one other as having been students at the same school. Jizhi, ten years Jiusi's senior, has gone on to win his jinshi degree and now has an official post in Nanking. The older man first invites Jiusi to move into his household and by chapter 4 has offered him a position in his private secretariat (mufu). Both of these offers come in the nick of time, Jiusi having exhausted what resources he has in a fruitless attempt to gain audience with his uncle so as to inquire what has become of the substantial capital that Jiusi entrusted to him. An emblematic episode that shows not only Jiusi's ignorance andJizhi's role in educating him but also introduces another important figure in the novel begins at the end of chapter 4· On one of the first days Jiusi is on his way to work at Jizhi's office, he happens to walk by the Nanking establishment of one Gou Cai, a Manchu of high rank (whose name is a pun

290

WU JIANREN AND THE REFLECTION OF STRANGE EVENTS

on a phrase meaning "dog-like nature"). Jiusi is extremely impressed by the marks of status and wealth that are on display, the elegant clothes and the many attendants. Upon arrival at Jizhi's office, Jiusi remarks on what he has seen, but Jizhi dismisses him with a rather contemptuous, "You're always so easily impressed!" (pp. 29-30) and turns the conversation to other matters in a deliberate attempt to avoid talking about Gou Cai. It is only later, when Jizhi and Jiusi are alone, that Jizhi tells him the actual circumstances of Gou Cai, something he had been unwilling to do with other people present. Jizhi informs Jiusi that Gou Cai, while once a high official, had alienated his superiors and is now mired in poverty, something Jiusi finds impossible to believe: I said, "You must have spoken in error. When I saw him this morning escorting his guest to the gate, not only was everyone in brand new clothes, but he had four or five men in attendance. How could he be in poverty?" Jizhi said with a smile, "You haven't been around very long, so how can you know what you're talking about? Manchus are masters at putting up a false front; no matter how poor they are, they'll still put up a front .... So when you consider that he was once a circuit intendant, how could he not come up with new clothes and attendants? As for the clothes, he clearly went to some trouble to conjure them up, but do you know about his 'servants?' When guests come they are servants, but when there are no guests around they eat at the table with the family." I asked, "Why is that?" Jizhi said, "The reason is that they are all his nephews and cousins and that sort of person, who rushed to him when they heard that he had become an official. None of them expected him to get poor, such that they would be obliged to act as his servants to keep up appearances." (pp. 43, 47) 33

For all the delicacy of the "Manchu question" during the period this novel was being written, it is clear that this presentation places more emphasis on the absurdity of Gou Cai (and of Manchus in general) than on the urgent depiction of the national question as such. 34 A turning point in the novel, however, occurs in chapters r8-2o, when Jiusi returns from Nanking to his home in the south in response to an urgent family telegram reporting that his mother is seriously ill. Upon arriving, he finds that his mother has not been sick and that she knows nothing of the telegram. She does, however, hint at family plots designed to swindle them. This, combined with Jiusi's own observations about the sinister environment at home, causes him to decide to dispose of the family property and take his mother and the proceeds from the sale to Nanking. He does this successfully in spite of the machinations of various conniving family members and acquaintances, including one of the men who had

Theodore Huters

earlier embezzled money from him at Hangchow soon after his father's death. 35 For all the good judgmentJiusi demonstrates in this episode, however, he still relies on others for information and advice. In this case, a relative of his Nanking patron Jizhi first provides him with the information ·On land prices he needs in order to resist his family's absurdly low offer, and then buys Jiusi's land. So while Jiusi does not fail to figure out the correct course of action, he would still have not been able to act without crucial support from the Wu family. It is not until chapters 32-3 5 that Jiusi actually encounters a problem and works out the solution completely on his own. 36 The long episode takes place in Shanghai and begins with Jiusi encountering one Li Jingyi, a former clerk in Jiusi's father's store. Before he describes Li Jingyi in detail, Jiusi first notes that he learned subsequently that Li was guilty of a series of horrible acts, thereby signaling the reader that he is not under any illusions about what he is dealing with this time. We soon learn that Li was complicit in the forced suicide of his own brother and that he then sold his newly widowed sister-in-law into prostitution. The narrative that follows recounts Jiusi's successful efforts to find the young woman, Qiuju, reunite her with her "step" (gan) parents, and oblige Li to release his financial claims on her. As part of this process, Jiusi meets her "stepfather," Cai Liisheng, and discovers him to be an upright scholar who has been unable to do well in the cut-throat world of private secretaries in Shanghai. 37 After solving Cai's problem, Jiusi in effect becomes his patron and finds him a job within Jizhi's establishment. In effect, Jiusi's deft problem solving for Cai has enabled him to assume the same patronage position vis-a-vis someone else that Jizhi had always played for him. With this ultimate indication of learning and maturation, Jiusi's diary would seem to fit the pattern of the standard account of a young man's education and intellectual growth, a form familiar in the West as the Bildungsroman. The text to this point can be interpreted as being consonant with this genre. Although the narration of Jiusi's personal and family life had been somewhat sketchy before this episode, we do see him pursuing his uncle and talking at length to his mother and his cousin about matters of real consequence, conspicuously including questions of how Confucianism was to be applied in these parlous times. And the text does, of course, constantly foreground questions dealing with the broader issue of how Jiusi personally will cope with the difficult task of gauging events. At this point in the novel, Jiusi takes on a significant new role. Almost simultaneous withJiusi's rescue of Cao Liisheng,Jizhi decides to set himself

292

WU JIANREN AND THE REFLECTION OF STRANGE EVENTS

up as a merchant, with a chain of shops in Yangzi River ports. Since Jizhi is serving as an official, it is illegal for him to engage in trade, and he asks Jiusi to manage the enterprise in his stead (chapter 40). 38 Since this job entails extensive travel, it serves as the device by which Jiusi gains the opportunity to visit other parts of China and, most important, to collect more stories. Having by this time achieved what might be called epistemological maturity, Jiusi is in a position to evaluate the significance of each story. The emblem of this ability to collect, to move between, and to follow up on stories is the modern steamship and telegraphy systems brought to China by the imperial powers. Jiusi moves about by steamship and communicates by telegram. His means of communication figure so prominently in so many episodes that it obliges the reader to assume that thematic significance has been invested in them. 39 Once Jiusi takes this job and begins his travels, however, the narrative of his personal life diminishes significantly. When he achieves a certain level of knowledge, details of his own life recede to the margins. A position of wisdom in the novel, in other words, is plainly identified with externality to events. To a certain extent this can be explained by the fact that his separation from close friends and relatives also separates him from his personal life. But this explanation comes up wanting in chapter 6 5, in which Jiusi returns to his old home with his mother and cousin for the purpose of his marriage. Whereas in the earlier chapters of the book we had long glimpses of the inner quarters of Jiusi and Jizhi's households, including long episodes in which Jizhi's wife appears, Jiusi's own marriage is dealt with tersely. We are told he will marry, he is teased a bit by his female relatives, and he spends three years at home after the marriage. At the end of this time, he leaves at Jizhi's call and does not return for the remainder of the novel. About his wife or any detail of the three years spent with her, nothing is written. 40 Jiusi's personal abstraction from the events he relates in the second half of the novel is indicated in a striking disclosure in chapter 6o. Jiusi returns from a trip to Canton and gives his diary to his friends to read. They read about events the reader has just seen in chapters 57 and 59, but one of the friends notices that a tag line in the text apparently refers to an episode not recorded in the diary. Jiusi confesses that this is indeed the case and that he had written down the tag in haste so as to remind himself to write up the episode when he had more time (p. 476). What is significant about this event is the confession to us readers that when we thought we were reading about Jiusi's "real" journey to the south and everything of significance about it, we were in fact reading only what was

Theodore Huters

293

set down in the diary. Calling attention to the tag line reminds us that we are reading only what has been chosen for us (for whatever reason) by a particular process of text production that must be kept analytically distinct from the actual account of Jiusi's own travels. This episode indicates that the author of the diary is ultimately shown to have almost complete freedom to build his text from events of his choosing. There could be no firmer sign of Jiusi's self-chosen exteriority. What Jiusi's bildung has gained for him, in other words, is access to more stories, and his function in the text after he gains perceptual maturity is reduced in effect to collector of tales. To be sure, embedded within these narratives is a vast array of detail about life and habits, the practical ideology of the governing classes, and how that ideology relates to daily life. The dramatic quality of the episodes also seems to become more powerful in the later chapters, as the level of depravity increases markedly. And each episode remains loyal to the theme announced by Sili taosheng in the very first chapter: "I had known that Shanghai was no good, but according to this [book], although the world is wide, it is as if there is no place that can accommodate one" (p. 3 ). But the trajectory of the narrative provided by the focus on Jiusi's education in the early chapters seems to disappear, and his new intellectual control proves inadequate to the task of providing a new sense of order to the events that he observes. There are two characters whose reappearance at various points in the novel serves as a moral gauge on events. These are Jiusi's reprobate uncle and the Manchu official Gou CaiY The growing outrageousness of their behavior and the behavior that surrounds them signals the precipitous descent into moral chaos that increasingly marks the second half of the book. The reader and even Jiusi have by the later chapters long recognized the uncle's character flaws, but in chapter 82 Jiusi ferrets out the information that his uncle had taken his wife's niece as a mistress, an action representing a whole new level of depravity to the character. This transgression involves an even greater violation of family norms than anything the uncle has done before. It is thus part of the demonstration in the final half of Strange Events that Confucian values are being eroded at their very core. The decline of the moral order is even more powerfully represented in the later depiction of Gou Cai. As described above, when Jiusi first encounters him in the early chapters, Gou serves both as one of the prime emblems of Jiusi's inability to discern the difference between appearance and the true order of things and as an object of ridicule. He is still primarily an object of ridicule in his next significant appearance, in chapter 44,

294

WU JIANREN AND THE REFLECTION OF STRANGE EVENTS

where he is invited to a party given by Jizhi. Gou Cai brings his concubine instead of his wife, hardly a momentous act, but he allows her to dress as if she were his formal wife. When his wife hears of this, she crashes the party to visit physical mayhem upon both Gou Cai and the concubine, an event portrayed even more broadly than are the early episodes involving Gou. 42 While the emphasis in this scene is on Gou Cai's absurdity, the Manchu official's act of moral transgression in jumbling family rules is of greater consequence than had been his earlier show of simply putting on airs. The level of transgression increases as the novel continues. Gou Cai next enters the narrative at the end of chapter 86, where his story becomes the principal focus of the novel for eight of the next ten chapters. 43 The core story in this sequence centers around Gou Cai's gift of his daughterin-law (the young widow of his recently deceased son) to a viceroy, one of whose concubines has just died. Gou does this in order to get back into official favor after having been dismissed from office in a scandal involving many corrupt officials. In carrying out this outrageous act against both their own family honor and all the rules of Confucian rectitude, Gou Cai and his wife transcend simple buffoonery. The two of them are still the same absurd people, but their urgent persuasion of their daughterin-law to accede to their wishes puts them at the wrong end of a very important moral register in which the stakes have suddenly become very serious. At this point the tone of the text takes on a new gravity, as we see characters from whom we had always expected clownish behavior put into the most sober sort of situation, a context that is extremely effective in conveying the queasy feeling of an encroaching moral chaos. Like the episode in which we learn that Jiusi's uncle has compromised the honor of a young female member of his family, Gou Cai's actions here contribute to the complication of tone that the novel as a whole is undergoing. Although Gou Cai's act gains him a temporary official reprieve, eventually a new viceroy takes office, and he is horrified at Gou Cai's violation of norms and dismisses him yet again. Gou Cai later schemes to get rich by literally building and operating q mint, and when the governor of Anhui agrees to sponsor the project, Gou's plan is put into operation. Although he eventually loses this job, too, Gou has in the intervening two years of managing the mint made so much money that he has been able to take "five or six concubines." After a few years more on the official circuit, Gou Cai decides to retire to Shanghai with his loot. He has acquired a minor sort of heart condition along with his money, and one of his motives in moving to Shanghai is to seek expert medical advice. Upon arriving in

Theodore Huters

295

that city, he visits Jizhi to get a reference to a capable physician. The physician that Jizhi recommends, Wang Duanfu, diagnoses Gou Cai's illness as minor, but he also notices that Gou continues to manifest symptoms. Wang concludes that someone in Gou Cai's household is trying to do Gou harm, and he decides to abandon the case lest he eventually be blamed for what he forecasts will be Gou's untimely demise. While both Jizhi and Jiusi recognize Gou Cai as a person who is thoroughly morally compromised, they agree that only a family member greedy for Gou's money would try to poison him. Jizhi visits the Gou household and sets a trap, telling Gou Cai's only surviving son, Gou Longguang, that abalone is the one thing that will do great damage to his father's health. A week later Jizhi and Jiusi invite Gou Cai to dinner, along with Wang Duanfu. When abalone is served and Gou Cai refuses it because he says he has been having it served to him every day at home, Jizhi and Jiusi realize that they are witnessing an attempted patricide. Jizhi tries indirectly to warn Gou Caito leave home or to send his son away, but Gou does not catch the hints. After Gou departs, Jizhi informs Wang Duanfu of his trap and what he learned from it, and the doctor, who had just delivered himself of a long after-dinner speech on the wisdom of separating parents and children, especially only sons, responds: "Isn't that it! So I wasn't doing an injustice to anyone when I said what I did. But if you already had such firm evidence, why didn't you tell him directly just now, instead of beating around the bush as you did? To look at Gou Cai in action, you'd think he knew what he's doing, but in fact he's pretty ordinary in his apprehension of things, so there is no reason to assume he'd get what you were hinting at." Jizhi said: "But if I had told him directly, it would have done damage to the relationship between father and son." Duanfu was furious, and said: "Enough of that! Don't you think the 'relationship between father and son' will be even more damaged if you don't tell him directly?" Jizhi suddenly saw the truth to this and said: "You're right. I'll go see him tomorrow and tell him plainly." (p. 833)

Perhaps the most striking thing about the encounter between Wang Duanfu and Jizhi is the contrast it reveals between Jizhi's ordinary Confucian tact concerning family propriety and its inability to encompass the outrage that is being committed. The ineffectiveness of morality in the face of this greatest of all outrages against Chinese values is striking, a crime explicitly linked in Chinese thought with utter decadence in political morality. 44 The offense is so great, in fact, that even the wretched Gou Cai wins our sympathy and Wang Duanfu's quite scandalous speech on

296

WU JIANREN AND THE REFLECTION OF STRANGE EVENTS

the value of breaking up families passes by unnoticed as a transparent statement of a situational truth. For all the horror of these final episodes, however, Jiusi maintains a relative distance from events-he continues only to observe and report (note that it is Jizhi who makes the effort to intervene on Gou's behalf). This distance also keeps events at one remove from the reader, if only because we are constantly being reminded that they are just stories. This episode becomes more ominous. When Jizhi visits Gou Cai the next day to tell what he has learned, he is sent away after being told that Gou Cai is too sick to see anyone. The next day he goes again and is told the same thing, and his many subsequent efforts to gain audience with Gou are unavailing. A few months later, he not unexpectedly receives word of Gou Cai's death. Jizhi eventually learns the full details of the plot, which he duly communicates to Jiusi. It turns out that Gou Cai's son had not been able to make a good marriage because of the family's tarnished image, and Longguang does not like the wife he is married to (although he does get on very well with her brother). When Gou Cai rejects his request for a concubine, Longguang and his brother-in-law take up with two of Gou Cai's concubines. When Gou Cai returns to Shanghai, they thus have a number of reasons to plot against the older man. They enlist the help of a shady doctor, by promising him a share of the spoils, but when Gou Cai eventually dies, they swindle the doctor out of his share. Although Gou Cai's sorry fate represents a working out of the karmic consequences of his treatment of his daughter-in-law, the actual events are so dreadful as to constitute the moral nadir of the book. On the other hand, Gou Cai's recurrence throughout the text does, as I discussed above, make his story more pressing than most of the others presented. In fact, Gou Cai's trajectory turns out to be ominous for Jiusi as well, as the final two chapters, which follow hard upon Gou's tale and almost equal it in finality, make clear. These chapters contain an extremely decisive conclusion, a closure that is as sudden and powerful as it is unexpected. In chapter ro7, Jiusi receives a letter from one of Jizhi's other employees telling him that his client Cai Liisheng is now serving as a county magistrate in Shandong; at the same time, in the next county, we are told, the sudden death of an uncle and aunt of Jiusi has orphaned their two sons. Jiusi sets out to rescue the two boys (after notifying his embezzler-uncle of the situation, who characteristically advises him not to get involved) and then goes on to pay a visit on Cai. Jiusi arrives in Shandong to bring the boys back, but is in the process robbed of the considerable sum of money he had been carrying. While all this is going on, he receives word that Jizhi's mother has died. When he

Theodore Huters

297

arrives in Cai's county, he finds that Cai has been unjustly relieved of office and has been fined by the government because he had distributed famine relief to the people of his xian against the wishes of his fellow magistrates in the region. It is surely significant that these events, which serve to unravel Jiusi and his social network, take place while he is in inland Shandong. The sluggishness of travel by horse-drawn cart and his isolation from sources of news in this remote, and perhaps more to the point, traditional place are foregrounded in this episode and usher in a new mood in which Jiusi is no longer in such fluid control of the narrative. Without access to modern transportation and the telegraph, Jiusi's ability to gather stories that reach beyond his own small circle suddenly grinds to a halt. 45 After this decisive interlude in the interior, Jiusi returns to Shanghai to try to raise money for Cai, only to be told that Jizhi's commercial empire has suddenly and utterly crumbled. This collapse leaves Jiusi liable, since he had accepted responsibility for the firm when Jizhi was serving as an official. The story ends with Jiusi journeying upriver incognito to avoid his creditors in order to settle the affairs of the embezzler-uncle, who has also died suddenly. Upon his return to Shanghai, Jiusi goes to the house of Wen Shunong, Jizhi's trusted employee and the bearer of many of these bad tidings, only to find the house burned to the ground. Finally, Jiusi flees to his ancestral home in the south, but first turns his diary over to Shunong, who turns out to be the man from whom the wandering Sili taosheng had received the text in chapter r. 46 Aside from ending the novel on as sensational a note as possible, with death, lamentation, and corrosive pettiness on all sides of Jiusi, these final two chapters also abruptly restore his personal narrative to prominence. This sudden reappearance of Jiusi's story destabilizes the comfortable pattern that the novel had fallen into and renders it again as uncertain as it had been in the early chapters. If the attenuation of his personal narrative had coincided with a new sense of cognitive control on his part, how does its sudden return to the foreground affect the overall shape of the narrative? It had seemed as if the text after chapter 40 was basing its determined collection and representation of tales onJiusi's capacity to understand and to organize them. In the final two chapters, however, Jiusi's own story sneaks up on him, as it were, in such a furious way as to cast doubt upon the idea that he ever had achieved the sort of capacity for practical understanding he had seemed to manifest earlier. Furthermore, the overpowering force that marks the return of his narrative after its long abeyance suggests some mechanism has been repressing it, which would help account for the nightmarish quality of its return.

298

WU JIANREN AND THE REFLECTION OF STRANGE EVENTS

This change of narrative pattern suggests that the cognitive equilibrium that marks the novel during most of its latter part is more unstable than the narrative voice is aware. If the record of events themselves is throughout rather pessimistic, Jiusi's new way of thinking and writing about it animate the grim account. In the final chapters, however, all the elements that had made up the new and had enabled Jiusi to move about and to contemplate his own (and China's) situation-steamships, telegraphs, Western examples, new ideas about the role of women in Confucianism, the reformed novel, and, above all, his superior position outside events-collapseY If actual observed events never turned out well, the only prospect for improvement resided in the representation of a consciousness that could conceive of alternatives; the final collapse provides a most graphic representation of how this promise evaporates. The slow accretion of Jiusi's practical capacity for understanding achieved textual figuration in the first-person voice, which represented the hopes for cognitive development. With the collapse of this voice in the almost gratuitously brutal cutting down of Jiusi's status in the world, a sense of the instability of events and the corruption surrounding them fills the resulting vacuum. The final chapters demonstrate retrospectively with maximum force the actual incapacity of Jiusi to manipulate, even symbolically, the events that he witnessed over the years. 48 The leverage that an external position seemed to promise turns out to be nugatory. In being for most of its course in sympathy both with Jiusi's critique of what he finds and with his desire to achieve practical understanding, the rhetorical posture of the text as a whole, however, clashes with its own dire conclusion. On the one hand, then, the impasse of Wu Jianren's writing represents a particularly powerful instance of an inability or an unwillingness to produce the sort of comprehensive guidelines to the world that the contemporary critical discourse on the novel persisted in demanding of the form. On the other hand, the very brutality of the ending signals the urgent desire in the writing voice for just such a transcendent position, even as it fails to locate one. The text comes to the conclusion that it cannot fulfill the demands of contemporary critics, even as it seems at some level to accept them. What the novel does not settle for, however, is the aesthetic mood of resignation characteristic of The Scholars, a book that remains content with creating a purely intellectual space somewhere between protest and almost disinterested contemplation. Strange Events embodies, in other words, an outlook on the world frustrated by its own inability to create a representation that has any hope of affecting the world.

Theodore Huters

299

Perhaps it was this tacit admission that there was no real alternative to surrender to the power of things as they are that so aroused the ire of Lu Xun. Even if one were to grant him the truth of his highly problematic remarks about the lack of forbearance in the late Qing novel, his condemnation of this intolerance does not ring particularly true if one seriously considers some of his own narrative work. It is difficult, for instance, to think of The True Story of Ah Q as exhibiting some vast reservoir of generosity toward its objects of scrutiny that is completely foreign to late Qing fiction. It is fairly evident from the remarks on Strange Events quoted above that Lu Xun's rejection of the work is based on what he perceives as its failure to present a sustained and comprehensive critique, its failure to gain, in short, a useful purchase on the unfolding of events. In a real sense, Lu Xun's frustration with the text is of a piece with the novel's own frustration with its inability to attain control over the events of its own representational universe. If, however, one understands Lu Xun's critique of the novel as based predominantly on aesthetic concerns (as most subsequent critics seem to have done), it would be easy to lose sight of the ideological nature of his paradoxical argument with Wu Jianren. Even the "hard-boned" Lu Xun (as Mao Zedong called him), in other words, was unwilling to accept the foreclosure of possibility represented by the collapse of the external perspective in Wu's text. The tendency to ascribe the novel's difficulties to simple aesthetic failure contributed to the notion that behind this failure lay a potential narrative fully capable of controlling a sufficient representation of events. Critics who took this line, such as Liang Qichao, seemed to be motivated by a vision that there existed somewhere a spectral text undergirded by a firm historical telos that would lift China out of the abyss. Lu Xun was clearly not the sort of person who indulges in teleological fantasies, but even he seems not to have been able to make his peace with the radical epistemological uncertainty that marks the end of Wu Jianren's novel, however much his own work echoed it. Wu at least presented forthrightly the confusion endemic to the period of transition in which he lived. Most later efforts to build narratives of modern China would, in fact, foreclose on this fearful asymmetry by imposing a unilinear external trajectory on the events they chose to depict. Ironically, however, the need to find a way toward a new kind of control over a desperately amorphous situationto "social engineer"-was to grow on its own failure, in social practice as in literature, as utopian thinking seemed to feed on the progressive collapse of ordinary civic activity as the twentieth century drew on.

CHAPTER I2

Confucian Agendas for Material and Ideological Control in Modern China R. Bin Wong

The Problem of Local Order in an Agrarian Empire Both Chinese officials and modern scholars have pronounced upon the difficulties and limitations of ruling an agrarian society at the local level in imperial times. We conventionally speak of the limited capacity of the government to penetrate local society and official reliance upon elites for whatever government success at local rule was achieved. Yet we should not underestimate the measure of success achieved by the Chinese state and would do well to remember that only a broadly achieved success at local control created repeated opportunities for the central state to meet the difficulties of ruling an agrarian empire. No other state in world history ever enjoyed the challenge of creating instruments of local rule over two millennia. During the second millennium of imperial rule, the settlement of growing numbers of people across ever larger areas demanded innovative strategies for local government, if China was to remain the largest demographic and territorial unit in Eurasia. While many techniques certainly elaborated upon older institutions and policies, at their core lay what we might call a Confucian agenda for social order that suggested to local elites new ways to promote social tranquillity and popular welfare. Elite projects, such as building schools and granaries or repairing temples and roads, suggest norms of socially appropriate and politically expected behavior. Zhu Xi (I r 3o-12oo) and his contemporaries championed the role of elites in buttressing social stability through good deeds. These activities could be alternatives to what had been deemed undesirable government

CONFUCIAN AGENDAS IN MODERN CHINA

activities. 1 More generally, however, officials viewed elite efforts not as substitutions for official efforts to maintain local social order but as complements to them. For their part, elites often pursued priorities that meshed closely with government concerns, but at other times their practices diverged from those anticipated by official sensibilities; the contrast of late Ming gentry patronage of Buddhism in Timothy Brook's chapter (pp. r6r-8r) 'with high Qing anxieties over Bixia Yuanjun in Kenneth Pomeranz's chapter (pp. r82-204) suggest some of the variations. But even when the late Ming gentry contributed to Buddhist monasteries, they supported Confucian projects of building granaries and schools. The mix of official and elite efforts to promote local order assumed particular configurations in eighteenth- and nineteenth-century China that reveal how political norms were translated into social practice. A selection of these efforts forms the principal focus of this chapter. What cultural understandings guided efforts at creating local social order? What factors shaped the range of possible roles that officials and elites played? How did official and elite efforts aimed at local order change during the eighteenth and nineteenth centuries? What legacy for local order in postimperial China do these practices offer? Answers to these questions should help explain the culturally specific strategies for creating local social order and subordinating a myriad of locales to an imperial state. By tracing these patterns during both the eighteenth and nineteenth centuries, we will be able to discern the changing manner in which the state's vision of community institutions plays a role in constructing local order. Because the logic of community institutions can be implemented by elites acting outside formal government, local order can also be created with little state input. But when neither officials nor elites sustain Confucian community institutions, as happened in the twentieth century, an alternative basis for social order must be invented. Strategies to maintain social order can be divided into three general categories-moral, material, and coercive. 2 Leaders can appeal to people's sense of what is right or proper, reward them with goods and services, or threaten them with physical violence. All three were pursued in governing rural China. Appeals to moral reason included exhortatory lectures and the formation of local educational institutions, whose major purpose was to school people in proper social relationships. Material appeals centered on promoting production and providing welfare mechanisms aimed at defending a viable subsistence for all people, no small task when pursued across an agrarian empire. Coercive efforts aimed at control through the registration of the population in order to tax all landholders and to keep track of all households among whom always lurked, in the official mind

R. Bin Wong

at least, potential troublemakers. The three kinds of control, though analytically distinct, were in practice often combined. The lijia system in the early Ming, for instance, was intended for both fiscal operations as well as more general surveillance; as the latter function declined within the lijia, an initially distinct baojia system for mutual surveillance was formed. Not only could these two systems become enmeshed, but baojia also became associated with a village lecture system called xiangyue, which in the Ming were community groups organized for moral instruction. 3 The importance of gentry households and other local elites to the maintenance of local social order has long been noted by scholars. Some forty years ago, Fei Xiaotong made the gentry the basic intermediary between officials and peasants. 4 More recently, G. William Skinner's observation that the size of the state relative to population contracted over time has often been the point of departure for scholars to expect elites to play a greater role in local affairs at the expense of the state during the Ming and Qing dynasties. 5 While American historians in the 1970s analyzed some features of state-gentry relations, the Japanese did far more to examine directly the interests and motivations of the gentry themselves. 6 Shigeta Atsushi outlined a theory of gentry control over local society that took issue with previous Japanese scholarship. He argued for the emergence of a historically specific kind of gentry landlord rule beginning in the late Ming. Other scholars have debated Shigeta's interpretation since his untimely death/ In general this Japanese literature highlights lower Yangzi experiences. Authors conventionally stress the gentry's assertion of control over local society. Some consider the gentry to be separate from the state, whereas others argue for close connections. 8 For Shigeta, local social order was basically rooted in the gentry's economic power over tenants. Japanese scholarship in the 1970s more generally recognized that there could be disagreement between gentry and officials over issues like water control and taxation. However, whether addressing the congruence of official and gentry interests in local social order or their conflicting positions over particular issues, Japanese scholarship on "gentry control" has overwhelmingly been concerned with coercive control and power over material resources. The Japanese literature on gentry control was followed by a newer literature focused on local elites taking on ever larger responsibilities. In contrast to the earlier Japanese literature on Ming and Qing gentry, some of the more recent Japanese and American scholarship on elites includes assessments of social management themes, in particular elite provision of social services. William T. Rowe's two books on the central Chinese riverport of Hankou during the nineteenth century present the emergence of a

CONFUCIAN AGENDAS IN MODERN CHINA

new kind of treaty port city in which the management of urban social order rests crucially upon the efforts made by merchants and other elites to organize the city. 9 Mary Rankin's work on parts of the lower Yangzi region in the decades following the mid-nineteenth-century Taiping Rebellion tracks the activities of local elites who asserted expanded roles for themselves as they re-establish social order. 10 Rowe and Rankin tie their analyses of social change in nineteenth-century China to the European notion of a "public sphere. " 11 Those who espouse a Chinese variant of "public sphere" relate nineteenth-century phenomena studied by Rankin and Rowe to the gentry society literature for the late Ming and early Qing. For those arguments that move from the late Ming forward, the spatial dimensions of the arguments are generally confined to regions along the Yangzi River, especially to places in the lower Yangzi. A second key locus of evidence comes from new urban centers in the nineteenth century. Missing from many considerations of these issues is much sense of the agrarian empire beyond the lower Yangzi and large cities. Nor is there much attention paid to the eighteenth century as a period of state activism different from both the seventeenth and nineteenth centuries; the image of elite expansion of social responsibilities linking late Ming and post-Taiping China misses much of what happened in the high Qing. This chapter reconstructs key components of the eighteenth-century program for social order in an attempt to understand how certain political and cultural norms became social practice. As a result, some of the arguments about elites, social change, and a "public sphere" will be viewed in light of an examination of particular institutions and practices, with special stress upon spatial variations across the empire. I seek to understand the Confucian agenda for local orde~ across diverse concrete circumstances. For their part, Qing officials well understood that the challenges of local rule varied dramatically. Placed firmly within an empire-wide context spanning the eighteenth as well as nineteenth centuries, an examination of local granaries and local schools shows how political expectations, social norms, and cultural practices were reproduced and changed during China's last dynasty.

Rural Granaries and Subsistence in Qing China Protecting peasants against subsistence uncertainties was an element in Chinese political thought from classical times on. In the Ming and Qing, official efforts spanned short-run responses to famines and long-term efforts to expand production. Granaries to lend or sell grain during the lean spring season according to need reached their zenith during the eigh-

R. Bin Wong

teenth century. The first Qing emperor envisioned ever-normal granaries (changpingcang) in county seats, charity granaries (yicang) in major towns, and community granaries (shecang) in the countryside. 12 Ever-normal granaries were to be managed by the magistrate's staff, who were directed to sell, lend, or give away grain in the spring and to make purchases, collect loans, and solicit contributions in the autumn. Community granaries, in contrast, were to be small in scale and scattered across the county, stocked by contributions of local rich landowners, and managed by local people. The early Qing ideal of a charity granary combined features of both evernormal and community granaries. As with the community granary, grain was to be mobilized through contributions; like the ever-normal granary, distribution was to take the form of reduced-price sales and loans. The granary system developed gradually. 13 Ever-normal granaries were the first to be established, beginning in the r 68os. Their formation relied upon contributions, funding from provincial treasuries, and the diversion of the grain tribute (caoliang), a tax-in-kind levied on a subset of China's provinces thought to be best able to supply the capital with rice. As evernormal granary reserves became widely established by the 1720S, the Yongzheng emperor (r. 1723-3 5) expanded granary promotion to include rural community granaries. Community granary reserves grew in several provinces in the early Qianlong (r. 173 6-9 5) years and, beginning at midcentury, were joined in Zhili and Shanxi by rural charity granaries that complemented community granary operations. The explicit logic of community granaries put responsibility for the creation and maintenance of these institutions in the hands of local people. The state's willingness to depend on the gentry and others to promote local grain reserves assumed a basic shared commitment to subsistence security as a key element in social stability. It also recognized the fiscal and organizational limitations of officials, who were unable to manage directly or finance fully a network of rural grain reserves. To guarantee the formation of community granaries, officials did, however, have to play major roles. The Yongzheng emperor permitted the use of official funds to finance the community granaries in those places where elite contributions were not forthcoming, but he firmly opposed officials becoming directly involved with granary operations for fear that they would take advantage of the resources for their own purposes. 14 Thus, there was a positive preference for elite involvement and sensitivity to the potential for local official abuse. In the early years of the Qianlong reign, officials in several provinces made efforts to expand the coverage of community granaries through the establishment of new community granaries with official funds or reserves

CONFUCIAN AGENDAS IN MODERN CHINA

diverted from ever-normal granaries. This practice was followed in Anhui, Guizhou, Jiangxi, Shanxi, and Yunnan. 15 Official efforts to expand community granaries in some provinces were accompanied by a growing recognition in others that the state had to establish monitoring procedures over community granary reserves if officials expected to integrate community granary holdings into their planning for subsistence security. The need to develop official monitoring of community granary reserves was a clear sign that elites were responding to official exhortations that called upon them to fund and manage local grain reserves. In Hubei, Hunan, and Jiangxi, a system of consolidated community granaries (zongcang) was established in which accounts for several community granaries were first totaled before transmission to county authorities, thereby reducing the number of ledgers the magistrate had to review. 16 County-level yearend balances of community granary reserves were aggregated at the provincial level and sent to the capital in memorials also detailing ever-normal granary figures. Community granary reserves thus became a part of a larger system of grain reserves used by officials to relieve subsistence anxieties. In provinces like Hubei, Hunan, and Jiangxi, it appears that the major responsibility for funding and managing community granaries rested upon the local elites. Officials promoted the formation of rural grain reserves and developed procedures to monitor their use but did not believe their roles undermined Zhu Xi's intent of relying upon local elites to run community granaries. In Gansu and Shaanxi, however, a second kind of community granary was founded according to the same rules applied to evernormal granaries in that region; both granary types were run by officials who made sales and loans for both consumption and seed grainY The larger official role in rural granary operations followed directly from the absence of rich elites to whom the responsibility for community granaries could be delegated. When officials so chose, they c::ould promote rural grain reserves on an even larger scale. In 1746 Zhili governor-general Nasutu suggested diverting part of the grain tribute to those parts of Zhili lacking adequate rural reserves; in the following year he called for the establishment of rural charity granaries, noting that as many as 70 or 8o percent of community granaries in the capital province were located near ever-normal granaries and thus unable to reach peasants at some distance from county seats. 18 Fang Guancheng, who became governor-general of Zhili in 17 49, outlined the dimensions of the Zhili charity granary system-a total of r,oos granaries were to be established to serve 39,687 villages. In the vast majority of Zhili counties, charity granaries were to be no more than

R. Bin Wong

r 5 to 20 li from each village. 19 Although we do not know how many charity granaries were in fact established and can easily imagine that far fewer than a thousand were created, charity granary reserves reported to the emperor did double from some 2oo,ooo shP0 in 1749 to some 4oo,ooo shi in the I77os and r78os. 21 The Zhili success prompted the emperor to have other north China governors consider implementing a similar system. Governors in Shandong, Henan, Shaanxi, and Gansu responded that a charity granary system was either not feasible or not necessary; their responses suggest a consistent lack of enthusiasm for taking on such a large task. Only Shanxi governor Aibida supported the idea and promoted the formation of charity granaries, for which year-end reserves began to be reported to the emperor. 22 The eighteenth-century officials who built the Qing granary system understood that the degree of government effort to mobilize and manage reserves outside county seats housing ever-normal granaries depended crucially on the presence or absence of elites who would form granaries as their own Confucian projects. In sharp contrast to those provinces in which officials took on a major role in creating rural reserves are provinces like Jiangsu and Zhejiang where official involvement in rural granary operations was minimal, even in the decades of greatest state activism. For Zhejiang, memorials on granaries report no figures for community granaries, and for Jiangsu, memorials on granary conditions report limited information on very modest community granary reserves located largely in the capital area. 23 The absence of reports on rural granaries does not mean there were no rural reserves; gazetteer data reviewed by Hoshi Ayao shows that there were community granaries in Jiangsu. 24 Rather, the lack of memorials affirms the small degree of official involvement with the rural granary efforts that did take place in the lower Yangzi region. We cannot assume, however, that the absence of active and systematic official involvement meant that granaries were of no interest to officials and that granaries had become solely an elite project. From an empire-wide perspective, eighteenth-century officials appear to have avoided establishing official oversight in rich provinces where such an effort seemed superfluous but worked hard to succeed at methods of monitoring and control in most other cases. In addition to Jiangsu and Zhejiang, there were four other provinces in which community granary reserves were not a well-developed part of the granary system-Gansu, Guizhou, Shandong, and Guangdong. Gansu and Guizhou appear to be cases in which the government attempted to manage the food supply without developing community granaries, perhaps

310

CONFUCIAN AGENDAS IN MODERN CHINA

because ever-normal granary development seemed to officials an adequate strategy or a sufficient challenge. 25 In Shandong, the absence of rich elites to form community granaries was partially compensated by ever-normal granaries that gave grain loans, which were usually not part of ever-normal operations but instead left to community granaries. The Guangdong case resembles the lower Yangzi cases in that the absence of memorials including community granary reserves in the state-managed system may not mean that such reserves were uniformly unavailable in Guangdong, but rather that officials were not systematically involved with monitoring and coordinating rural granary activity. For Guangzhou prefecture specifically, Matsuda Yoshiro has assembled Qing dynasty data on 90-odd granaries and benevolent halls, with information on their dates of founding, management, and functions. 26 Since the local official's role was circumscribed in much of Guangdong by powerful corporate lineage structures, it is not surprising to learn that state participation in granaries was largely limited to urban settings. As in the lower Yangzi cases, the absence of official oversight reflects the partial reliance on local elites for food supply management; elites largely shared the Confucian agenda for local order that the eighteenth-century state promoted more directly in other parts of the country. From an empire-wide perspective, community and rural charity granaries figured prominently in the eighteenth-century state's system of subsistence management. In Hubei, Hunan, Jiangxi, Shaanxi, Shanxi, Yunnan, and Zhili, they accounted for some 20 to 40 percent of total provincial granary reserves. Beyond the sheer size of community and rural charity granary reserves (more than 5 million shi for the seven provinces listed above during the r 770s ), the pattern of management and monitoring reveals clear spatial variations. Elites and officials played varying roles in creating and sustaining community and rural charity granaries as components of a larger granary system in the eighteenth century. Zhu Xi's ideal of local management was complemented by official oversight in many provinces; without this effort to track community and rural charity granary accounts, officials would have been unable to conceive of these granaries as a first line of defense in short harvest seasons. In several of those areas without rich and responsible local elites, eighteenth-century officials played an even larger role in developing rural grain reserves. In contrast, officials in the lower Yangzi provinces of Jiangsu and Zhejiang, as well as in the south China province of Guangdong, kept no systematic account of community granary reserves. Thus, where elites were most able to take care of community granary matters, the state was least likely to oversee granary operations.

R. Bin Wong

pr

The spatial variations in eighteenth-century China's civilian granary system suggest that the "public sphere" some scholars have found in China needs additional specification. Both Rankin and Rowe make clear the geographical origins of their examples, but they do not tell us much about situations outside their samples of places with a "public sphere." But it matters to our general understanding of a "public sphere" to know what exists in those places without much of a "public sphere." The eighteenthcentury civilian granary system suggests that officials would depend on a "public sphere" in places where they could expect it and would substitute for it in other places. The awkward nature of speaking of officials "depending" upon a public sphere points to the poor fit the concept may have for thinking about Chinese politics and society. The powers of European states were contested and constrained by elites in their public spheres. Public opinion became actively engaged in confirming and challenging different government policies. In contrast, the Confucian agenda for local order in eighteenth-century China did not pit state and elites against each other in a European manner. Nor did it privilege either of them as the guardians of local order. It admitted both in spatially variable ways to sustain social stability. The eighteenth-century state's system for subsistence management included monthly price reports for grains in every prefecture of the empire as well as weather and harvest reports. On the basis of information in these documents, officials made decisions on how best to use granary reserves and influence commercial shipments to balance demand and supply within small locales and across great expanses. The nineteenthcentury decline of the granary system was occasioned by factors both within and without the system. From within, the granary system was undermined by the difficulties officials faced in repeatedly mounting the organizational efforts needed to keep the system running. The Jiaqing emperor (r. 1796-r82o) took an important step in 1799 when he freed local community granaries from having to report their activities to local officials, who in turn no longer reported on these granaries to provincial officials. From without, there emerged growing demands for grain among the military; these depletions required additional funding for restocking, which was rarely forthcoming. Gone was the eighteenth-century system of official monitoring and coordination of granary activities within provinces and among them. 27 But this does not mean that officials and elites ceased their efforts to build granaries. Officials and elites in several provinces continued to makeJocal efforts to fund and manage granaries during the nineteenth century. Compared to eighteenth-century practices, granaries in some places did become more

CONFUCIAN AGENDAS IN MODERN CHINA

of an elite responsibility since officials by themselves were increasingly unable or unwilling to sustain ever-normal granaries. But more salient on an empire-wide level are the overall decline in reserves and the absence of a centrally monitored and coordinated system. Reliance on local elites to sustain rural granaries was hardly new in the nineteenth century. During the late Ming and even under the Yongzheng and Qianlong emperors, community granaries in Jiangsu and Zhejiang were not subject to the kind of monitoring that became common for much of the empire. What the nineteenth-century granary situation suggests, however, is that the Jiangnan model of elite activism may have become more common, with less official leadership and certainly a collapse of systematic official oversight; individual local officials might still play important roles, but the larger bureaucratic structure of the state did not matter as much as before. Thus, a Chinese "public sphere" of elite activism emerged as much from the reduction of official activism as it did from an increase of elite efforts. This dynamic bears little resemblance to early modern European elites seeking to stake out a set of claims against the state in their public sphere. Spatial variations similar to those found for the Chinese granary system can be seen among local schools.

Local Schools in Qing China The Chinese, more than any other state in the early modern world, made the principle of instruction (jiao) basic to its conception of political rule. During the Zhengtong ( I4 3 6-49) reign of the Ming dynasty, community schools (shexue) were established in many counties as an official undertaking; by early Qing times a few remained, at least in name if not always in practice. 28 The most common form of local school in Qing China was the "charity school" (yixue). These schools were established by the government in border regions to civilize local populations, and local elites responded to official exhortations to build yixue in interior parts of the empire. 29 Angela Leung's recent work on charity schools during the Ming and Qing demonstrates the central role played by the leaders of local communities rather than by officials in the creation and maintenance of these institutions. Although her research is focused primarily on lower Yangzi examples, she includes evidence from other places as well that suggests a major role for local elites in local education outside the lower Yangzi as well.3° But how common was this lower Yangzi pattern? In frontier provinces, the state clearly played a larger role in charity school formation. In the Huidian shili (Collected rules and precedents}, one finds frequent mention of official efforts to promote charity schools

R. Bin Wong

313

in Guizhou, Yunnan, and the more peripheral parts of other provinces such as Sichuan, Hunan, and GuangdongY Charity schools were especially associated with the education of minority peoples. The best-documented case of successful government promotion of charity schools is the famous eighteenth-century official Chen Hongmou, who in Yunnan spearheaded the founding of 6 so charity schools. As William Rowe's research has shown, Chen made his subordinate officials responsible for developing charity schools. Through repeated questioning, he made them increasingly uneasy about their failures to establish these educational institutions. He suggested that they turn to a variety of sources to fund the schools-local budget surpluses, customary fees, rents from estates attached to their bureaucratic posts, revenue from newly opened and previously unregistered land, and contributions from local elitesY Whenever possible, Chen sought land endowments for charity schools so that they would have a secure source of funding. Elites played a role in establishing charity schools through financial contributions as well as gifts of land, but their efforts were not the driving force behind the expansion of charity schools. In sharp contrast to the dynamics in the lower Yangzi, officials in Yunnan, responding to the governor's insistent initiatives, clearly played the leading role. The contrasting importance of elites and officials in charity school formation in the lower Yangzi and Yunnan parallels their relative significance to community granary formation in economic cores and imperial peripheries during the eighteenth century. In between these two polar cases, charity school formation involved a mix of official and elite benefactors, just as community granaries did. For Hunan, the I 8 8 5 provincial gazetteer lists some I66 charity schools formed during the Qing dynasty, not all of which were still functioning in the late nineteenth century. Information on the roles of officials and elites is limited, but for eighteen cases in which someone is mentioned as having played a role either in founding or repairing a charity school, sixteen involve officials and only two involve elites. 33 Although these numbers represent far too small a proportion of the total for us to infer these ratios of official and elite involvement to the larger set of charity schools, it is likely that officials played a considerable role in charity schools formed in Qing dynasty Hunan. For Sichuan, the I 8 r 5 provincial gazetteer lists 46 early and mid-Qing instances of charity and community schools being formed or repaired; of these, eighteen cases mention officials and eight mention other people. 34 Thus, both officials and elites participated in creating and sustaining the province's local schools. In the Shaanxi provincial gazetteer of I934, there are 24 instances of officials being cited for founding or sustaining a

CONFUCIAN AGENDAS IN MODERN CHINA

local school, but only seven cases of gentry and other non-officials are noted. 35 For Guangdong, the r822 provincial gazetteer lists some 6oo community schools from the Ming and early to mid-Qing periods, but offers virtually no information on the individuals responsible for themin six cases officials are mentioned and in only two cases are other people mentioned. 36 The Hubei provincial gazetteer from 1921 is much more informative. It gives cases throughout the Qing of both official and elite participation in the founding and repairing of charity schools-44 instances mention officials, and another 56 suggest non-official sponsors. In addition, the seven schools in one particular county are reported to have been established by both officials and virtuous peopleY Quite clearly a mix of official and private efforts was responsible for the local schools in Hubei. The different sets of provincial data on local schools for Hunan, Sichuan, Shaanxi, Guangdong, and Hubei collectively document the presence of both official and elite efforts to form schools. Whereas officials were clearly most important in frontier areas and least important in economically developed areas like the lower Yangzi, for much of the empire local schools were the joint product of elite and official activity. Even if records of official sponsorship exaggerate the role of local officials, who may have sometimes played more honorary than real roles, the actual participation of both officials and elites in local schools generally is undeniable. Schools and granaries complemented each other ideologically. The promotion of education and economic welfare were twin commitments of the local magistrate; thus jiaoyang, "to instruct and nourish," was considered a basic feature of government. 38 Moreover, the variable roles of officials and elites across the empire were also similar for both. But the organizational formats developed in the eighteenth century to coordinate these activities were separate vertically structured institutions rather than a single horizontally integrated local system. Local community granaries fit within a larger granary system that was itself simply one major component of a broad repertoire of state options to influence food supply conditions; local granaries had the straightforward function of feeding peasants during a particularly lean year and occasionally providing seed grain for the next year's planting. Local schools, however, had multiple functions. State-sponsored schools, especially in the empire's frontiers, were aimed at "civilizing" minority groups by indoctinating them in Confucian moral and social precepts. But local schools in the interior were often conceived as the first step of study for boys with scholarly promise; some of those who did well in local schools could expect to continue their

R. Bin Wong

studies beyond the curriculum locally available in order to pursue the greater goal of passing the civil service examinations. One indication of the different social and political qualities of local granaries and schools comes from consideration of the role that lineages played in each. There are a few cases of lineages' managing grain reserves as part of a charitable estate, especially in the nineteenth century, but even then it does not appear to have been a common practice. 39 Moreover, the government does not appear to have pushed lineages to take on this task. 40 One official who did consider lineages for this purpose was Chen Hongmou. While serving as governor of Jiangxi in the early 1740s, Chen suggested that lineages could be the units for local granaries instead of some less clearly defined territorial units. He reasoned that families who made loans to their kin were not always able to collect repayment. If grain were put into a community granary and the accounts forwarded to officials, then lineage members, according to Chen, could reap the benefits of a community granary without sharing grain with non-kin. 41 However, in succeeding provinces where he served as governor, Shaanxi (17 4 346), Hubei (1746-48), Shaanxi (1748-51), Henan (1751-52), Fujian (1752-54), Shaanxi (1754-55), Hunan (1755-56), Shaanxi (1756-57), andJiangsu (1757-58), he does not appear to have made similar suggestionsY For local schools, lineages played larger roles. Supporting local schools appears to be one of the most common lineage undertakings outside those directly centered on the ritual recognition of the kin group. Not only did rich corporate lineages in Guangdong and Fujian often maintain local schools, but far weaker ones in the Yangzi valley and in North China also supported local education. 43 The term yixue (charity school) was applied to some of the schools founded by lineages; thus, lineage-based schools were linguistically equivalent to those schools promoted by officials in the border regions and founded by elites in other interior areas. Since elementary education really meant moral indoctrination, the principles taught in charity schools affirmed the kinship relationships of lineage organizations and promoted broader visions of Confucian social order that lineages found compatible with their own roles in local society. Community and charity schools reached culturally distinct target groups across the empire, as they served the overlapping interests of a government seeking to promote a common ideology in diverse settings and of elites and lineage organizations committed to securing the moral foundations of local social order. The mix of official and elite efforts to build schools varied across space just as it did for local granaries. As the

CONFUCIAN AGENDAS IN MODERN CHINA

eighteenth-century granary system declined, local granary operations began to resemble more closely local schools in both the eighteenth and nineteenth centuries in terms of variety. Both granary and school activities became more variable in the nineteenth century as centrally stimulated and guided efforts at subsistence management and moral indoctrination were replaced by provincial and subprovincial initiatives. Local efforts at creating and sustaining charity schools, like those aimed at mobilizing grain, involved both officials and elites. The fuzzy distinction between official and non-official roles in education is captured by the county "studies official" (xueguan), whose position, in Alexander Woodside's words, was "incompletely bureaucratized. " 44 Neither a regular official nor truly non-official, the "studies official" makes a precise distinction between official and non-official difficult. Thus, the "public" in the socalled public sphere in China contained varying participation by officials and elites, whose efforts existed along a continuum that admits of no sharp distinction between what is state and what is socialY In the shift from an eighteenth-century mix of official and elite efforts promoted and coordinated by higher levels of government to nineteenthcentury patterns of local officials and elites working without much guidance or control from above, the effective delegation of power to elites from officials is easy to see. Equally important, and more difficult to recognize, are the more explicit local limitations of nineteenth-century government practices. Obscured by a Confucian rhetoric that masked these changes, the bureaucratic structures of the high Qing for creating rural social order in an agrarian empire disappear from view and perhaps even consciOusness.

The Eighteenth-Century Search for Central Control and Its Legacy for Modern China Scholars have long noted the manpower and organizational limitations of the late imperial Chinese state that precluded its effective penetration of village society. This observation has often led us to expect, implicitly if not explicitly, that the state could succeed at few efforts to order local society and that if the government did make such efforts, it could do so only through delegation of responsibility to local elites. When this train of thought is joined to arguments that local elites had their own priorities and desires often quite separate from and even in competition with those of the state, we are easily carried to the conclusion that the state was at most of limited relevance to local order. The evidence for separate and

R. Bin Wong

even competing agendas is drawn out most clearly for taxation issues. 46 For other issues, there often seems to be the assumption that if programs were not funded by the state, then the interests they served must be separate from the state. This chapter argues to the contrary that in the realms of material welfare and moral indoctrination the state and local elites shared a broadly common agenda, which, during the eighteenth century, was implemented across the agrarian empire through a mix of official and elite efforts coordinated, if not fully controlled, by the central state through vertically integrated reporting proceduresY Officials defined the dimensions of their involvement in these concerns according to their assessments of local social structures and economies. In general what distinguishes eighteenth- and nineteenth-century government efforts at sustaining social order is the degree to which higher-level officials sought to monitor local society. The eighteenth-century granary system demonstrates the state's capacity to construct a massive and sophisticated structure to influence material welfare across diverse locales. Although certainly a fragile achievement in fiscal and organizational terms, it was nevertheless sustained for many decades as a complex system of grain mobilization, storage, transfer, and distribution. Not surprisingly, a government capable of this kind of success also attempted to impose vertical control in other areas as well. In the realm of moral indoctrination, the Qing emperors paid considerable attention to implementing the village lecture system (xiangyue). The term "xiangyue" had referred in Song and Ming times to a local community organization within which social harmony was to be promoted through proper instruction and material aid. In the Southern Song, the xiangyue was conceived as a unit within which people could organize to fight fires, repel bandits, and care for the sick and poor. 48 During Ming times, the Hongwu emperor (r. 1368-99) called for the establishing of xiangyue as a local educational institution in 13 88. Famous figures like Wang Yangming (1472-15 29) and Lii Kun (15 36-1618) also championed the xiangyue as a local institution that twentieth-century scholars have seen as an instrument of self-government or local elite control. 49 But as Maurice Freedman noted more than 25 years ago, the early and mid-Qing rulers gave the term new meaning as a kind of public lecture system subject to official oversight. 5° In his "Sacred Edict," the Kangxi emperor (r. 1662-1722) offered sixteen maxims that counseled people to work diligently, spend carefully, treat kin and neighbors with propriety, promote correct learning, remit taxes promptly, and organize into groups to prevent thefts. Through the

CONFUCIAN AGENDAS IN MODERN CHINA

village lecture system (xiangyue), people were to listen to lower-degree holders (shengyuan) who elaborated upon these themes. 51 Not satisfied with his father's maxims offered in terse classical Chinese, the Yongzheng emperor issued his own amplified instructions. This imperial effort was augmented by many popularizations. 52 These practices depended on elites without formal bureaucratic positions who recognized their Confucian responsibilities to preach correct behavior. The state made efforts to organize and monitor these lectures as an activity distinct from other local operations. In certain cases, the magistrates themselves were expected to give these lectures. Already in the eighteenth century there were complaints about the limited efficiency of the village lecture system. In part, limitations followed from the absence of an analogue to the granary accounts that tracked the mobilization and disbursals of grain. Even if the village lecture system could not claim the organizational sophistication and success of the granaries, the political intent remains clear-the village lecture system under eighteenth-century rulers was to be a vertically integrated system that specialized in moral indoctrination. Officials were expected to keep track of these lectures as a specific activity. Separate from these material and moral systems, the eighteenth-century state mounted a considerable effort to implement the baojia system to provide for mutual surveillance and security. For all three categories of local control-material, moral, and coercive-the eighteenth-century state aimed to create distinct vertically integrated and functionally specific institutions through which a centralized state could organize local social order. These systems created explicit rules for elites, whose levels and dimensions of participation varied across the empire. 53 The success of high Qing Confucian strategies to promote local order clearly depended on official ability to secure the participation of elites to play roles defined for them by the state in accord with Confucian moral sensibilities. We might expect that the major weakness of these systems was an antipathy elites shared toward official efforts to bend their wills to a state-defined social agenda. For the late imperial elites we know best, the gentry elites of Jiangnan, there is a record of resistance to official interference in local affairs that begins in the late Ming and certainly carries forward into the early Qing; gentry elites were able and willing to manage their own locales without state interference. 54 But it is not so clear that the state perceived this situation in the same way. If in fact elites succeeded in sustaining local social order without official participation, this merely simplified the tasks of ruling the agrarian empire and allowed official energies to be expended on other types of problems. The

R. Bin Wong

absence of official oversight would be least anxiety provoking in those areas where elites were both able and willing to adopt Confucian social and moral outlooks. More generally, the limits on eighteenth-century practices of vertically integrated, functionally specific systems of sustaining local order were imposed less by antagonisms between officials and elites on these matters than by constraints on bureaucratic capacities and shifting ideological commitments. To return to food supply management, the bureaucratic costs of surveillance of local operations could be substantial. In the case of granaries, gathering the accounts of many separate rural granaries, checking to see that the figures balanced each year, and then aggregating them for reports up the official hierarchy was an operation that demanded skill and patience from local officials. As granaries became more common with larger reserves, the difficulties of performing these operations grew accordingly. If mistakes were made, it often became difficult to uncover them in future audits, and accurate monitoring became increasingly less likely. The difficulties of monitoring local activities like rural granaries were not merely an issue of organizational capacities. Higher levels of government could worry about the misappropriation of resources by local officials who shifted either money or grain to some alternative use. Fears of potential meddling limited the role the Yongzheng emperor assigned to local officials in rural granary management. Worries again surfaced when the Jiaqing emperor decided in 1799 to end state monitoring of local rural granaries because he felt it was impossible to sustain official oversight without also encouraging corruption in some county yamens. 55 The central government had to decide whom it trusted more-local elites who would take care of local affairs on their own under the positive force of moral suasion to channel elite self-interest or local officials who would perform surveillance tasks as one of many duties they felt too understaffed to perform. 5 6 The relative importance of material and moral means to achieve local order, not surprisingly, varied over time. Heading toward the midnineteenth century, state commitments to ideological and material order became overshadowed by threats of rebellion and the consequent challenges of mobilizing men to form militia and extracting sufficient revenues to meet military expenses. The shift in fiscal priorities was accompanied by an increased reliance on local officials and local elites, who were expected to coordinate the many activities necessary to first defend and then sustain local order. This meant a strategy for greater local coordination of material, moral, and coercive means of control with less vertical integration of

320

CONFUCIAN AGENDAS IN MODERN CHINA

functionally specific activities. These changes diminished the abilities of the central state to manage local affairs, although many provincial and local officials continued to confront problems of irrigation, river conservancy, and famine relief. 57 Indeed, flanking the purely military crisis of rebellion were social problems that continued to demand state intervention, but because official actions in the more peripheral provinces were driven by military considerations, we conventionally consider these areas separately from the economically more central regions, where the more familiar story of local political and social changes is told. We focus largely upon what might be considered a Jiangnan model of local rule in which elites play prominent roles in providing the services that create social order. What I call a Jiangnan model of local rule was by no means limited to Jiangnan in the late nineteenth century, but the practice of this strategy for maintaining local order had its strongest tradition in Jiangnan, where elites had played such a salient role in local affairs since the late Ming. But the empire-wide context within which Jiangnan elite practices took on broader social significance changed between the late Ming and late Qing. In the eighteenth century, a strong drive for various types of vertically integrated bureaucratic control subordinated elite efforts to official oversight; in those areas without strong elites, officials played more active roles. This spatial mix was coherent in the eighteenth century. The empire's political integration of vast territories at the local level was predicated upon central government recognition of varying roles for elites and officials. The success it achieved depended on establishing a degree of coordination, oversight, and control that proved impossible to sustain amid changing social and political circumstances. Jiangnan elites were those least affected by the bureaucratic changes of the eighteenth century; the shared agenda of local order was met by elites in Jiangnan with little official intervention, even in the eighteenth century. When the eighteenthcentury system fell apart, a Jiangnan style of social order in which elites, sometimes with local official involvement, created local institutions without vertically integrated bureaucratic oversight appears to have become more common. Between the eighteenth and nineteenth centuries, government-coordinated institutions like granaries weakened. Masking this institutional decline is the rhetoric of the Confucian agenda for local order, which did not explicitly require any central government role in the creation of local order. The basic logic of Zhu Xi's local institutions, subsequently elaborated in complementary ways, even by scholars such as Wang Yangming

R. Bin Wong

321

whose philosophical views are generally seen as a reaction and challenge to Zhu Xi, enlisted local elite participation in the project of creating local order under local magistrates. The central state played an active role in promoting these efforts for much of the eighteenth century, but played much less of a role in the nineteenth century. As a result, the Confucian agenda for local order could be articulated on different spatial scales with no change in the basic rhetoric of representation. In this sense, it possesses a "fractal" quality-no matter what spatial scale we look at in late imperial China, we may find expressions of a Confucian agenda for social order. 5 8 There were, of course, disagreements over how best to achieve local order, many of which were articulated with respect to "feudal" (fengjian) and prefectural (junxian) alternatives. But even these different institutional preferences were expressed within a common ideological framework. This "fractal" quality admits of no easy dividing line between state and society. Terms like "public sphere" and "civil society" are hard-pressed to embrace this "fractal" quality, which is foreign to the early modern European expenence. The fractal quality of the Confucian agenda for local order obscures the diminishing spatial scale on which policies to promote local order were implemented in the nineteenth century. Earlier historiography spoke of a shift of power from central to provincial leaders; more receritly the expansion of elite activities has been highlighted. This chapter stresses, as a basic change from eighteenth- to nineteenth-century practices, the collapse of the vertical axis linking center and locality, with consequent challenges to any post-imperial state seeking to create agrarian social order.

Imperial Decline into Modernity and the Emergence of "Civil Society" From the point of view of local institutions, China's late nineteenth- and early twentieth-century history offers a misleading fit with our notions of modern social change. The further decline of granaries removes from view a kind of institution quite foreign to our European-based notions of modern social change, which privilege urban spaces and stress dynamics creating sharp divisions between the state on the one hand and the private economy and civil society on the other. The absence of consideration of late imperial Chinese practices in European models encourages us to focus on Chinese developments easily labeled "modern" because they resemble patterns found in Europe and North America-urban social changes with new

322

CONFUCIAN AGENDAS IN MODERN CHINA

social classes, political organizations, and economic activities. But European trajectories of change did not leave behind the networks of local institutions found in late imperial China. For twentieth-century China, modern change necessarily had to include strategies for local order spanning agrarian villages. How these would be similar to or different from the Confucian agenda prominent in the previous centuries was not obvious, especially because the conceptual categories for Confucian norms and practices was directly challenged and largely undermined by new Western ideas. And yet, some set of strategies and principles to secure local order in agrarian China had to be articulated, as can be seen in the following chapters by David Strand on Sun Yat-sen (pp. 326-45) and by Ann Anagnost on village compacts (pp. 34 6-6 5) in contemporary China. The creation of strategies and choices among them would not (and could not) be shaped simply or entirely by Western dynamics of social change or the categories of interpretation introduced from Western thought. Changes in education also fit the conventional story line of a collapsed empire. The demise of Confucian education and its displacement by Western subjects meant that local education as moral indoctrination lost its basic appeal. Moreover, the state lost its intimate identification with education. Ironically, however, traditional institutions such as lineages took a lead in changing curricula. Over time post-imperial local governments took on educational responsibilities, displacing elites and lineages that had previously controlled local schools and had ushered in the initial phases of curricular transformation. This trajectory clearly differs from the European experience of educational changes, perhaps captured best by the role of lineages. And yet, this plot can be reduced to the notion of the emergence of a "civil society" in which the government takes charge of basic activities like education. If we move forward just a few decades, the Chinese parallels to European changes become less persuasive. Rural subsistence management reemerged after I 949 as a basic political challenge. Although the specific institutional arrangements and the ideological representation of these activities have not closely resembled Qing practices, the persistence of food supply issues in local agrarian society is unmistakable. 5 9 Politics and education have also become intimately enmeshed again since 1949· The state's outlook in the area of elementary education has mimicked late imperial perspectives, as the post-1949 state has generally sought to impose its own moral vision of society through the education of young children. Of course, elementary education now includes many kinds of

R. Bin Wong

323

instrumental knowledge first learned from Westerners. But the presence of these "Western" subjects does not speak unequivocally for Westernization more generally. Nineteenth- and twentieth-century Chinese social and economic changes with their roots in the West are obviously different from earlier historical parallels. Western changes became models, opportunities, and constraints for China, especially after the second half of the nineteenth century. Although China had a long history of urbanism, nineteenthcentury treaty ports clearly added new dimensions to the scene. Urban centers became the most obvious places to expect a "public sphere" and a "civil society." Certainly in Europe it is the cities that house these social formations. But where politics in Europe were largely an urban phenomenon with classical roots in urban polities, Chinese politics were agrarian and imperial. The institutions reviewed in this chapter were rural in orientation. The new Chinese urban developments, including twentieth-century analogues to civil society, would have to be linked in some manner to the rural society that China had been for so many centuries. Discussions of a twentieth-century Chinese urban "public sphere" and "civil society" make far more sense, I think, than those addressing a "public sphere" in previous centuries. The connections between the two have been asserted, but remain problematic. 60 The principal problem, as I see it, lies in specifying the dimensions of similarity and difference for the term "public sphere" in a manner that supports claims for discovering similarities even as it clarifies basic differences. The danger of working with a term like "public sphere" is that insufficient contextual framing will elicit Eurocentric expectations for Chinese patterns of social or political change that are not in fact well founded. Assessments of "modern" political change in China focus, not surprisingly, often appropriately, and perhaps even unavoidably, on ideas and institutions of Western origin. The fate of "modern" political change easily becomes entangled with the success or failure of Western political ideologies and forms of government organization. Although there is no doubt good reason to defend our political values and be prepared to make normative judgments about regimes we like and do not like around the globe, our analytical capacity to discern historical dynamics in specific cases should not be restricted to the possibilities defined by those values. Conflating "foreign" with "modern" and "domestic" with "traditional"a correspondence often asserted by political actors as well as later analysts-still allows us to account for important tensions and choices that

CONFUCIAN AGENDAS IN MODERN CHINA

political actors conceive. But the clarity achieved from this perspective covers a narrow landscape; much remains poorly illuminated and therefore dimly seen. The European epic of modern state formation embraces not only the story of parliamentary institutions and democratic ideology but also the aggressive centralization of power by growing national state bureaucracies that increased their claims on social resources through fiscal extraction. 61 Indeed, the two processes proceeded as much in uneasy tension as in harmony. For the modern Chinese case, there is no comfortable parallel to the centralization of power through the creation of new centralizing bureaucracies and armies to displace formerly more autonomous political forms. China, unlike Europe, was not a patchwork mosaic of more than a thousand independent political units in 1300 that would become gathered into a few national states by 1900. 62 Instead, the late imperial Chinese state labored to implement a Confucian agenda for local order that was partly responsible for the continued reproduction of the agrarian empire. How the ideological and institutional dynamics of this kind of state and local social order intersected ideas of local self-government obviously differed dramatically from the way in which national state formation collided with similar notions in Europe. Philip Kuhn's article on local government set out clearly some of the likely possibilities for local government development in the early twentieth century; he argued for the constraints imposed on participatory and representative forms of local government by a national state committed to maintaining vertical control over local authorities. 61 Other works appearing in English since Kuhn's piece have explored the interplay of Western political ideas and Chinese political concepts, as well as the institutional features of local government operations in the late Qing and Republican periods. 64 Yet it remains easy-because of our reliance on European models-to ignore the larger edifice of political structure and ideology within which local government and social order issues were constructed in late imperial times and which informed the range of possibilities for post-imperial politics. One of the scholars to take the European case most seriously, Prasenjit Duara, analyzes aspects of Chinese local and provincial government in terms of modern state formation. His work has justly received attention for the insights it offers on Republican period politics and society, but he offers no sustained analysis of the central government in a book about the "modern state." In part this is understandable precisely because the late imperial Chinese state offers a starting point for developing a "modern" state fundamentally different from that of the early modern European

R. Bin Wong

political formations. Yet, without consideration of the central state and its relationship to provincial and local levels of government, Duara's discussion of China's state-making process is crucially truncated. More generally, if our analyses of the modern Chinese state are to become more sensitive to the world of political possibilities that exist in a Chinese context that is itself increasingly part of a larger universe, we cannot implicitly or explicitly limit our sense of the possibilities to those defined by a partial distillation of European experiences. The collapse of the Confucian agenda hardly means that it has been displaced by a cluster of political and cultural norms that naturally recognizes European categories such as those of state and "civil society." The continued association of wuguanfang (non-official) with buzhengtong (unorthodox) affirms the difficulties of imagining, let alone creating, an autonomous sphere for the expression of group interests in contemporary China. 65 The variations in late Ming, high Qing, and late Qing strategies to sustain local social order speak to the uncertainties of how officials would value vertically integrated control measures to support centralized government. The more officials sought to coordinate activities at lower levels, the less likely they were to promote effective vertical oversight and control. Since the Confucian agenda, I argue, had a "fractal" quality, this agenda for local order could be articulated on different spatial scales without alteration of its basic ideological principles. In the high Qing, an empirewide scale with a coordinated mix of official and elite efforts made the Confucian agenda an imperial vision. In the nineteenth century, the agenda shrank to provincial and subprovincial units. In Republican times it floated uneasily without a sound mooring on any spatial scale. JiangJieshi's program to persuade rural China to obey Confucian rules of hierarchy and obedience was less than compelling when the positive obligations of elites to act paternalistically were not equally promoted. With the twentiethcentury collapse of the Confucian agenda, a void at once institutional and ideological was created. The successes and limitations of the Communist effort to fill this vacuum have become clearer in recent years. The struggle to create a new vision continues.

CHAPTER 13

Community, Society, and History in Sun Yat-sen's Sanmin zhuyi David Strand

J

£} un Yat-sen's career as a revolutionary bent on destroying the polit-

ical and cultural bases of imperial China and constructing a republic suggests a role as iconoclast and Westernizer. However, despite his use of foreign ideas and organizations, including nationalism and the modern political party, in seeking to act on these principles and use these instruments, Sun was increasingly driven to define his terms and locate his programs in a Chinese cultural context and to sinicize his arguments for Chinese audiences. He was also struggling to move a nation he, and other leaders of the period, often found difficult to grasp and control. From the beginning, Sun relied on Western political theories and models, such as republicanism and democracy, that posited a contractual relationship between state and society. But his final vision of modern China stressed communal ties and familial metaphors. This turn to archaic and nativist thinking did not cause Sun to jettison Western models and technologies. Rather, he sought to give Chinese provenance to Western ideas and modern applications to traditional cultural practices. In the double effort to revitalize and Westernize China, a "narrative of community" became central to Sun's social and political thought.

A Dou or Zhuge Liang? In the last months of his life, Sun Yat-sen made a gesture that reinforced his reputation as a populist and a democrat in the conventional Western sense. Having journeyed north to Beijing in an attempt to negotiate an end to China's civil wars, he insisted that any national "reconstruction" conference convened to settle constitutional issues include a wide variety

David Strand

of groups and associations. He opposed northern warlord-politician Duan Qirui's plan for a council of purely national and regional politico-military elites. In an open telegram to Duan, Sun argued for participation by "representatives of civic bodies [renmin tuanti] such as modern business groups, chambers of commerce, educational associations, university student federations from various provinces, industrial and commercial societies [guilds and unions], agricultural societies, and other groups. " 1 Broad participation and direct civic action would prevent a cabal of militarists and politicians from subverting the renewed constitutional order that seemed in the offing. Despite his willingness to negotiate with Duan Qirui and others like him, Sun had scant respect for the generals who dominated so much of China's national political life. In a series of lectures given in Guangzhou from January to July r924, transcribed and published as the Sanmin zhuyi (Three Principles of the People), he had castigated those whose autocratic ambitions ("the mentality of longing to be emperor" [xiang zuo huangdi de xinli]) promoted civil strife. 2 Sun's commitment to the "principle of democracy" (the other two principles are nationalism [minzu] and the people's livelihood [minsheng]) was rooted in his observation that even anti-Manchu revolutionaries had often displayed imperial longings for power. 3 In one of the rhetorical inversions he was so fond of, Sun had declared that, after all, the reason that a republic had been founded in r9r2 was so that China's "four hundred million" people might be emperor instead. However, in illustrating what this new democratic role involved, Sun took pains to rein in any ambitions for a direct role in governance the populace itself might have. He made it clear that the citizenry as sovereign must surrender power to men of ability and likened the Chinese people to the child-emperor A Dou described in the Romance of the Three Kingdoms, who "has authority but lacks ability. " 4 A Dou, being "simple and stupid" (yongyu) and "without a spark of talent," allows the geniusstatesman Zhuge Liang to govern on his behalf. Sun went to challenge his audience: "Everyone think a minute. As regards political authority [zhengquan], who do today's Four Hundred Millions really resemble? [A Dou or Zhuge Liang?] To my way of thinking, these Four Hundred Millions are A Dou. China today has four hundred million A Dous." Having depicted Emperor A Dou in terms usually reserved for the ignorant masses, Sun endows these hundreds of millions with authority without conceding them a mental life much beyond the Mencian division separating those who work with their minds and those who do not. 5 Critics of Sun would later point to the A Dou reference as evidence that his revolution was

SUN YAT-SEN'S SANMIN ZHUYI

fundamentally elitist in nature. 6 What kind of populist or democratic position can be constructed from such seemingly unpromising material? Sun Yat-sen's late writings and speeches, especially the Sanmin zhuyi, suggest a rejection of an independent role for individuals or groups and serious reservations about anything resembling a free or open society. Sun, finally, was more statist than populist, as befits a politician who in 1924 superintended the conversion of the Guomindang into a Leninist party. But he was also keen to root government in a national community that gives its cultural or communal consent, just as A Dou was supposed to have done. Like many Third World politicians and thinkers, Sun was interested more in telling the story of the Chinese nation as a community facing threats of extinction than in promoting a society organized to empower individuals, groups, or localities against the state. This "narrative of community" was Sun's main concern and stands behind his theory of history and politics.? · Sun Yat-sen was by no means a settled, or even always coherent, political thinker. Sun's apotheosis as ideological founder and framer after his death obscures the fact that he was, in the judgment of one nonpartisan biographer, "not a great thinker. " 8 He lacked originality and rigor, 9 and his writings and speeches are full of contradictions. He ridiculed Western notions of "natural rights" as ahistoricaJ1° and then proceeded to construct his own mythic tale of a distinctively Chinese society emerging from a state of nature. He attacked, as a democrat, the "I am the state" formula of the absolutist kings while in the next breath inviting fellow citizens to participate in "my revolution. " 11 He depicted state power as an affair of force and violence and as the expression of a peaceful devotion to "mutual aid." Sun insisted both that Chinese under the emperors had had too much freedom 12 and too little. 13 His love of analogy, allegory, and metaphor led him to compare government to a machine, a joint-stock company, and a family. Of course, contradictions and inconsistencies might reflect a subtle awareness of the complexity of human affairs instead of simple confusion or carelessness. A recent study of Sun's social and political thought has characterized this taste for the contradictory as the expression of a thorough-going "dualism." 14 Sun's ability to see the state in mechanistic, contractual, and familial terms might suggest a richness of perspective rather than slipperiness. But there is also in his statements a tendency to deal with such contradictions by conflating terms and acting as if real differences did not exist. He insisted, for example, that his notion of "people's livelihood" was the same as socialism even though he had profound differences with Communists and socialists over issues such as

David Strand

class conflictY He attempted to reconcile Confucianism and Communism by simply declaring the existence of affinities between the two. 16 Accepting the loose, improvisational quality of Sun's thought as more an exercise in political rhetoric rather than formal theorizing is perhaps a better strategy for appreciating the significance of what Sun said and wrote. One might expect as much from the Sanmin zhuyi given its lecture formatY In this sense, two large problems conditioned Sun's attempt to persuade his audience in Guangzhou and the wider Chinese public of the value of his leadership and ideas. The first was his precarious political position. As a result of his revolving-door relationship to political power, Sun had become concerned, not to say obsessed, with the problem of imposing order on what he saw as China's political chaos. This obsession with order and unity is evident in his attempt to package Chinese history and culture within a formal structure of ages and epochs and to reconstruct China's population into a political community. Sun's second problem was the need to overcome his image as a Westernized Chinese more in tune with foreign things and ideas than with his own country. His Sanmin zhuyi lectures are loaded with references to such external cultural and technical phenomena as Plato's Republic, textile industry inventions, the American Civil War, obscure European texts, and the policies of foreign statesmen like Bismarck. As a "true son of the littoral," in Paul Cohen's phrase, Sun's career had been devoted in great part to bringing the West to China. 18 As a sign of his power in that realm, Sun seems to have felt obliged to tutor his audience in the ways of this larger world. But, as Cohen observes, Sun's "assault" on "hinterland culture" also required "indigenous validation" in order to win acceptance for organizations like the Guomindang and ideas like the Three Principles. 19 And so Sun was also determined to show Chinese pedigrees for ideas like communism and anarchism 20 and to identify distinctively Chinese variations on universal themes like democracy.

Sand and Rock: The Problem of Social and Political Solidity In the Sanmin zhuyi, the most frequent image of contemporary Chinese society is that of a "sheet [or plate] of sand" (yipian [pan] sansha). In a rhetorical sleight of hand, Sun held out this vivid, disturbing simile of a China in chaos as he simultaneously offered the promise of a unified nation or race (minzu). 21 He even seemed to suggest that the one issues from the other, encouraging, perhaps, a literal-minded inquiry into the country's sand-like state. However, in following Sun's argument it becomes apparent

330

SUN YAT-SEN'S SANMIN ZHUYI

that Chinese are not atomized into separate particles. Once he ventured past metaphor a bit, Sun described Chinese society in quite different terms by focusing on the way individuals are bound to family and clan. Expressed in a style that reflects his affinity for processes involving evolution, development, and changes of physical state, Sun's primary concern was what China could be if it were "solid" rather than sand-like. China would be a rock. If we pick up a handful of sand-regardless of how much-each grain shakes around a lot, unrestrained. This is what a sheet of sand is. If we then add cement to the sand, it forms a rock, a solid body Uiangu de tuanti]. Once transformed into rock, this body is quite solid, and the sand has no freedom [of movement]. So it is readily apparent that if one compares sand and rock, rock is basically something formed from sand. But once inside the solid body of a rock, sand cannot move. It has lost its freedom. One way of thinking about freedom, simply put, is that freedom is the ability to move about within a [larger] body. 22

Sun went on to say that the Chinese are free only in the sense that they have long been innocent of political obligation to the nation. Unlike Europeans who experienced absolutist rule, the Chinese have never been provoked into thinking about liberty in this negative sense. 23 They have freedom without being conscious of it. Sun's definition of freedom as movement within limits imposed by a larger collective body is elegant and could even be construed as liberal. That this model of freedom does not follow from the simple physics of his sand-rock analogy is an indication of where his priorities lay. In China, freedom and a sand-like state mark the absence of solidity and unity. Here Sun was following the lead of Liang Qichao, who earlier declared that "with four hundred million people there are four hundred million countries. '' 24 According to Sun, the Chinese reserve their true commitments and solidarity for family and lineage. Foreign observers say that the Chinese are a sheet of sand. Why might this be? The reason is that the people only have family-ism and clan-ism. They have no nationalism. The rallying power of the Chinese toward family and clan is extremely strong. They often sacrifice themselves in order to protect the clan. For example, members of two feuding lineages in Guangdong will sacrifice untold life and property without any thought of stopping. This is all because of the depth of clan mentality [zongzu guannian]. Because this kind of doctrine is so deeply rooted in the psyche, sacrifice for the clan is ever-possible. In contrast, as regards one's country, there is nothing [comparable] in the form of a spirit of real sacrifice. And so, the rallying power of the Chinese can extend only to the clan and no further. It has not yet reached the nation. 2s

David Strand

33!

The Chinese would actually be like particles of sand if they were not obligated or cemented to a different set of communal structures. This is neither a Lockean world of independent-minded, self-interested individuals nor an anomie, mass society. In Sun's view, it is an uncompleted community in which the absence of the higher stage of national unity renders the Chinese as citizens as weak as if they were mere scattered, individual atoms. In identifying family and lineage sentiments as rivals of nationalism, Sun might have been laying the groundwork for an attack on the traditional family system. Such views were commonplace in the May Fourth period, and Sun has sometimes been associated with such criticisms. 26 Sun might have followed the logic of a Tocqueville who foresaw that societies of atomized individuals formed the perfect basis for statist and nationalist systems. Instead, Sun accepted the permanence of family and clan and, in one of the strangest and most arresting sections of the Sanmin zhuyi, imagined a national polity organized on the basis of lineage. In his fifth lecture on the Principle of Nationalism, given in February 1924, Sun speculated on how a "national body" (minzu tuanti) might be formed from family and clan units. I noted earlier that China had very solid family and clan groupings .... For example, when Chinese meet on the road, after engaging in conversation, they will ask each other's family name. If, as a result, they discover they are from the same lineage, they become extremely affectionate and, sharing the same name, regard each other as close relations. Building on this worthy sentiment, we could extend clan-ism to nationalism. If the nationalism we have lost is to be restored, we must be unified, unified on a larger scale [lit. "have a very large group or body"]. If we are to form a great body, we must first have small foundations. Only from such a foundation can [the process of] uniting together meet with ready success. The foundation we in China have for this purpose is the lineage group. In addition, we have the hometown [or native-place; jiaxiang] foundation. The native-place outlook of the Chinese is [also] very deep. Persons from the same province, county, or village find it particularly easy to join together. As I see it, if we take these two worthy perspectives as the foundation, we can certainly bring everyone in the whole country together. 27

This double focus on lineage and locality is basic to Sun's description of Chinese society. Since he did not expect ordinary people to engage in political philosophy, he needed to identify a psychological or cultural mechanism that would begin the process of combination leading to unity. Sun's search for a group-based foundation for Chinese society placed him in the company of many contemporary activists and social critics. The premise of the group as social and political building block had been

332

SUN YAT-SEN'S SANMIN ZHUYI

advanced decades before by "fanatical 'groupists"' like Yan Fu and Liang Qichao. 28 A February I 9 I 9 article by Li Dazhao argued for a new society and a new world based on small, democratically run units able to coordinate their actions at higher levels. 29 In an essay entitled "A Grand Union of the Popular Masses" ("Minzhong de da lianhe") published in summer I9I9, Mao Zedong had imagined "small unions" like study societies, guilds, labor unions, and native-place associations (tongxianghui) forming the basis of a national polity. 30 Arif Dirlik has pointed out that Mao's essay exhibited a "faith in the efficacy of the 'small group' as a means of social transformation" typical of anarchist and socialist currents in the May Fourth period. 31 The word tuanti itself was flexible and elastic, with a basic meaning of a compact, coiled, and enclosed mass and its micro and macro applications as group, organization, and "great national body" (minzu de da tuanti). Some like Mao sought a kind of social united front of old and new bodies. Chen Duxiu, prior to his conversion to Marxism, had considered a fusion of American-style self-government with China's guild system. 32 Others attempted to create new, utopian societies entirely free of tradition. 33 Although one might ground one's small-group perspective in anarchist or socialist theory, the concrete inspiration for thinking about society in this way was everywhere apparent in urban China. The youth movement and urban nationalism produced highly visible gatherings of such groups in protest rallies or formally organized "federations of all circles" (gejie lianhehui). Each "circle" (jie)-merchant, worker, student, journalist-enclosed numerous unions, guilds, societies, clubs, and associations. The group was much easier to see and appeal to than more abstract, less visible notions of class, mass, and nation. As such, it became for Sun and many others a touchstone for theorizing about Chinese society and a highly practical means of mobilizing this newly imagined society. However, Sun's exclusive focus on family, lineage, and native-place groups sets him off from most other social and cultural critics. In I9I9, Zhang Dongsun, in an essay entitled "Why Must We Discuss Socialism?," observed that "I often say that China does not have a society, just a collection of families. " 34 In that case, some basis other than the family must be found for constructing a new society. For Sun, "a collection of families" and similarly constituted lineage and localist groups were the foundation of choice. Voluntary associations, and a society of associations, require an emphasis on the individual as "unit" appropriate to foreign countries but not to China. In foreign countries, the individual is the [key] unit, and the law regarding the individual rights of parent and child, elder and younger brother, elder

David Strand

333

and younger sister, and husband and wife is such that each is accorded his or her own protection. In the case of a lawsuit, you would not consider the family's situation but only that of the individual. And so the state is a [direct] enlargement of the individual. Between the individual and the state, there is no solid and common intermediary social [ground; zhongjian shehui]. So we can say that in the case of the structural relationship between citizen and state, foreign countries differ from China. Because in China, in addition to the individual, we attach such importance to the family, if there is some issue of consequence, we naturally ask the head of the household. Some say this kind of arrangement is good. Some say it is bad. From my standpoint, the structural relationship between Chinese citizens and the state [guomin he guojia jiegou de guanxi] begins with the family, extends to the clan, and only afterward [becomes] the nation [guozu ]. This kind of organized, stepby-step [in the sense of ascending stairs] enlargement, in perfect order, would be, in the structural relations between large and small, very solid and concrete. 35

With his discussion of individual, family, zhongjian shehui, and the state, Sun ventured into the realm where Western theories of "civil society" take shape. But after looking about, Sun professed to find little of interest. He suggested that in the West the space between individual and the state, imagined by writers as diverse as Hegel and Tocqueville to contain complex strata of associations and contractual relationships, is in fact empty or weak. Sun's "intermediary social ground" is the place where civil society should be if it exists. Sun claimed to see only the seamless, spiritless incorporation of the individual into the state. Despite the cursory style of his investigation of the social basis of the state in Europe, Sun did identify a point of tension in Western formulations of the state-society relationship. Partha Chatterjee, in a recent critique of the application of the "civil society" concept to countries like India and China, has argued that even in Hegel, an author of the concept, there is a deep ambivalence concerning what is missing from this intermediate zone. 36 Chatterjee reads Hegel's insistence on the family as a "natural or immediate phase" of ethical life as a "suppressed narrative of community, flowing through the substratum of liberal capitalist society, which those who celebrate the absolute and natural sovereignty of the individual will refuse to recognize. " 37 Neither the ethical supremacy of the state nor the role of civil society as the supra-familial realm of economic and organizationallife fully compensates for losses incurred. "Hegel's arguments on the family remind us of the irreducible immediacy in which human beings are born in society: not as pure unattached individuals free to choose their social affiliations ... but as already ascribed members of society. " 38 Western theorists, while obliged to identify and accept the rupture between

334

SUN YAT-SEN'S SANMIN ZHUYI

family (or community) and individual, could not help recording the loss of "immediacy" involved in such a transaction. Of course, the figurative use of family and communal ties by the state and other institutions has assumed a crucial legitimating role, encouraging one to find a place in these surrogate or "imagined" communities. 39 At the same time, the pretense of the nation-state as mother- or fatherland can be punctured or set aside by individuals who assume the role of citizens with a political or social agenda. The unquestioned loyalty of citizens to the nation-state (terms fused in patriotic rhetoric) can coexist with the liberal-democratic discourse of state and society (terms decoupled to allow for an adversarial and contractual relationship between the state and individual). While many of his contemporaries were attacking ascribed membership in family and clan as an obstacle to participation in the immediacy of voluntary association and democratic politics, Sun celebrated and extended kin ties while making the individual unit appear as unpromising as possible. At one point, in direct answer to the question "Why can't we govern China?," Sun began a graphic lament about the image the individual Chinese presented to the Westerner by way of "spitting, farting, long fingernails, and stained teeth. " 40 While urging self-discipline and cultivation on his audience to correct these faults, he argued that the individual (especially the ordinary individual) is the wrong place to look for China's strength. Foreigners miss the social power of China because they see only the individual and not the family or group. 41 The implication is that to Westerners schooled in gauging the power of a nation by individual appearances (the typical "Englishman"), China's potential for organization is invisible. This assessment by Sun is in marked contrast to one made a few years earlier by Liang Qichao, who praised the qualities of individual Chinese but complained that in groups his compatriots compared poorly to Americans and Europeans. 42 Where others might see a lack of progress in forging a rational, legalistic world, Sun sensed hidden, invisible strengths based on traditional connections and groupings. For example, he noted with satisfaction that in Chinese commercial practice, where one might expect impersonal, contractual relationships, the presence of "good faith" (xinyi) means "there are no contracts." 43 If that were true, the need for the kind of displacement of social function to impersonal arenas assumed in the concept of civil society would be eliminated, or at least softened. 44 Sun's exercise in comparative cultural analysis barely conceals an attack on May Fourth-era urban political society as empty, shallow, and unChinese. This was an extraordinary stance to take at a time when cities were brimming with insurgent unions, associations, and federations and

David Strand

335

Sun's own reorganized Guomindang had absorbed energy, activists, and ideas from the May Fourth movement. 45 But the spectacle of free association as practiced in republican China appears to have offended Sun's sense of political aesthetics. Pluralism was simply not consistent with Sun's machine-like and architectonic vision of a balanced, functional polity. Groups so constituted had a way of jumping out of line. Students were too disorderly, 46 workers too independent-minded. 47 Merchants, as Sun would be reminded in his clash with the Canton Merchant Corps in October 1924, were among the most effective and stubborn practitioners of the arts of both traditional and modern association. Federalists played into the hands of the militarists. 48 Even his own party on occasion proved unreliable in terms of loyalty. 49 Sun used the same structure of ascending levels of social integration employed by Mao and others to explain how a nation of four hundred million citizens might be put together in a more orderly fashion. But the "perfect order" he hoped to achieve was to be based on traditional feelings of solidarity rather than voluntary association. 5° If we use the lineage as our unit and reform its internal organization to construct a nation, we will have a much easier time of it than foreign countries have had with their use of the individuals as the unit. With individuals as units, in a given country, at the very least one must deal with millions. In China there would be four hundred million. Gathering that many together would naturally be very difficult. But consider making lineages our units. The number of family names in China is commonly said to be a hundred ... at most there might be four hundred .... If we use clans as the small foundations to do the work of strengthening the nation, dealing with China's four hundred clans would be like working with four hundred people. From each family name, using the original lineage organization., we could, in the name of the clan, rally people at the village and county level first and then expand to the provincial and national levels. Each family name would then become a great body. Take, for example, the name Chen. Organized in this fashion, people named Chen would join together at the village, county, and provincial levels. After two or three years, the Chens would have a great organization. Once each clan had such a body, consolidated in comparable fashion, the result would be many huge groups. Once such lineage groups knew of the disaster about to befall us, of the threat of extinction we face, they would all unite into one great Chinese Republican national body [guozu tuanti]Y

Sun counted on fear on the part of lineages at the prospect of the family line dying out being translated into a direct identification with racial threats faced by China. But Sun was seeking more than a turning of lineage-

SUN YAT-SEN'S SANMIN ZHUYI

based solidarity outward to the political community. He also wanted these feelings organized within the matrix of China's administrative hierarchy. And unlike functional groups of farmers, merchants, and workers or strictly local or regional groupings, these clan orders would have no obvious interests to articulate or press aside from their own survival. Sun's fully imagined and articulated political community stood in contrast to and competition with a partially constructed and operational civil society. As Sun sketched out this Chinese gathering of the clans for his audience in Guangzhou, the country's associational life had already moved up in scale from locality and region toward "all-China" (quanguo) chambers of commerce, labor unions, peasant unions, bankers associations, and the like. And these groups of course did have specific interests and claims on the political system. In his 1925 message to Duan Qirui, and similar remarks made a few weeks earlier to the Shanghai press, Sun was prepared to accept this more conventional collection of functional and ideological groups as participants in a national conference. 52 He did so for some of the same reasons of efficiency. Sun regarded Chinese society as a mover might a particularly bulky piece of furniture. Where were the best handholds? 53 It would be easier to work with even several hundred organizations than the Four Hundred Million. But given the opportunity to design his own polity, which he provided himself in the Sanmin zhuyi lectures, he plainly preferred a communal approach, "employing traditional Chinese groups" (guyou de tuanti) rather than the new civic organizations founded in the I9IOS and 192os. 54 The resulting structure "composed of millions of families, would be a great family of the people [dazhong de yige da jiating]. " 55 This preference for traditional groups and networks can be traced to the pattern of Sun Yat-sen's career and the heavy use he made of family and native-place ties. However outlandish the final image of a grand council of Chens, Wangs, Zhangs, might be, Sun's experience of political organization and ascent included working closely with lineage and hometown-based bodies. Some of his earliest political support groups among overseas Chinese in Hawaii were made up almost entirely of people from his home county in Guangdong. 56 His heavy reliance on this kind of informal, local tie, instead of formal examination degree or national elite status (which also, of course, included active use of lineage and nativeplace connections), 57 earlier in his career had marked him as marginal and an outsider. 58 Now he was able to make these original signs of limitation and marginality central to his definition of what distinguished Chinese from foreigners and the indigenous from the littoral.

David Strand

337

In addition, lineage organizations, if not as radically expandable as Sun imagined, had become more flexible during the late imperial period. Thus there was an empirical and practical basis to Sun's seeming flight of ascriptive fancy. Changes in the Song and Yuan periods led to "the development of the true nco-Confucian lineage organization, which was catholic in membership, including junior and senior branches, commoners and officials, all on an equal basis. " 59 The tendency toward ever-larger kinship organizations was such that in 1764 the Qianlong emperor forbade same-surname groups from taking an entire prefecture as their membership base. 60 Fei Xiaotong, in his classic study of traditional or "ruralistic" (xiangtu) Chinese society, cited a "special quality of elasticity" that permitted families to be "very small" or "as big as small kingdoms." 61 As a result, these "highly elastic social circles ... can be expanded or contracted according to a change in the power of the state." The flexibility or elasticity of Chinese social organization encouraged national visionaries like Sun to imagine stretching and shaping the small and the familiar into giant, new and exotic forms and configurations. Native-place ties also could speed and structure the process of forming more modern-appearing leagues, federations, and unions. Vera Schwarcz, in her study of the May Fourth era, notes that "one of the most important sources of solidarity within the student generation was provincial origin." 62 Jeffrey Wasserstrom makes a similar point about the student movement in Shanghai, quoting one activist who asserted that "if you did not join [a native-place association J, you could not get anything done. " 63 By comparison with student movements in Europe and North America, these Chinese "traditional" or "socializing" bodies tended, according to Wasserstrom, to support student mobilization. 64 Bryna Goodman has shown in her study of native-place connections in late imperial and republican Shanghai that place of origin was one of the important and supple criteria available for joining city people together in common cause. 65 Perhaps Sun hoped that by emphasizing the common, deep structure of social organization undergirding student, merchant, and worker politics, he might both restrain and gain control of these formally autonomous bodies. It is revealing that Sun believed that the May Fourth movement itself had been started by one or two individuals who had influenced public opinion and triggered the student movement. 66 If he could tap into mass psychology, a popular concept in China by this time, 67 through principles of his own, Sun might have comparable influence and control. Finally, a community-based argument is most powerful (and controversial) when communities are under siege by an absolutist or colonial

SUN YAT-SEN'S SANMIN ZHUYI

state or aggressive and triumphant capital. In the absence of specific human or natural disasters, local Chinese communities may have been gaining strength, not growing weaker, at the turn of the century. 68 Given traditions of merchant cooperation with literati in the governing of localities, even the capitalist reorganization of major urban centers was led by "business leaders whose geographical, family and professional connections reflected the traditional principles upon which Chinese society as a whole was organized. " 69 China had largely escaped colonization and failed to create a strong state. These circumstances, together with early, indigenous commercialization of the economy, made family and communal appeals of the kind Sun was making a discourse that could run with the grain of social structure rather than against it. As such, it was less a "suppressed narrative" than an open and realistic appreciation of the strength of familial and communal attachments. The idea of a national gathering of clans may have strained credulity. But even so, this tall tale of the Chinese political family ascendant was concocted of realistic sentiments heightened until they approximated the intense nationalistic feelings of Sun's audience. Just as Confucianists had used intimate social sentiments like filiality to explain larger ethical notions like benevolence (ren), so Sun could render the nation more vivid by making kinship and hometown his touchstones. 70 The persistence of community in China made the feelings associated with the nation as political community less a substitute for a lost world of immediacy than an extension, as Sun argued, of intact kin and kin-like solidarities. 71 If Sun suppressed anything, it was the society of associations and groups quite familiar to Sun that he was not sure how to control.

States of Nature and States of History In addition to describing the assembly of a national body or group from family and local components as a spatial and social process, Sun Yat-sen also depicted social and political development in linear or evolutionary terms. He devoted considerable space in his 1924lectures to outlining the ages of human history and demonstrating how his plan for China's future was determined by past events. As Wei Jieting observes in his study of Sun's "social history," "From Sun Yat-sen's standpoint, the whole of human social history is a developmental process of change from simple to complex, chaos to civilization, ignorance to knowledge, and depths to heights. " 72 His four epochs, in ascending order, are prehistoric chaos, theocracy (shenquan), monarchy (junquan), and democracy (minquan). 73

David Strand

339

Sun formally rejected Social Darwinism because of what he judged to be the negative implications of the survival of the fittest principle for social organization within a polity. 74 But he found the "apes to men" idea appealing and larded his account of human history with a naturalist emphasis on origins and the survival of useful traits. In addition, given his acutely racialist interpretation of competition among states and peoples, Sun was prepared to see the separate development of China and other countries in an evolutionary light/5 History for Sun traveled at two speeds: the rapid pace of material progress and the much slower development of political and social ideas. This second, slower rhythm allowed him to insert ancient Chinese ideas into contemporary discourse about politics. His evolutionary perspective encouraged him to see these old cultural predispositions as having a kind of genetic hold .over contemporary history. Sun presented these continuities in part as a matter of "great ideas" still valid today. For instance, he cited the continued value attached in the West to Plato's Republic as an example of the way in which political philosophy lags behind material change.76 If Plato, why not Confucius? But Sun, in this part of his argument, was interested less in individual greatness itself than in the genius of the race. Sun, as is well known, liked to imagine China pulled together by railway systems and electrical power grids. But for the "invisible machinery of politics" capable of "controlling society," one needed to rely on psychology and take into account basic and ancient cultural differences.7 7 According to Sun, the key Chinese psychological and culture difference appeared early on. In Sun's first stage of history, in which "people fought with wild beasts," Sun already detected a uniquely Chinese leadership principle, even though "the state's organization was not yet complete" and people still lived in clans. 78 Whereas elsewhere in human history these leaders were invariably fighters, in China such chiefs or kings were men of cultural and technical attainment. There is no explanation for this departure save the special nature of the Chinese mind emerging from Sun's quasi-historical state of nature in which humans fought wild beasts with their bare hands. The resulting "psychology" possessed by Chinese causes them to "believe that men of great leadership capacity should be emperor. " 79 This retrieval of ancient sentiments and mentalities fits Sun's criticism of Chinese "drunk on new culture." 80 These iconoclasts fail to understand the hidden reserves possessed by China because of its unique historical experience. Not suprisingly, Sun uncovered communal values capable of trumping Western-style claims to individual or group autonomy. 81 These

340

SUN YAT-SEN'S SANMIN ZHUYI

values will make it possible to organize China on a "mutual aid" or cooperative basis radically different from liberal economic and political models. 82 Despite Sun's oft-stated hostility to anarchism, 83 a challenging alternative to many of his own positions, he owed a great deal to Kropotkin's social anarchism and the notion of a natural, human basis for mutual aid. 84 According to Sun, the capacity for mutual aid distinguishes humans from other animals, creatures wholly driven by Darwin's principle of competition. 85 Although animal competition remains a residual feature of human behavior, mutual aid functions as the motive force in human development. 86 Sun also seems to have taken to heart the anarchist assessment of the state as essentially violent. He saw politics as a double thing: natural and pacific in terms of the emergent nation (minzu) and artificial and violent in matters of state (guojia). States are founded through violence. Nations emerge as "the product of natural forces. " 87 China is unusual because the nation as a unified people (dated pre-Qin) early on gave birth to a state. Unfortunately, this precocious joining of nation and state was lost when China's sense of minzu was overshadowed by Confucian empire-builders who prized cosmopolitanism more than racial solidarity. 88 Later conquests by the Mongols and Manchus created a state based on force. Despite Sun's formally historicist approach, the particular (and controversial)89 twist he gave to Chinese history allowed him to reclaim the original naturalism of the nation (community)-state relationship. This goes against the grain of the universalizing global forces propelling China toward democracy. But Sun was willing to give China dispensation in this regard by assuming the separate working out of internal and external historical forces. Events like the replacement of the Manchu monarchy by the Republic resulted from the requirement that Chinese history align itself with world history, as iron filings line up within an enveloping magnetic field. But China has its own internal, natural impulses that operate independently and lead to the possibility of a special kind of Chinese democratic state. And so, just as a national political "body" completes the promise of basic-level kinship and community organization, Western forms and theories help recover and complete ancient impulses and practices. In one of his most theoretically oriented essays, "Action Is Easy and Knowledge Is Difficult," Sun discussed elements of culture and social life such as food, medicine, language, and money in order to show the relationship between the evolved genius of the Chinese race and the received scientific

David Strand

and theoretical wisdom of the West. 90 According to Sun, European and American cooking cannot match Chinese cuisine for quality, variety, and subtlety, but the West has theories of diet and nutrition that permit scientific analysis and management of food production on a huge scale. The Chinese have green tea, soy sauce, dried lilies, wood ear fungus, bean sprouts, and, above all, bean curd, a food with significant patriotic connotations.91 The West has Louis Pasteur and the biological sciences. Beancurd as the "natural" product of Chinese culture and theories of nutrition as the outgrowth of Western science together produce, among other things, the scientific validation of beancurd as an ideal food (and, incidentally anarchist Li Shizeng's Paris Doufu Company). 92 As a result, "the culinary skills of the Chinese and [Western] science and hygiene are in serendipitous agreement [anhe]. " 93 In analogous fashion, China has long had language without linguistics, money without a theory of finance, and alchemy without chemistry. With the addition of theory, these cultural achievements can be harnessed to new purposes that improve on nature, including the human body. The ancients have said that a human being is a small world [xiao tiandi] ... [but] I prefer to say small state [guojia ]. The various internal organs function as parts to serve the whole just as the various departments of the state help govern the political affairs of the nation. The organs of the body are complete as organized and nimble of movement, but they come nowhere near what today's state is capable of achieving. 94

Comparing the modern state to the human body is less surprising than asserting the superiority of the one over the other. This was Sun's way of evoking the power of modernity as an artificial, mechanistic force that enhances rather than obliterates the pre-existing natural and social world. In like fashion, the modern state can enhance, rather than supplant, ties based on family, clan, and native place. The natural, or communal, body yields to a political or mechanical body armed with new powers. Culture is put to the service of the state. Following from Sun's insistence that "action is easy and knowledge is difficult," these new theories are not only Western but, more important, esoteric, a realm of knowledge closed to ordinary people. The modern A Dou eats, and perhaps makes, beancurd as he always did. The modern Zhuge Liang manages nutrition and diet for the entire nation based on scientific knowledge, as Sun admiringly said the German government did during World War J.9 5 This kind of governance depends on the masses remaining politically passive and culturally active, leaving elites free to manipulate the world to the common advantage. Sun downplayed the

342

SUN YAT-SEN'S SANMIN ZHUYI

possibility of culture and science clashing, as happened, for example, later in the republican period when municipal authorities in Shanghai sought to vaccinate residents, some of whom feared the injection of foreign (in two senses) substances into their bodies. 96 Instead he predicted the continued acceleration of knowledge-based change ("the last hundred years already supersede the last thousand, the last ten years the previous hundred") until, guided by the principle of mutual aid, "today's bitter world is transformed into a blissful heaven. " 97 In taking steps like these toward "indigenous validation" of his modernist political theories, Sun also focused on the conventional distinction between "enlightened rule" (wangdao) and "rule of force" (badao). 98 "Chinese say wangdao follows nature. In other words, wangdao is a natural force. The nation is a body created through the use of wangdao. Badao is force of arms. The state is a body created through the use of badao." 99 By Sun's account, the nation, which precedes and creates the state, is itself created through wangdao, a quality conventionally associated with kings rather than the people. Long before the people became "like A Dou," they were A Dou in that the community of Chinese was endowed with moral and racial authority. The concrete form this took included the wangdao-like principles of cooperation and mutual aid. In this spirit, Sun later defined "society" ahistorically as "the structure of mutual aid" (huzhu de ti). 100 He also claimed that local communities in China have "naturally endowed democratic ability" even if "they have not attained the perfection of Western governments. " 101 Sinking this moral quality deep into Chinese society (as well as history) was another one of Sun's ways of positioning the Chinese people to have what he felt was the proper democratic relationship to the state. At the same time, as suggested by Franz Fan on in another context, this turn toward a timeless national culture promised the "secure anchorage" denied to, but often craved by, the cosmopolitan. 102 Sun excoriated the literati (wenren) for being co-opted by the Manchus and took this as a sign of how cosmopolitanism eventually makes the Chinese elite unreliable as representatives of the nation. Instead, the "jewel of nationalism lay deep in the basest levels of society" as secret societies carried on the anti-Manchu struggle "by word of mouth rather than by writing. " 103 The way of the ruler becomes the moral practice of the community. The civil commitments of the elite become the province of the illiterate. Here is a kind of populism informed by anarchism but directed at the fusion of nation and state through communal rather than contractual means. 104 Sun took considerable pride in revealing invisible group strengths obscured by hostile critics of Chinese culture and locating hidden reserves

David Strand

343

of national identity buried in the past. Whereas a May Fourth skeptic might strip away culture and ideology to reveal the controlling power of interests and politics, Sun stripped away politics to reveal and revitalize governing cultural principles. Conclusion Some scholars have pointed to "Sun Yat-sen's bottom-upwards model of political development, with its vibrant populist tone and 'self-government' prescriptions" and the failure of later regimes to follow through on this vision. 105 Others have seen in seemingly "pro-democratic" writings like Sun's "a far stronger preoccupation with national goals and deep-rooted cultural attitudes unfavorable to democracy." 106 In considering the two narratives-community on the one hand and civil society on the otherSun clearly favored the former and in a fashion that seems to preclude most liberal forms of democracy, despite his formal support for elections, referenda, and recall ballots and his stated belief that authoritarian forms of "political tutelage" were temporary and transitional. 107 On a visit to America in 1842, Charles Dickens observed that "the mass (to use our monarchical term) are miserably dependent in great things and miserably independent in small things. " 108 Sun might have agreed with the latter part of Dickens's formula, given his account of the empty "freedom" enjoyed by Chinese in their communities. What he hoped to do in contrast was make the Chinese people happily dependent in great things by getting them to accept their ascriptive status as members of the Chinese nation. His plan to leave lineage and local community intact would grant depth to his imagined political community and ensure the survival of the natural moral structure he found so compelling and so Chinese. Of course, in the absence of his gathering of the Chinese clans and given his distrust of May Fourth-style associational politics, he chose the "bottom-downwards" logic of Leninism as the structure to connect the small and great worlds he spoke so vividly and passionately about. As Myron Cohen has recently pointed out, this was a prescription for demoralizing the small, familial, and local elements in Chinese society and rendering shallow and superficial the great minzu imagined by Sun and his contemporaries. 109 Political leaders of Sun's era faced fateful choices as they contemplated the ascending structure of social and political organization in republican China. If they left standing what Keith Schoppa has analyzed as the "primary" structures of social and politicallife-"kin, patron-client and

344

SUN YAT-SEN'S SANMIN ZHUYI

brokerage relationships"-higher levels of "composite and complex political structures" like parties and states might remain hostage to local, personal, and informal forces. 110 This summoned up the long-standing statist (junxian) fear of local society as a "tail too big to wag" (weida budiao). 111 The junxian perspective had been supplemented by a localist or "feudal" (fengjian) view that stressed elite-led community mobilization linked in common cultural cause with the project of empire. 112 But as Prasenjit Duara has recently shown, by the turn of the century, the fengjian tradition had been overpowered by the aggressively statist narratives of nation-builders. 113 Sun sensed a powerful resonance between local communities and the nation-state, but a mechanism for fusing or, better, balancing the two eluded him. Sun missed or ignored, because of his national and visionary role, real possibilities for reconciling the two narratives of community and society at the local level. Since at least the post-Taiping reconstruction, old and new local elites and their mixed constituencies of communal and associative groups had been working to maintain such a balance without, however, curing the miseries associated with their uncertain independence. 114 Although Sun defended his choice of Leninism as a matter of picking the best organization for carrying out his "doctrine" (zhuyi), that doctrine itself went a long way toward preparing primary structures like the family and the native place for A Dou-like submission to the party as the modern embodiment of Zhuge Liang. As for civil society, associational life was merely a stepping-stone, to be solidified (jiangu de tuanti) in the interest of central control. Sun admired modern society for its efficiency rather than for its political boisterousness. The noisy republic he operated in, with its chanting protesters and wickedly ironic editorialists, required "solidification." Sun encouraged journalists to organize in groups to support his political efforts 115 and criticized them for their "habit of attacking the government. " 116 By privileging state interests and discourse, Sun can be seen as helping to lay the ideological groundwork for severe restrictions on his era's emergent public sphere. A "dialogic community" in which citizens could and did criticize their government and communicate with each other was not compatible with Sun's republic of A Dous. 117 The Madisonian principle of extending the community to the nation and grounding the enterprise in common values keyed to citizenship is quite close to Sun's depiction of the Chinese republic. 118 But since Sun's citizen is constructed almost entirely of cultural tendencies and psychological characteristics, the sense of agency required for dialogue is missing.

David Strand

345

This is not to say that this kind of narrativizing of community at the expense of civil society left Chinese citizens completely defenseless against the regime. Citizens can be made to act like A Dou only as long as leaders appear as capable as Zhuge Liang. Sun's government by moral or communal consent required leaders to demonstrate their command of the esoteric and their ability to move the nation. Predictably, failure in these areas of competency could breed contempt consistent with family-like immediacy. If the family and small group could be stretched by leaders to the scale of the nation-state, the regime could be shrunk in the public mind to the dimensions of a fractious, poorly run household. What ideology promised to solidify, social irony continues to dissolve. Studies of popular protest in China from the May Fourth movement on have noted the ability of Chinese citizens to turn celebrations of the authoritarian state into ironic demonstrations of the state's incapacity, from official student visits to Nanjing in 193 I suddenly transformed into protests against the Nationalist regime to the 1989 funeral of disgraced leader Hu Yaobang that sparked the democracy movement. 119 In the latter case, in a new inversion of Sun's people as emperor analogy, student protesters depicted the imperious Li Peng as a "very stupid" A Dou.l2° We are left with Sun Yat-sen and his thought caught between a commitment to democracy undermined by hostility toward individual and group autonomy and a tendency toward authoritarian politics restrained by a romantic faith in a republic based on cultural consent. The result is a mix of elitist and populist sentiments-an aesthetics of domination in which every Chinese in and out of power is both potential supplicant and critic.

CHAPTER 14

Constructing the Civilized Community Ann Anagnost

07' hang Yimou's film Qiu Ju da guansi (The story of Qiu Ju) is the tale ~ of a young peasant woman's unrelenting search for legal redress of what she regards as excessive abuse of power. The village head (cunzhang) has kicked her husband in the groin, temporarily incapacitating him for labor and arousing concerns about his future fertility (given the ever hopeful prospect of a change in the birth policy). The injury is minor, but the question of adequate compensation looms large for most of the film, as Qiu Ju pursues her grievance at ascending levels of political authority. The case is settled at each level in her favor and yet she fails to feel satisfaction. The reasons for this are clear. She does not seek financial compensation for her injury; rather, she wants a public avowal of where the limits to power lie. In Zhang Yimou's own words, "What Qiu Ju wants is shuafa [sic]-a word used in the film that does not mean an 'apology' but an answer, an explanation, a clarification. " 1 Were the village head to recognize this demand for accountability, he would "lose face" (diu mianzi) and therefore his authority to govern in the still highly personalized modality of power in reform-era China. The efforts of mediators at every level to "save face" for the village head by persuading him to give Qiu Ju a monetary compensation are consistently derailed by the village head's own intransigence. He throws the money at QiuJu's feet and tells her that she must bow her head to him as she bends over to retrieve each bill, a movement that would force her to ape the ritualized subordination of the ketou ("kowtow"). This humiliation she refuses. Only when the village head redeems himself at a moment of need does Qiu Ju feel satisfied, but not before the ponderous gears of the legal system have been set in motion and lead to his arrest. The legal

Ann Anagnqst

347

means of redress, therefore, are never appropriate to the situation; they are either insufficient or in excess of what is desired. Qiu ]u does not obtain satisfaction in the form she wants. The village chief's freedom to strike a citizen is never challenged by the legal system. Only when it is proved that the blow inflicted a documentable physical injury does the legal apparatus kick into action in a dazzling display of punitive force. The question of power and its limits is thus effectively reduced to an economy of injuries and compensation. Qiu Ju's story offers a pretext for exploring the articulatory practices of the state from the perspective of narrative address. Zhang Yimou's film attempts to solicit a recognition in his audience of a different kind of relationship to power in the reconfigured community of reform-era China by providing a model of agency workable within the constraints of the present political system. This model highlights for us, if only by contrast, important questions about how the party-state "addresses" the mass of Chinese citizenry. More specifically, my argument focuses on the promotion of village compacts (xianggui minyue) drafted by township and village governments in the post-Mao period. These compacts attempt to reinstitute a mode of normative rule by means of an economy of "face" in communities threatened by the disintegration of collective organization and the new mobilities of post-reform rural society. Although these compacts bear some resemblance to the xiangyue of the late imperial period, the question of historical antecedents is a vexed one and must be put into the context of the "heated" debates about Chinese culture and popular sovereignty beginning in the mid-198os. The question at the heart of these debates is whether the Chinese "past" is a valuable resource for the construction of a modernity specifically Chinese or a weighty burden that stifles the progressive movement of China in history. The terms of this debate suggest a politics of representation implicit in the revival of the village compacts. The official promotion of these compacts as an institution of "self-government" imaginatively positions the rural masses as both their author and their object. Indeed, the official discourse on the compacts effectively confuses Homi Bhabha's distinction between the pedagogical and the performative aspects of subjectivity. The "masses" in effect are subjected to a pedagogical practice that they, as active subjects, are said to be the authors of.2 Bhabha's formulation replicates at the level of a national body or subjectivity Foucault's dual sense of the subject, both as "subject to" a heteronomous rule and as an agent "tied to his [sic] own identity by a conscience or self-knowledge. " 3 What Bhabha calls the "pedagogical" refers to a set of practices and discourses

CONSTRUCTING THE CIVILIZED COMMUNITY

that calls into being a subjective identification of individuals with a national body, a citizenry. The "performative" refers to the situation in which this national body begins to take on an agency of its own, an agency that does not necessary derive entirely from the discourses that formed it. These are, of course, idealized locations. Subjectivity never lies on one side or the other but in the uncertain oscillating space between them. The compacts are primarily pedagogical in promoting a normativized vision of the ordered community, but they do so under the guise of the performative. They are thus akin to the power of constitutional reforms as a mode of what Duara has called "writing the people into being," the creation of a subjectivity in the very act of "representing" one. 4 The story of Qiu Ju disrupts this circular production of subjectivity by inviting a critical gaze at the highly personalized modalities of power exemplified by the ironic slippage in Chinese between "personal power" (renquan) and "human rights" (renquan). The film is all the more pregnant with this irony in its apparent depiction of the operations of the legal system as operating exactly as it ought, with officials fulfilling to a heightened degree the Communist party ideals of earnest and self-sacrificing service-so much so that no less a personage than Li Ruihuan, the head of the Propaganda Ministry, praised the film as one worth advocating through the government propaganda apparatus! 5 Indeed, the film's exploration of power as it operates according to its most idealized self-representation explicitly invites the Chinese "masses" to question the everyday exertion of power and to imagine their own agency in the political process. The Village Compact The village compact, as a model for community reform, lies at the heart of the issues of popular sovereignty and history raised above. 6 It is a "split" object, simultaneously old and new, a quintessentially "Chinese" approach to the social order and a novel instrument of "democratic reform." This splitness suggests the working of ideology, the suturing over of an undecidable contradiction within the party's own discourse. The recent history of the village compact associates it closely with the post-Mao reforms. A handbook on propaganda work identifies it as part of the Civic Education movement (wujiang simei) that began in 1982. 7 In that same year, a People's Daily commentary marked the beginning of a national campaign to draft village compacts. 8 However, this commentary accompanied a report of a model brigade in Yunnan province that had drafted a compact in 1981. 9 In at least one other case, a village compact

Ann Anagnost

349

in Fujian was described in the early 198os as having a twelve-year history. 10 In any case, from 1982 on, the village compact was quickly incorporated into a succession of officially defined models for local political and social reform. Most ubiquitous of these is the "civilized village" model (wenmingcun). As early as 1982, the drafting of a village compact was listed as one of the ten criteria for achieving model status in a "civilized village" campaign in Hebei province.U More recently, the village compact has been associated with the "construction of socialist spiritual civilization," beginning in 1986.U These compacts also figure importantly in recent campaigns to increase peasant understanding of the law (pufa jiaoyu). In 1991, in the course of my fieldwork on the "civilized village" in areas outside Nanjing, Wuxi, and Kunshan, I discovered that village compacts were not only still in effect, but had, in the period since 1989, undergone radical revision to address some of the new problems emerging from the heightened pace of the economic reforms in the late 198os. 13 The village compacts give quasi-juridical form to the definition of the model of the "civilized village." They provide a punishing specificity to areas of everyday life that are too trivial or too pervasive to be dealt with adequately by the state legal apparatus. They have therefore a quasi-legal status in that they are intended to address issues of social order that are beyond the scope of the law or the present capacity of the legal system. The discouragement of "unhealthy practices" such as feudal superstition or gambling, as well as issues of social order, family quarrels, interpersonal disputes, the fulfillment of state contracts, vandalism, and petty theft, are among the issues addressed. They are supposed to be drafted by the people themselves, or by the newly constituted village committees (cunwei), and indeed they are often touted as a model for self-government, as a basic democratic process of the masses. But it is also clearly evident that village compacts are quite often initiated and implemented under the leadership of local party organizations and personnel. 14 The ambiguous status of the village compact is reflected in the effort made to clarify what differentiates them from law. In a handbook for disseminating legal knowledge in the countryside, the line between them is defined with great care. It states that such compacts (gongyue) or regulations (shouze) are stipulated by the twenty-fourth article of the constitution as a "democratic form" in the construction of "a socialist spiritual civilization." They are drawn up by the people themselves, not by a national legislative body. Their jurisdiction is limited to the community that drafted them, and their "binding force" (yueshuli) depends on education and criticism, supplemented when necessary by economic sanctions. 15 Whereas

350

CONSTRUCTING THE CIVILIZED COMMUNITY

the law depends on the guarantee of coercion, the compacts must depend on the force of public opinion and social consciousness. The village compacts are therefore distinguished from state law, in terms of their institutional status and their enforcement, and yet when party policy becomes promoted through this extra-legal means, such as stipulating penalties for out-of-plan births, it blurs the boundary between politics and law. This boundary is further confused when village regulations appear to function as a replacement for legal procedure to keep the immature legal system from becoming overloaded or even to substitute for it in places where legal institutions do not yet exist. 16 Although the compact is often described as democratic in nature, a voluntary contract, and a means of "self-government" and "self-education," in fact, it has become an important sign that the construction of a socialist spiritual civilization has been "grasped firmly" (zhuajin) and therefore figures importantly in the accumulation of political capital at a number of levels in the party organization. At the same time, however, the party also sees the compact as a means for restoring its damaged relationship with the masses (dangqun guanxi). Therefore, although the initiative may come from the party organization, the responsibility for drafting and implementing the compact is often delegated to the village committee (cunwei) or shared with persons of exemplary moral standing in the community outside the party organization itself. The surveillance that these compacts entail is thereby represented as coming not from the party but from the "people." This takes institutional form in the localized proliferation of new organizational entities created to enforce the compacts as well as improve the general moral atmosphere of the community. These new bodies are variously titled "social ethics appraisal committees" (shehui daode pingyi hui), "village people's educational activities group" (cunmin jiaoyu huodongzu), "councils on weddings and funerals" (hongbai xishi lishihui) (to discourage ostentatious family rituals), or, simply, "civilization committees" (wenming zu). These groups are generally composed of a mix of party personnel and persons of communally recognized high moral standing, selected under the direction of the local party organizationsY In cases where this wider participation is actively practiced, one might speculate that this more participatory mode might have an important moderating influence on the ability of local officials to abuse their power. However, in practice the compacts can be quite punitive, giving local party leaders tremendous power in implementing policy. For instance, they are used widely to enforce local birth quotas by exacting fines for

Ann Anagnost

out-of-plan births. And yet, these fines may vary wildly due to their localized character. Cadres who are overly dependent on heavy fines to enforce village regulations are criticized in the press as unimaginative and lazy. Education and persuasion are the preferred means of enforcement, but require time and energy in pursuit of uncertain results. Moreover, one suspects that the fines often do not really act as a deterrent, but as a tax on out-of-plan births. The unabashed readiness of those who have to pay for their extra children leads one to suspect that the ability of more prosperous households to pay the fine may be a new form of conspicuous consumption in the countryside. This suggests new dimensions of resistance to the allocative power of the state in which its penalties become the basis for status enhancement. Because of this capacity for abuse, local compacts do come under scrutiny from higher levels of the party organization, especially in cases where abuses of power have led to confrontations or other legal problems. The dimensions of this abuse were compellingly described in one especially detailed commentary. 18 The use of physical coercion to enforce the compacts was deplored. A minority of cadres "with a weak concept of the law" were described as using the charters for arbitrarily seizing or imprisoning people, searching their homes and confiscating their property. The implicit sanctioning of cadre use of corporal punishment was also censured. And yet, in one township charter, the fine for striking a cadre was 20-50 yuan, but there was no provision for punishment when a cadre beat a village resident. 19 Overdependence on fines was also deplored as were overly punitive measures. The editorial cited a compact from one township in which 29 articles out of 50 stipulated a total of 30 different fines ranging in amount from r to 4,ooo yuan. 20 A village resident who cut thatch in a neighboring village was to be fined 5 yuan. If the fine was not immediately forthcoming, another fine called an "attitude penalty" (taidu kuan) of 300 yuan was added. Moreover, in what appears to be an almost incredible display of punitiveness, the malefactor was required to treat 40 persons to a banquet; and his pig, sewing machine, bicycle, and desk clock were to be confiscated. One wonders how often such extreme sanctions were enforced or whether their mere excessiveness "on the books" was intended to intimidate possible offenders who might otherwise disregard them. And yet, in at least one case reported in the same editorial, sanctions that were far less extreme provoked an act of ultimate protest. A peasant who had illegally cut two trees was fined So yuan, according to village regulations. Reportedly, he was unable to pay this amount and hanged

352

CONSTRUCTING THE CIVILIZED COMMUNITY

himself instead. His death brought his case to the attention of higherlevel authorities, who then arrested the local officials held responsible for his suicide. 21 This case was used to illustrate that village cadres themselves are subject to the rule of law when they overstep the limits of their power (ill-defined as they are). And yet, this case also begs the question of why this unfortunate peasant had no legal recourse other than to hang himself to bring his wrong to the attention of the higher authorities! This example beckons toward Zhang Yimou's observations about the lack of avenues of appeal: "[The Story of Qiu Ju] is a very ordinary story that happens all the time in Ci-.ina. One never knows who to talk to, what to do, where to go. Most problems are not so bad to start with, they only become so because of the workings of the bureaucratic system and the ordeals you have to go through. " 22

The Norm as Model Undoubtedly, the norm is related to power, but it is characterized less by the use of force or violence than by an implicit logic that allows power to reflect upon its own strategies and clearly define its objects. 23

The village compacts represent the party's will to reform everyday practice, to ensure the extension of its norms throughout society. And yet this disciplinization is represented as the party's retreat from its overwhelming presence in everyday life and its replacement with a more participatory local politics. The compacts are marked by a concern for ordering social life not as an end in itself but to channel and magnify an ethic of production. And yet these goals must be accomplished at a time when rural communities are themselves undergoing dramatic transformations. In this sense, the re-creation of the ethical community does not mean the continued existence of the closed community encysted within the collective organization of the Maoist era. But with the opening up and fragmentation of these communities, the newly "liberated" economic actors must now be subsumed within the ethical community as it is promoted in a more global sense. If the face-to-face community is to continue to be a binding force in society, it can no longer be phrased in terms of the "little community," but must be seen in a more universalizing ethic that continues to classify individuals within the architectonic spaces of a moral geography that may not exist outside the discursive construction of the party's ideological practice. The compacts are, in this sense, a disciplinizing

Ann Anagnost

353

intervention at a time when the organization of labor and production have been otherwise dispersed in the encouragement of a market economy. In this sense, the village compacts would seem to fall within the definition of biopower, which Foucault describes as "a power bent on generating forces, making them grow, and ordering them." This is a power that contrasts with the murderous display of juridical force that "seizes things, time, bodies, and ultimately life itself," preferring instead "to incite, reinforce, control, monitor, optimize, and organize the forces under it." It contrasts the power of the sovereign and his "power of life and death" with the rationality of the modern state as "managers of life and survival. " 24 Foucault distinguishes two poles around which biopower organizes itself: the body as a machine that can be rendered docile through the disciplinization of work and of everyday life and the population as a body politic that defines the rationality of the state in its efforts to improve its "quality," by regulating and intervening into its biological processes. The post-Mao Chinese state consciously aspires to this regulative ambition, perhaps more than any other modern state. One need only look at China's birth policy to get a sense of the scale of this ambition in which "the large-scale phenomena of population" becomes the site where power is exercised.ZS It is clearly expressed in the party's determination to "improve the quality of the people" (tigao renminde suzhi), which suffuses its educative efforts. Indeed, the party's discourse on the compacts in the early 1990s clearly locates them as embedded in this larger project of national improvement. Foucault suggests that the development of biopower resulted in the growing importance of the norm over the deployment of juridical power. "Such a power has to qualify, measure, appraise, and hierarchize, rather than display itself in its murderous splendor." 26 The village compacts incorporate practices of surveillance and evaluation; their interventions into everyday practice are disciplinizing. And yet the way in which the norm is constructed here differs from the disciplinary procedures that emerged out of the development of capitalism in Europe. 27 The existence of a norm implies a standard of measure or evaluation, but the function of the normative in Foucault's sense is the distribution of values around an average. However, this definition of the norm is one of relatively recent development. Ewald has detailed its origins as the ideal medium of communication in the homogeneous space-time of capitalist production. The norm is "the means through which the disciplinary society communicates with itself. " 28 It engineers the "controlled insertion of bodies into

354

CONSTRUCTING THE CIVILIZED COMMUNITY

the machinery of production. " 29 And yet this was not always true. Until the beginning of the nineteenth century, the norm implied a "principle of valorization" directly linked to the notion, not of the average, but of rectitude. In the transformations of industrial capitalism, the play of oppositions specified by the norm shifted from an ideal of rectitude and its antithesis to the "normal" versus the abnormal or the pathological.3° Ewald draws attention to how this "law of the average" gives rise to a positivism that is both objectifying and emptied of signification. [Facts] do not signify, they simply are ... they are comprehended solely in terms of probability, the statistical category. The statistical category operates as a system of classification rather than as an identifying denomination. The average man is a "fictional entity," not a real individual, nor a model or original that serves as the standard for all men but the reference point common to them all. ... He is society itself as it sees itself objectified in the mirror of probability and statistics.Jl

However, in China, the rule of the norm operates within a system of moral signification in which the norm is predicated on a common standard or model of rectitude, rather than an average. In the norm as model, categories are not a neutral system of classification but identifying denominations that signify values in a moral system that is prescriptive rather than normative. 32 In much of the literature defining the construction of a socialist spiritual civilization, an earlier opposition between zheng ("rectitude") and wai ("heterodoxy" or "crookedness") is refigured into the postMaoist narratives of the model. These narratives relate transformations of entire communities to "civilized village" status (wenmingcun) from a disvalued state-a "quarrelsome village" (shifeicun), a "war chaos village" (zhanluancun), or a "pirate village" (zeicun). The intensity of these designations clearly puts the normative value of what is "civilized" into a system of moral signification. In noting the differences between these two constructions of the norm, I am not trying to argue that one is more "modern" than the other but merely taking into account their sociohistorical specificity. Nor am I trying to make an argument for an essential "Chineseness" intrinsic to this way of operating, despite strikingly similar technologies of control in the Chinese past. Indeed, its present deployment is self-consciously constructed as a "Chinese" solution to social-order issues responsive to China's "special characteristics" in its course of economic development. Both these caveats are all the more necessary in that the specific modes of discipline I describe here are becoming superseded in certain discursive domains, especially in the Chinese academy, where Western modes of disciplinary practice are

Ann Anagnost

355

rapidly gaining sway in the importation of a highly empirical, positivistic Western social science theory that represents itself as the indispensable condition of the "modern" and therefore empowering. 33 If the village compact fits within this Chinese conception of the norm as model, how do we interpret its functioning as a disciplinary practice of the modern Chinese state? Does it merely bear the imprint of a Chinese tradition, the heavy weight of which forms an almost insurmountable obstacle toward attaining "true" modernity? Or can we allow it to possess a modernity that is "other" to the Western model? To the extent that such practices are identifiably "Chinese," they may well be a nativist rejection of Western modalities of power, a way of asserting a "Chinese" modernity that rejects the disempowering effect of a historical teleology in which "difference" always means "backwardness" when compared to the West as a universal standard of modernity. And yet this difference between East and West is intricately intertwined with the opposition between capitalism and socialism and the latter's construction of the role of the regulative state as a superordinate rationality. At the same time, if one wants to argue for the modernity of the power of the Chinese socialist state in terms of its similarity to the Western model, one must beware of assuming its operation as precisely the same. At first glance many of the practices of the modern Chinese state look very similar to the practices of the Western disciplinary state. But are they the same? And must they be, in order to qualify as "modern"? The village compacts apparently operate in ways similar to Foucault's concept of the panopticon, which distributes individuals across a grid and exposes them to a central gaze. This grid is composed of the decimal units of surveillance and control, the creation of bureaucratic categories marking degrees of compliance that isolate the objects of the party's normalizing practice. However, there are some important differences. Foucault has drawn attention to biopower as the quintessence of the modern disciplinary state. The norm exercises its influence invisibly and anonymously as a "natural" state of being widespread in society, the abnormal is an "unnatural" state against which the technologies of normalization must be mobilized. In China, the exemplar or model, not the average, is the basis for these practices. The entire social body becomes the raw material to be worked on by the projection of a conscious will to transform society from its present state of backwardness, of lack. The opposition between "civilized" and "uncivilized" sets the terms for a projected transformation of society as a whole (rather than the normalizing of aberrant individuals,

CONSTRUCTING THE CIVILIZED COMMUNITY

even when these technologies are applied directly to individual subjects). Power is not disseminated in quite the same form as described by Foucault for the modern European state. It is not anonymous, invisible, and silent, but personal, extravagantly visible, and accompanied by the hushed buzz of voices or the clamor of gongs and drums. The disciplines of the bodyits bearing and demeanor, its productivity and reproductivity-are not influenced through the application of invisible and anonymous forces through the orchestration of the rhythms of capitalist production, although these modes of discipline may increasingly be found in the industrial sector of the Chinese economy. They are applied by means of a direct "address" by the party as a transcendent authority that bears a recognizable face in the person of the local party secretary. The orchestration of bodies, practices, spaces, is done at the level of conscious signification, where it can also be consciously resisted. However, despite this difference in the operation of the norm, is there anything less "modern" about this power, in its representation of itself as the triumph of a rational will over the powers of darkness and in its dreams of effecting a profound transformation of space and time and the practices that inhabit them? I suggest that the village compact is precisely a self-conscious attempt to construct an "alternative modernity" in the face of the universalizing claims of Western industrial culture that imposes its own norms of political legitimacy. Moreover, Zhang Yimou's film suggests that it is possible to contest this economy of power within the terms of its own self-representation without having to appeal to Western models of popular democratic government. Qiu Ju's search for an answer is framed by notions of reciprocity and an ethic of personal relations deemed missing from a "Western" industrial culture. But the film also draws attention to the gap between representation and practice in which we see the much more complicated picture that emerges when the model of normative rule collides with the juridical use of force. The compacts assume the form of willed consent but operate by means of the arbitrary exercise of power. In practice, the compacts may provide the means whereby local cadres can reconstitute their power in the face of their declining direct control over production. The delegation of power to local officials carries with it a high tolerance for the abuse of power by local officials in the forcible seizure of body and property-the "crude methods" (tu fazi) of "local emperors" (tu huangdi). Foucault suggests that the dawning of the modern disciplinary state came with the consciousness of a "badly regulated distribution of power" in which officials in the lower jurisdictions could ignore procedure and

Ann Anagnost

357

exercise their power arbitrarily without sufficient supervision. 34 This dysfunction was the result of a monarchical "super-power"-an overconcentration of power at the center. What reformers of the late eighteenth century wanted was a new economy of power that would render "it more constant and more detailed in its effects, to increase its efficiency, to punish less and punish better. " 35 In other words, they wanted to disseminate power more effectively throughout the system. This reforming zeal is analogous to the party's tradition of rectification, which targets the commandist and "patriarchal" tendencies of local cadres. But this personalization of power in the person of the party secretary is not just the unevolved product of a feudal past, it is also very much a legacy of the tactics of guerrilla war and revolution that were repeatedly reprised in the Maoist period as a highly effective mobilization strategy for rapid economic development, a guerrilla tactic for national survival under cold war conditions. The personal power of local officials is defined as abuse and yet the center depends on it to distance itself from the deployment of force. In other words, it is a feature of the distribution of power in contemporary Chinese politics that has been actively reproduced throughout the socialist period. In the current debates about whether local party officials still retain power in the wake of decollectivization, the questions we ask should perhaps employ a more mobile conception of power that can take into account the changing loci of its exercise. This suggests a more radical reading of the death of that hapless peasant as a protest against his excessive punishment-suicide defines perhaps the only limits of a power that hides its violent aspect behind the guise of willed consent. And yet, even this act of ultimate protest is swept up to provide grist for the mill of power's self-representation. The Problem of History As mentioned above, the village compacts are represented as a split, as both old and new. This "splitness," in fact, enables a denial of history. The propaganda materials promoting village compacts are marked by a willful amnesia, a refusal to recognize their historical antecedents. 36 In propaganda work, the compacts are portrayed as novel institutions not imposed from above but "invented" (chuangzaole) by the people themselves as a spontaneous response to their transformed circumstances in the reform period. However, there is an interesting bifurcation in historical consciousness in which recognition of this past is in fact acknowledged

CONSTRUCTING THE CIVILIZED COMMUNITY

but only within certain discursive domains. In academic discussions of the compacts, the historical connection with pre-socialist forms is noted and recognized as a problem. The problem is, stated simply, how a state that calls itself "socialist" and "democratic" can justify the use of an institution long identified within Marxist historiography as a tool of the ruling classes (tongzhi jieji). This curious dilemma begs interrogation of the operation of power in the postMao state, for implicit within this split consciousness is a suturing over of the representational function itself. By attributing the village compacts to the spontaneous ingenuity of "the people," their collective genius becomes a metonymy for an indigenous folk culture that stands apart from the state and that is, therefore, authorizing and legitimating as a specifically "Chinese" approach to the social order, the primordial determination of a socialism (or democratic institution) with "Chinese characteristics." At the same time, their identity as "masses" (qunzhong) or "the people" (renmin), in the sense of the "body politic," is indeed a product of the nation-state, which in its modern guise, rests its authority precisely upon the production of these subjectivities. 37 The confusion of authorship imparts a legitimating function to the pedagogical activities of the state as emerging from a culturally specific model already present among the people. And yet, the academic discourse on the compacts recognizes their long history as an institution of imperial statecraft, dating back at least to the Qing, a history that openly acknowledges the project of pedagogical subjection to norms imposed from on high. Once this history is acknowledged, it becomes a problem. How do we situate the discourse on the village compacts and the attendant discussions about "Chinese" models of community within the cultural politics of the reform era? In other words, what is this history made to mean in the recent contestation of the Chinese past as either burden or resource for the imagined future of the nation? In some respects, this dilemma is perhaps shared by all nationalisms. However, it clearly must be situated within the context of reform-era China. Homi Bhabha has suggested that rather than thinking in terms of histories, we should instead think in terms of "temporalities" that resist "the transparent linear equivalence of event and idea that historicism proposes" by providing "a perspective on the disjunctive forms of representation that signify a people, a nation, or a national culture. " 38 This focus on history in its refiguration or "rememoration" of a national past forces us to think in terms of what is at stake in connecting to this past in ways specific to the present "time" of the nation.

Ann Anagnost

359

This problematization of history has emerged from recent work on nationalism, beginning with Benedict Anderson's Imagined Communities and further developed in a critical post-colonial historiography in which the Enlightenment notion of "history" itself is seen as inextricably intertwined with the emergence of the nation as the reigning political form of the twentieth century. 39 This insistence on the "invented" nature of the nation opens up an awareness of the "gaps" sutured together to give national identities a primordial and internally homogeneous quality. "If nation-states are widely conceded to be 'new' and 'historical,' the nations to which they give political expression always loom out of an immemorial past, and, still more important, glide into a limitless future. " 40 Indeed, it is precisely the adoption of an Enlightenment history that produces the nation as the natural product of the progressive unfolding of inherent essences or Spirit, a narrative form by which "non-nations were converted into nations. " 41 Duara, in discussing the implications of the adoption of Enlightenment history in the early development of Chinese nationalism, notes the uneasy relation that nationalist discourse has to its past and its present. "The nation as the subject of History is never able to completely bridge the aporia between past and present. " 42 The village compact betrays this split quality both as an artifact of a modernizing project, as "democratic," and as the sign of an essentially Chinese conceptualization of the social order. These sorts of split object are never unambiguous; they are both a reassurance of the primordial basis of national identity and a nagging symptom of the nation's failure to progress. This tension is readily apparent in the semantic play between "culture" (wenhua) and "civilization" (wenming) in the heated context of the cultural debates (wenhua re) of the past decade. Wenhua here refers to the notion of reclaiming a Chinese essence in the project of moving toward wenming, a state of civility closely identified with the advanced industrial cultures of Asia and the West. 43 Indeed, although wenming is often identified with Westernization, much popular interest is directed toward Japan and Singapore as two nations that have successfully achieved civility apparently without losing their cultural identity; indeed their economic transcendence becomes retroactively the stimulus to (re)essentialize a cultural basis for that very success. A refusal to link the appearance of village compacts to some essentialized model of Chinese culture is in a sense to take a position within the culture debates of the r98os. The controversial television documentary series Heshang (River elegy) criticized the Chinese predisposition to look toward the past as a "repository ... of society's ideal form" and

CONSTRUCTING THE CIVILIZED COMMUNITY

argued for a wholesale rejection of many aspects of Chinese "tradition," its overcentralization of power, insularity, and submissiveness to authority.44 This critique is counterbalanced by a Confucian revival that, in the wake of the abrupt economic takeoff of the 1990s, may yet add strength to a post hoc argument for a "Confucian capitalism" assumed to have a magical affinity with the capitalist logic of the late twentieth century. "The official project of the Confucian revival is thus potentially dangerous in that it operates in the mode of the 'metaphorical identification' of the subject with the state, of the private with the public spheres, of morality with loyalty, and of past with present. " 45 This identification between subject and state raises issues concerning representation in its dual sense of both "proxy" and "portrait," as well as how, in the rapidly overheating economy of the 1990s, popular sovereignty should be constituted. 46 The Problem of Representation The uneasy relationship that these compacts bear to the past is closely tied to the issues of popular sovereignty and representation. This was clearly evident during my fieldwork research in 1991 when I gave a presentation of my preliminary research results on the village compacts to an audience of Chinese intellectuals. The discussion period included two questions from Chinese participants that were addressed with some heat. The first asked why village compacts had "lost" their efficacy in dealing with problems of social disorder. The second raised the problem of the potential dissonance between the ethico-moral values of the state and those of the local community. The specific example used to illustrate the second question was the sometimes violent village sanctions imposed upon women accused of extra-marital relations. The question about "lost efficacy" seemed to echo a widespread concern about the disintegration of community norms in villages that had, until recently, been encysted within the vertical hierarchies of the Maoist centralized state. These villages, under the economic reforms, have increasingly opened out to the world, making more porous the administrative membrane that surrounds them to admit a steady flow from the interior of a growing mobile rural population in search of some foothold in the massive economic transformations highly localized along the eastern coast of ChinaY To ask this question suggests a growing lack of confidence in the party and its methods. To the extent that the village compacts had once been considered as "efficacious" (you ling), their present promotion

Ann Anagnost

by the party when they are no longer considered effective seemed to suggest a sort of tired inertia in the party's own morale that keeps it reaching back for the means by which it had once attained its ruling hegemony. Perhaps this is why the village compacts were suddenly "discovered" to have a history dating from the Yan'an period, despite their representation, up until quite recently, as a spontaneous popular innovation of the early r98os. In a short piece that appeared in the People's Daily, the author claims to have received "from a comrade in Yan'an" a copy of a village compact published in the Liberation Daily in r 94 3. This comrade goes on to suggest: "This sort of village compact certainly has a relevant significance for our present project of establishing 'spiritual civilization. "' 48 This move suggests a return in the political culture of the 1990s toward a nostalgia for the Yan'an period and the early years of the PRC, a phenomenon Lisa Rofel has referred to as "liberation nostalgia. " 49 This nostalgia was actively cultivated in the officially generated celebration of the film ]iao Yulu, which resurrected the image of a model party secretary from the early r96os. To ensure large audiences, this film was required viewing for political study in spring 1991, and work units distributed large numbers of free tickets. The visible consumption by cancer of the party secretary's body as he labors for the people's well-being recalls the "thematization of excess" of traditional moral tales of filial piety. 5° This nostalgia for the political commitment of the Liberation and even the Cultural Revolution was combined, nonetheless, with an ironic recognition that history has incapacitated one for this sort of blind political certainty. The response to the film] iao Yulu, on the part of young intellectuals I knew, was: "Zheiyang de ren tai shaole!" (That sort of person is all too rare!). One senses here the desire to believe in the model of Jiao Yulu and an awareness of the impossibility of such a model. It is important to note that liberation nostalgia is not limited to party circles alone, but is more generally distributed among the population at large; and yet this media blitz of nostalgic films and television dramas cannot but play ironically in the present period. Zhang Yimou's film rewrites the subtext for this theme of consuming labor by juxtaposing the model of the self-sacrificing local official with its apparent costs: the unquestioned submission of the masses to authority and the overdependence on an economy of power that personalizes its "face." These representations of power in its most idealized form legitimate abuse and ensure that "new" democratic modes of self-government actively reinforce the authority of local officials they were ostensibly created to displace.

CONSTRUCTING THE CIVILIZED COMMUNITY

The second question, which raised the issue of village justice, is no less disturbing in its pessimistic view of "the people," a mass noun that also shares the splitness of being "between" past and future. Duara notes that at the heart of this split in the time of the nation is the concept of the peopleY Intrinsic to the idea of the nation as an identity that is deeply historical, the "people" likewise retain an identity that is primordial in character. And yet to constitute a popular sovereignty that is intrinsic to the "modern state," the "people" have to be remade as part of a massive pedagogical project. This project includes not just the formal educational apparatus, but also, as Duara suggests, the creation of a modern literature and emergence of folklore as a motivated recuperation of the national essence, as well as, of course, historiography itselfY The notion of popular sovereignty is legitimated not by popular politics in the streets but by "a higher impersonal mechanism: the written constitution. " 53 The "people" are not just remade; they are written into being. This emphasis on writing "the people" into being suggests to me Derrida's notion of writing as the supplement to speech, in which writing is somewhat ambiguously regarded as both adding to and filling in "what is missing" to cover over the limits of any totalizing claim. "The supplement has not only the power of procuring an absent presence through its image, procuring it for us through the proxy of the sign, it holds it at a distance and masters it. For this presence is at the same time desired and feared. " 54 This ambivalence that mixes together desire and fear appears to me to express itself through this bifurcated historical consciousness that only partially obscures the presocialist antecedents of the village compacts. To suggest that the compacts are the innovation of the people is to suggest that they derive from an indigenous folk tradition that has no explicit relationship to the "ruling classes" of the prerevolutionary past. And yet the concern remains that this indigenous culture also harbors the persistent traces of an unenlightened "culture" that is feudal and patriarchal, that bears evidence to China's failure to advance according to the laws of Enlightenment history. This construction of the rural masses as unenlightened has been highly elaborated in the past half-dozen years by means of the issue of population quality (renmin suzhi), which circulates not just within party rhetoric but also in oppositional intellectual discourse and in the language of everyday life. It represents a real shift in the discourse of population from one merely of quantity to that of quality, and it constructs "the people" in terms of a lack so profound that it makes necessary a vast array of pedagogical initiatives and orthopractic measures on the part of the party apparatus.

Ann Anagnost

And yet this nostalgia for the compact as a manifestation of the popular will also somewhat complicate the rendering of the "feudal" past as uncompromisingly backward. Does it not also betray a nostalgia for community that is not just "civilized" as defined by the enlightenment neologism of wenming but also "moral" in the Confucian sense of daode? After all, the two notions are often linked together in the hybrid phrase wenming daode as the quality that defines China's present lack, not only what it needs to develop as a "civil" society, but also what it needs to recuperate from its past. So the compacts are, to me, the projection of an impossible desire that both desires and fears the presence of the masses in the political process. They are the authorizing presence of the party's sovereignty, but they also harbor all the darkness of a subjectivity marked as the other of modernity that therefore cannot be invested with political sovereignty until it has been subjected to an extensive and prolonged process of re" modeling. Conclusion Lydia Liu's discussion of the discourse of individualism in May Fourth literary and philosophical debates shows the multiple articulations of what that category has been made to mean in recent Chinese history. Her purpose is not to decide which social theory is most desirable for China or even which one is most "Chinese," but merely to examine carefully how the terms of the debate are set and the uses to which they are put. As she warns, such amnesia or forgetting of the discursive history of such powerful constructions threatens to enact a "return of the nightmare. " 55 Her careful retrieval of this history actively deconstructs any attempt to identify the terms of the debate as belonging to the unproblematized abstractions of "China" and "the West." What then, does such scholarship enable? If scholarship is to constitute a truly political praxis, as Michael Taussig has suggested, must it not go beyond establishing the truism of the "constructed" nature of ideological categories? What if, instead ofmerely deconstruction, we were to direct our practice to nothing less than "more" construction? "What a sociology that would be! " 56 I am not sure that it is my place to imagine some proleptic vision of a popular sovereignty "good for China." Nor do I wish to claim the possibilities opening out of the present transformations of Chinese society for any universalizing theoretical project. Rather, I prefer to follow Liu's model of noting how the terms of debate are set and the uses to which they are put, while at the same time, in sympathy with Taussig's argument for a politically engaged

CONSTRUCTING THE CIVILIZED COMMUNITY

scholarship, I wish to note the possibilities opened up by new articulations of constructed categories that cannot easily be contained by a universalized abstraction such as "civil society." If we view The Story of Qiu Ju as an allegory about the emergence of "individual agency" in reform-era China, then surely we must recognize the presence of the marketplace as an enabling, but perhaps not sufficient, "cause." QiuJu paves her way up the hierarchy of the legal system through the sale of hot peppers, which she and her family have produced as a household specializing in the production of a single commodity for sale (zhuanye hu). Part of the comedy is the repeated motif of her setting off for yet a higher level of the legal system, accompanied by her young sisterin-law and a load of peppers. Her new-found geographical mobility is mirrored in her economic ability to buy "city clothes," which look odd especially when worn over her padded peasant clothing, an effect accentuated by the ungainliness of her pregnant body. Despite the comic effect of this masquerade, do we not also have a hint of the rapid acquisition of an ability to embody new subjectivities through the acquisition of commodities?57 However, the role of the market is not limited to the frame of the film, but indeed becomes a factor enabling its production. Just as Qiu Ju had to go over the head of the village chief and seek recognition of her grievance at a higher court of appeal, so Zhang Yimou had to depend on foreign investors and his growing international acclaim first to make his film and then to ensure that it could be shown to Chinese audiences. 58 If the market is an enabling factor not just in Zhang Yimou's overcoming of political obstacles to make his film but in his ability to "imagine" Qiu Ju's coming into a realization of her own agency, then what does this have to say about the current debates about China's political present and future? Do we just assume the liberal narrative of the power of the market to lead inevitably toward an increasing rationalization of political power or do we concentrate instead on the multiple and uncertain discursive possibilities opened up by a new coming together of forces? Implicit in this question is the opposition between universalism and particularism. In analyzing the relations of power in the Chinese socialist state, the question should not be whether this modality of power is somehow more "Chinese" but rather how the concentration of power in a myriad tiny centers somehow produces the effect of a single center. Who can dispute the unmistakable aura of power in the personage of a local party secretary? What happens when we multiply this "local emperor" (tu huangdi) or "red sun" (hong taiyang) several thousandfold? Is the

Ann Anagnost

concentration of power in this personage not a legacy of the campaign mentality of a centralized party faced with the conditions of guerrilla war, succeeded by the guerrilla conditions of rapid economic development during the Maoist era and beyond? The relative autonomy of a thousand tiny centers to deliver results up the hierarchy of command is perhaps more a legacy of how the modern Chinese state, under conditions of imperialism and cold war politics, took a form that was in some respects overdetermined for it rather than deriving from some long-standing and immutable Chinese conception of the social order. If the market economy has unleashed new forces that threaten the party's monopoly of political power, what might that new economy of power look like? There is no clear answer in Zhang Yimou's film and yet the film goes beyond being merely an invitation to reflect passively on the arbitrary nature of power. It actively seeks to call into being a new subjectivity that acts on those reflections. At the end of the film, Qiu Ju is faced with the realization that she has succeeded in her quest only too well. Having made peace with the village head, she must now face once more the arbitrary working of power. 59 One can only imagine her starting out on a new series of journeys, this time on behalf of her old enemy to seek his release from prison. It is clear that Qiu Juhas not completely rejected the communitariar norms of face, nor the obligations of reciprocity (baa); rather, she merely seeks to ensure that power is made accountable to them as well as ordinary citizens. There is a gulf between the kind of family justice that she claims-the village is like a big family-and the justice dispensed by the Court .... [But] if you don't ask a question, nobody will ever give you an answer. You always have to fight in order for something to be done. In China, you have to try twenty times, and spend years in order to solve the most minor problems. Officials don't make any mistakes really, but in the end, there's never any answer. To request that something be done is the beginning of democracy. With this film, I wanted to say that every Chinese-and not only the peasants-should do the same thing: to fight for their rights and discover themselves in the process. 60

Whether the "village compacts" and other organizational forms remain signs of an imaginary discourse or whether they provide the means to perform into being a new kind of political subjectivity, beyond the pedagogical intentions of the center, remains to be seen. Zhang Yimou's film as a new articulatory practice is perhaps a start, but it is no guarantee, nor should we underestimate the ability of such new articulatory practices to be reabsorbed by the very power they seek to question.

REFERENCE MATTER

Notes

For complete author names, titles, and publication data for works cited here by short forms, see the Works Cited, pp. 441-71. Introduction The authors thank Kenneth Pomeranz for his comments on the Introduction. r. A. H. Smith, Chinese Characteristics, p. 237. Robert Hart was more emphatic: "They are well-behaved, law-abiding, intelligent, economical, and industrious, they can learn anything and do anything,-they are punctiliously polite, they worship talent, and they believe in right so firmly that they scorn to think it requires to be supported or enforced by might" (Robert Hart, These from the Land of Sinim [London, 1901], p. 54; quoted in L. K. Little, "Introduction," to ]. K. Fairbank, K. F. Bruner and E. M. Matheson, eds., The I. G. in Peking: Letters of Robert Hart, Chinese Maritime Customs, r868-I907 [Cambridge, Mass.: Harvard University Press, Belknap Press, 1975]). 2. Michie, The Englishman in China, pp. 369-70. 3· Hsiao Kung-chuan, Rural China, p. 3· 4· Ibid. pp. 8-ro. 5. Wolf, "Gods, Ghosts, and Ancestors," p. r 3 5. 6. Two fine examples are Dennerline, "Fiscal Reform and Local Control"; and Polachek, "Gentry Hegemony." 7· Originally published as two separate articles in]imbun kenkyu 22.4 (1971) and the Iwanami koza sekai rekishi, vol. 12 (1971), these became a single essay, "Kyoshin shihai no seiritsu to kozo," in Shigeta's Shindai shakai keizaishi kenkyu, later translated as "The Origins and Structure of Gentry Rule," in Grove and Daniels, State and Society in China. 8. For American scholarship, see Esherick and Rankin, Chinese Local Elites and Patterns of Dominance, which contains several excellent articles on elite power in both late imperial and post-imperial China. 9· Kuhn, Sou/stealers.

Notes to Pages 9-33 Fischer and Lundgreen, "Recruitment and Training." Stone, Family, Sex and Marriage, p. 99· I2. See Terada, "Min shin nori chitsujo." I 3. Even when people enjoy a commonly agreed upon set of meanings, they can contest the actions of other parties. In an agrarian society like late imperial China, this meant tax resistance, grain seizures, and rent protests. 14. They may also persist amid far greater disparities of power, as James Scott's work suggests; see Weapons of the Weak and Domination and the Arts of Resistance. IO.

I I.

Chapter r I thank Anthony De Blasi, Hilde De Weerdt, Paul Smith, Thomas Lee, and members of the Columbia University Seminar in Nco-Confucian Studies for their comments on an earlier version of this chapter. Epigraph: Xu Ji, fie xiao ji, 29. I I b. I. One might note that the assumption that there was an orthodoxy includes both those who are not sympathetic to it-see, e.g., the introduction to Kwang-jing Liu, Orthodoxy in Late Imperial China-and those who apparently are, such as Wm. Theodore de Bary, Neo-Confucian Orthodoxy and the Learning of the Mind-and-Heart. I agree with Willard Peterson's ("Review") point made in regard to the first work-that we can distinguish different interests, each of them trying to impose its own orthodoxy on particular others, without assuming that there was "an orthodoxy"-and would extend it to the second work. Some hold that the court cared little about correct belief ("orthodoxy") and much about correct practice ("orthopraxy"); see Watson, "The Structure of Chinese Funerary Rites." 2. The Yuan exams also represented patronage for those who thought of themselves as shi versus men who did not believe that "reading books" mattered and were perhaps of other social backgrounds. From this perspective, the exams ought to have been acceptable to the shi whatever their content. However, in fact the restoration of the examination system went together with a decision to adopt a Nco-Confucian examination curriculum at the expense of the literary curriculum popular in the Southern Song and Jin. 3. X in Tang shu 4 5. I I So. This was almost half the total of all groups eligible for eventual entry into the civil service. The eligibility list of 73 7 is translated in full in Huang Ch'ing-lien, "Recruitment and Assessment of Civil Service Officials," pp. 248-49· 4· Du You, although concerned more with practical talent than with ethical conduct, approved of the objections to the literary jinshi examinations and quoted several attacks in full in his Tongdian. Xie Qian, for example, in contrasting "ethical conduct" (de xing) with "talent for language" in 692 cited LiE's attack on southern literary writing under the Sui and asserted that Yang Di's creation of the jinshi examination led students to value superficiality ("the floating and empty"; fu xu) (Tongdian I7.409). After 755 Zhao Kuang, a follower of the Spring and Autumn Annals teachings of Dan Zhu, calling for a general overhaul

Notes to Pages 33-3 8

371

of the system, objected that the jinshi degree was valued by the age yet passing was determined by clever skill in composing poems and rhapsodies (Tongdian 17-419). Also after 755 Shen Jiji thought of ways to recruit "substantial talents," men with "talent and virtue" rather than "literary phrasing" (Tongdian r8.446). Du You himself admitted, however, that from the Kaiyuan (713-41) period on, "All shi, worthy and unworthy, were ashamed not to succeed on the basis of literary composition [wenzhang]" (Tongdian r 5·3 57). 5· ]iu Tang shu I I9.3430-32; Quan Tang wen 33 r.14b-I6b. For dating to 763, see Tongdian I5.358. 6. Jia Zhi's memorial can be found in ]iu Tang shu I I9. 34 3 2-4. For others who defended the importance of wen, even while calling for redefinitions of "good" wen, see Bol, "This Culture of Ours," chap. 4• 7. For the creation of new examination fields after 7 55, see McMullen, State and Scholars, pp. 95-97, I 51-53, I98-99. For an example of the association of particular values with texts, see Liu Zongyuan's discussion of how he tried to embody the textual tradition in his own writings in his letter to Wei Zhongli (Liu Zongyuan, ]i 34·87I-74). 8. It was not until mid-century that the number of jinshi degrees began to exceed the degrees for memorization fields; see Chaffee, Thorny Gates, pp. I9293· 9· Song huiyao, "Xuanju" 3.5b. IO. Chen Shan, Menshi xinhua 6a. II. Han Yu, Han Changli ji 5.20.23, farewell preface for Chen Tong; cf. Hartman, Han Yu, pp. 22I-22. The key verb here is wei (to do, engage in, realize). I2. Zu Wuze, Longxue ji 8.42. I3. This, I think, is the central message of Han Yu's famous essay "On the Origin of the Way," Han's intellectual history of the sages' dao. I4. Han's refusal of Zhang Ji's request that Han Yu spell out his doctrine is a well-known example. I5. Xu zizhi tongjian changbian 53.II68. Zhang argues that the current practice of drawing themes from a wide range of texts encouraged candidates to develop the breadth of reading necessary to recognize the themes. This practice may have been a response to the complaint in 983 against those with skill in composition but without extensive book learning; see Song huiyao, "Xuanju" 3·4b. I6. For proponents of curricular narrowness, see Song huiyao, "Xuanju" 3.I6b-I7a from I029/If2 and 3.I8b-I9a from I038/I/8. On testing prose first, see Song huiyao 3.IIb from IOI7/9f28; 3·I7b-I8a from I034/3/r; 3·I9a from I038Irl29. Note that examiners during these first decades of the eleventh century resisted a change that increased their workload and denied the value of modern literary forms; see Song huiyao, "Xuanju" 3·3 I a. I7. The rules for the new system were drafted by a committee headed by Song Qi. The rules are preserved in Song huiyao, "Xuanju" 3.23 b-29b; for the imperial edict inaugurating the reform, see Song huiyao, "Xuanju" 3 .29a-3oa, dated I 044/3/r 3. The three goals of the reform are stated in both documents. r8. Zhou Dunyi explained what he means by "wen yi zai dao" in his Tong

372

Notes to Pages 38-40

shu. There are other examples. Bi Zhongyou (Xitai ji 6.9b): "Wen is a device for conveying the dao, and the historian has responsibility for conveying the wen." Lii Tao (]ingde ji 20.23b-24a): "The Classics are that by which the dao is conveyed, and the dao is the path to order." See also Wang Ling, ]i I 9. 3 3 I. Note also the use of the phrase "transmit the way" (chuan dao) ("and illuminate the mind") by Wang Yucheng, Xiaochu ji I8.I53, and Mao Pang, Dongtang ji 6.3b-5a. 19. Chen Xiang, Guling ji I6.7b-8b. 20. For more on this position, see Fuller, The Road to East Slope; and Bol, "This Culture of Ours," chaps. 6 and 8. Zhu Changwen (Lepu yougao 6.4aIoa ) notes the guan dao position. 21. For an example of the first position, see Lii Tao, Jingde ji I 3. 3 b-4a. For the second, see Huang Shang (I044-II3o), ¥anshan ji I6.Ia-2a, I9.9b-roa, 20.1oab, 21.2a-3a, 21.3a-4b, 23.1a-5b. 22. Tang Geng (I07I-II2I), Meishan wenji 8.Ia-3a. Tang went to argue that the guwen style should be required in the exams since it is free prose yet, because it contains embedded rhyme and meter, requires discipline. 23. Wu Chuhou, Qingxiang zaji 8.Ia. 24. Chen Guan notes this position (I am not sure he really defends it) in his "Debate over Wen," in Sheng Song wenxuan quanji 3 2. I I a-I 2b. See also Zhao Dingchen (1070-II29), Zhuyin ji shi ji 11.5a-9a. 25. Bi Zhongyou, Xitai ji 5.4b-6a. See also Mao Pang, Dongtang ji 6.3b5a; and Li Fu ( I052-I 128), Jueshui ji 5.12a-I4a. 26. Examples include Liu Ban (I023-89), Pengcheng ji 34.2b-4a; Kong Wuzhong (ca. Io4r-ca. I097), Qingjiang san Kong ji I5.1Ib-12b; Zheng Xia (ro46-III9), Xi tang ji 2.I4b-I8a; and Lii Nangong (I047-86), Guanyuan ji II.2a-7a, 15.Ia-3b. 27. Song huiyao, "Xuanju" 3.43b-44b (I07Ihlr). Cf. Jin Zhongshu, "Bei Song keju zhidu," pp. I07-8. However, it seems that the candidates remained partial to poetry, for half of them chose the Odes and only one out of ten the Documents. In I078 the court promulgated a quota in response, requiring that henceforth 30 percent of passes be reserved for the Odes and the Change each, 20 percent for the Documents, and IO percent each for the two Rites. See Song huiyao, "Xuanju" 3·45 (Io78/7h5). 28. Song huiyao, "Xuanju" 3.43b-44a. This document also appears in Wang's collection; see Linchuan ji 42.450. 29. Local schools were ordered in nor to replicate the three grades of the Imperial University, the Three Hall system; those graduating from the local "upper hall" gained admission to the Imperial University directly. Then, from I I 04 to II2I, while the regular examinations continued, tests were held to graduate a small number of upper-hall students at the University directly into the civil bureaucracy. See Thomas H. C. Lee, Government Education, pp. 126-29. 30. Ibid., pp. n6-I9, 126-32. Although the number of buildings increased by 5,ooo, the total number of students in no8 had declined from over 2Io,ooo in II04 (ibid., p. I32). I do not know how to account for this. 3 I. See Winston Lo, "Philology" and "Wang An-shi". For an account of the

Notes to Pages 41-45

373

new commentaries, the Zi shuo, and the new curriculum, see Cheng Yuanmin, "San jing xinyi". 32· Wang Anshi, Linchuan ji 39·4IO-II. 33· Ibid., 64.678. 34· Shen Liao, Yunchao bian 8.73b-74a. 3 5. Elsewhere I have treated Wang's view of himself in greater detail and depth; see Bol,, "This Culture of Ours," chap. 7; and "Government." 36. Wang Anshi, Linchuan ji 77.812. 37· Zeng Gong,Ji 12.197-8. 38. Wang Anshi, Linchuan ji 67.714. 39· Ibid., 66.707. 40. Wang Anshi, Linchuan ji 73·779· 41. Ibid., 84.88o. 42. Most of the objections are found in Song huiyao, "Xuanju" 3.48b-s8a. For a detailed reliable account of the history of Wang's curriculum, attacks on it, and defenses of it from the ro8os into the mid-twelfth century, see Cheng Yuanmin, "San jing xinyi." 43· SuShi, Su Dongpo ji 4.12.8.43· See also Chen Guan in Sheng Song wenxuan quanji 323. I Ia-12b; XuJi,Jiexiao ji 29. Iob-IIa; Li Zhaoqi (ca. roseca. II 26), Le jing ji; Tang Geng, Meishan wenji 1.5a-7a; Mao Pang, Dongtang ji 6.p-8a; Liu Anjie (ro68-II r6), Liu zuoshi ji 1.2a-2b. 44· Fan Zuyu, Fan Taishi ji 35.15a-r6b. This echoes a memorial on the examination system from Chief Councillor Liu Zhi (I030-1097), Zhongsu ji 4.2oa-23b. This accusation was not a slander, as the themes from palace examinations under Huizong make clear; see Song huiyao, "Xuanju" 7·33a-35b. For a defense of Wang's Zi shuo as illuminating the principles of both written characters and "human nature, destiny, and morality," see Cheng Yuanmin, "San jing xinyi," pp. 259-60. Zou Hao (Dao xiang ji 22.roa-b) at one point explicitly compared the Mencian zide with Huayan enlightenment. 45· Mencius 4B14. 46. Zheng Xia, Xi tang ji 6.ra-Iob. 47· See the palace examination questions for the years III5, rrr8, II2I, and I I 24 in Song huiyao, "Xuanju" 7·3 3 b-37a. The writings of Wang's critics were officially proscribed in I 102. 48. Hua Zhen (rosr-rrro+), Yunqi jushi ji 23.27a-3o; cf. 24.13b-r6a, 24.16a-r8b. 49· Chen Guan, Siming zun Yao lu, 32.3b-8a. 50. SuShi, Su Dongpo ji 2.6.30.11. Elaborated on in his commentaries on the Documents and Change. 51. See, e.g., Su Che (I039-III2), Luancheng ji 20.447-48; Chen Shidao (1053-IIOI ), Housh an ji 14.9a; Chao Buzhi ( ro 53-1 IIO ), file ji 37 .ra-b; and Zou Hao, Dao xiang ji 9.4b-5b. See also SuShi, SuShi wenji 7.204; Tang Geng, Meishan wenji 1.5a-7a; and Mao Pang, Dongtang ji 6.5a-8a. 52. Wang's commentaries were attacked but not forbidden in the early Southern Song. In I r 3 8 the court explicitly rejected a proposal to rely exclusively on

374

Notes to Pages 45-47

Wang's texts and decided instead to allow scholars to use traditional commentaries and their own opinion in essays on the Classics. Even Qin Gui, associated to some extent with Wang's learning and attacked by the followers of Cheng Yi's teaching, agreed both that Wang's failing was wanting others to agree with his opinion and that literati need not stick to one view. See Cheng Yuanmin, "San jing xinyi," pp. 273-80. Zhu Xi wrote an examination reform proposal (but never submitted it) with an eclectic Classical studies curriculum. 53· See Su's letter to Xie Minshi, partially translated and discussed inK. Smith et a!., Song Dynasty Uses of the I Ching, pp. 9 5-96. 54· Yuan shi 8r.2o18. The edict was drafted by Cheng Jufu; see Yuan wen lei 9.112-13. 55. Yuan shi 81.2019. For more detailed accounts of the split between literary intellectuals and Daoxue advocates in intellectual culture under Mongol rule, see Abe, Gendai shi, pp. 44-53. DeBary (Neo-Confucian Orthodoxy, pp. 2027, 3 8-44, 5 3-6o) notes the division between literary and Daoxue scholars over the examinations. See also Lam, "On the Yuan Examination System." For the spread of Daoxue under the Mongols, see Wing-tsit Chan, "Zhu Xi and Yuan Nco-Confucianism." For a comprehensive account of Mongol and Yuan examinations, see Yao Dali, "Yuanchao keju zhidu." Literary learning in Jin intellectual culture is discussed in Bol, "Seeking Common Ground." Some Yuan literati had a similar view of recent intellectual history; see, e.g., Yu Ji, Daoyuan xue gu lu 22.549, and Xu Youren, Zhizheng ji 34.1a-2b. 56. Su Tianjue (1294-1352), Cixi wen'gao 30.II67, speaks of changing customs. See also Jie Xisi (1274-1324), Quanji 5·3 I 8. As in the Song, it seems likely that Yuan men saw the examinations as a means to demonstrate that they were indeed shi or literati. For the Song case, see Bol, "The Examination System and the Shi." 57· Hsiao Ch'i-ch'ing, "Yuandai de ruhu." Writing in 1305, Xu Qian (Xu Boyun xiansheng wenji 4.1a) noted a debate over whether to "select shi from the schools" or "nurture shi in the schools." Xu favored the first; Lu Wengui (Qingdong leigao 7.7b) the second. John D. Langlois, Jr. ("Introduction"), and others have noted Chinese support for the Yuan; Qian Mu ("Du Mingchu") unhappily notes this about literati active during the Yuan-Ming transition. 58. Zhang Zhihan (1243-96), Xiyan ji 13.6a-7a. Wu Shidao (jinshi of 1}21; Wu libu wenji 10.27b) explained why shi should be the leading social group in "On the Origin of the Shi." Li Cun (1281-1354; Fanyang Zhonggong Li xiansheng ji 28.1oa-14b) argued that the examinations have been restored to end the corruption of government by the clerks and the attendant popular suffering. For the desire of literati to distinguish themselves from clerks, see, e.g., Wu Shidao, Wu libu wenji 10.27b; and Lu Wengui, Qiangdong leigao 7.7b. In a begging letter Huang Jin (1297-1357; Huang Wenxian gong quanji 3.48a) contrasted himself with the layabout sons of nobility and the rich people while assuring his patron that his family never descended to the ranks of farmer, artisan, or merchant. 59· Yu Ji, Daoyuan xue gu lu 31.549. The difficulty of getting the "right men" through examinations was itself a subject discussed in some examinations; see questions on the subject by Pu Daoyuan (I26o-1336), Shunzhai xiansheng

Notes to Pages 47-52

375

xianju conggao 13.IIa-12a; Lu Wengui, Qiangdong leigao 3.3b (cf 7.6b); and Wu Shidao, Wu libu wenji 19.2ob-21b. Some noted the traditional moral objection that examinations were inimical to self-cultivation because they encouraged literati to be "new and different"; see Yang Weizhen (1296-1379), Dong Weizi ji 8. I a-b. 6o. Li Cun (I28I-I354), Boyang Zhonggong Li xiansheng ji 28.IOa-14b. For Wu Cheng, see deBary, Neo-Confucian Orthodoxy, p. 6o. 61. ChengDuanli, Weizhaiji 5.1b. 62. Wu Shidao, Wu libu wenji 19.26b-28a. 63. Lun Yu yinde 1.6. 64. Zheng Yu, Shishan xiansheng wenji, preface. His collection now goes by the title Shishan ji. 65. Cheng Duanxue (1278-1334), ]izhai ji 3.17a; Xu Youren, Zhizheng ji 34.1a-2b; Wu Shidao, Wu libu ji 19.31b-32b; Wei Su (1303-1372), Wei taipu wenji 6.ua-112a. Gong Shitai (Gong libu wanzhai ji 7.3b-5b) suggested that both the Classics of the Former Kings as difficult reading and the "writings of the unique individual" (yijia zhi yan) of Han, Tang and Song were impediments; Tang Yuan (1269-1349 ); Yunxuan ji 10. I 8a, stressed the problem of traditional Classicism. 66. See, e.g., Gong Shitai, Gong libu wanzhai ji 7.9a; Hu Han (1307-13 81), Hu Zhongzi ji 2.ua; Liang Yin, Shimen ji 10.4ob-41b. 67. Liang Yin, Shimen ji 10.4ob. 68. Cheng Duanli's curriculum, Chengshi jiashu dushu fennian richeng, is an example. 69. See for example, Li Cun's account of his own conversion, in Boyang Zhonggong Li xiansheng ji 28.1a-4b; cf. ibid., 28.Ioa-14b. 70. For a discussion of these two sides of Neo-Confucianism and Wu's efforts to reconcile see Gedalecia, "Wu Ch'eng." 71. Cheng Duanxue,]izhai ji 3.17a-b. Cf. Chen Lii, Anya tang ji 8.19b; Wu Lai, Yuanying Wu xiansheng ji 9.9a-1 I a; Pu Daoyuan, Shunzhai xiansheng xianju conggao 20.4b-8a; Hu Han, Hu Zhongzi ji 5.2a-2b. 72. Xu Youren, Zhizheng ji 34.1a-2b. 73· Gong Shitai, Gong libu wanzhai ji 7.3b-5b. 74· Cheng Duanli, Wei zhai ji 3.7b-9a. 75· YuanJue, Qingrongjushiji 18.6b. 76. Ibid., 18.1a. 77· For examples of literary men who were aware of the place of Daoxue in intellectual culture but largely ignored it, see the literary collections of Chen Lii, Anya tang ji, and Zhou Tingzhen (1292-1379), Shichu ji. 78. Of particular value are Langlois, "Political Thought" and "Jin-hua Confucians"; and Sun Kekuan, Yuandai Jinhua xue shu. Those active during the Yuan-Ming transition are treated in the context of other arguments in Dardess, Confucianism; and Qian Mu, "Du Mingchu." See also Rong Zhaozu, Mingdai sixiang shi. 79· Sun Kekuan (Yuandai ]inhua xue shu) treats Xu Qian, Liu Guan, and Huang Jin at length.

Notes to Pages 52-56 So. Sec Langlois, "Yu Ji"; Hok-lam Chan, "Wei Su"; and Ouyang Xuan's biography in Yuan shi I82. 8 I. Huang ]in, Huang Wenxian gong quanji 6. I sa, 6. 3 2a-3 3 b. 82. Ibid., 3.9b, Ioa, Izb. For an equation of Nco-Confucianism with the "learning of the nature and decree," see ibid., 6. 3 2a-3 3b. 83. Langlois ("Jin-hua Confucians," pp. I 84-8 5) notes this and limited evidence for Sun Kekuan's case about Huang's Nco-Confucianism (see Sun Kekuan, Yuandai ]inhua xue shu, pp. 145-6). 84. Huang Jin, Huang Wenxian gong quanji 7.1p, 29a, 32b, 36a, 4Ib, 43a, 46a, 53 a, 61a, 72b. In his prefaces for collections of writings that he himself said are devoted to spreading Nco-Confucian teachings, Huang devoted at most a sentence or two to the man's thought. Instead, he wrote at length about the author as a social being, situating him in networks of acquaintances, his native place, his times, and even literary traditions; see ibid., 6. 30b-3 2a, 6.3 2a33b. 85. Ibid., 5.I2a-Ip. 86. Liu Guan, Liu daizhi wenji 13.1p-16a. 87. Ibid. 88. Huang Jin, Huang Wenxian gong quanji 6.I Ia-I2b. 89. Xu Qian, Xu Boyun xiansheng wenji 3·14b-16a, letter to Wu Shidao. 90. Yu Ji, Daoyuan xue gu lu 7, 31.523-25, 33·549· 91. Ibid., 15.267-68 and 31·523-35· 92. Ibid., 7.136-37. 93· Ibid., 7.137; cf. 33·549· 94· Ibid., 7.134-36. Yu was also concerned that too much emphasis on the internal basis of morality could lead literati to forsake seeking the knowledge of historical change necessary for political accomplishment. 9 5. Liu Guan, Liu daizhi wenji q.21a-22b. 96. Ouyang Xuan, Guizhai ji 7.9b-1 1a. 97· Wu Lai, Yuanying Wu xiansheng ji 6.7b-9a. 98. Ibid., 1r.1a-2a. 99· Langlois, "Jin-hua Confucians," pp. 154-67. 100. Following the text in Guo Shaoyu and Wang Wensheng, Zhongguo lidai wenlun xuan, pp. 229-3 2. 101. Ibid., p. 232. 102. Mote ("Sung Lien," p. I226) takes a similar view: "This tradition, in so far as it is identifiable, was rather narrowly orthodox, considered to be directly in the line from Zhu Xi and Lii Zuqian, through an important intermediary figure, Xu Qian. Yet it was characterized by an unusual breadth of intellectual concern, as well as literary refinement." 103. Wang Wei, Wang Zhongwen gong ji 2.466I-69. This account of the Zhu Xi lineage reappears in Wang draft for the Rulin biographies for the Yuan dynastic history; see Yuan shi r r.sr 31-32. 104. Huang Jin, Huang Wenxian gong quanji, Song Lian's preface. 105. Song Lian, Song xueshi wenji 8.I49-150. Io6. Wang Wei, Wang Zhongwen gong ji 15.5333-42.

Notes to Pages 56-59

377

I07. Ibid., 2.4697-98. Io8. Ibid., 1.465o-54· I09. For a contrasting view based on the writings of Liu Ji, Wang Wei, Song Lian, and Hu Han, see Dardess, Confucianism, pp. I 3 I-83. This book warrants a much lengthier response than I can provide here. Chapter

2

Research and writing for this paper were funded by the Fulbright Foundation (Foundation for Scholarly Exchange in Taiwan), the Japan Foundation (Tokyo and Kyoto universities), the Republic of China National Science Council (Tsing Hua University), and the Pacific Cultural Foundation during the I990-9I and I99I-92 academic years. Portions have previously appeared in T'oung Pao (Leiden) and Chtlgoku bunka to shakai (Tokyo) and appear here with the permission of the editors involved. 1. For the locus classicus of the ancient account of the Duke of Zhou's sagely behavior as regent in assisting the infant son, King Cheng (r. I II 5 ?-Io78 B.c.), of his brother King Wu (r. I I 22-I I I 5? B.C.), see Shangshu zhengyi, compiled by Kong Yingda (547-648) eta!., in Shangshu leizhu chuji (Taibei: Xinwenfeng Press, I984), I: I98-2o2. See also Sima Qian (145-90? B.c.), Shiji (Beijing: Zhonghua Bookstore, I972), 33.I5I8. 2. See the account of this confrontation in Mingshi jishi benmo, p. 209. See also the I609 "Xu" (Preface) by Guo Zizhang to Lian's collected works, entitled Lian Zhongcheng gong wenji (late Ming Wanli ed.), 3a;and the I762 "Xu" by Zhou Huang to Lian Zhongcheng ]inchuan ji ( I762 ed.), Ia-4a. Lian's collected writings were first put together in J49I by Wang Zuo (1440-I5I2) under the title Jinchuan yuxie ji. Later Wang was able to add additional writings by Lian called "Zining yigao" that had been hidden by one of Lian's townsmen. The collection was republished in I 543· Li Mengyang (14 73-1529), when he served as an education official in Jiangxi in I 5 I I and restored Zhu Xi's famous White Deer Academy there, also prepared a shrine for Lian near Lian's birthplace and helped print his writings. See Li's remarks on Lian in "Lian gong yishi," appended to Lian Zhongcheng gong wenji I7a-I7b. Cf. Frederick W. Mote's account of Lian in Dictionary of Ming Biography, hereafter DMB, pp. 9 I I-I 2. Mote thinks the confrontation possibly apocryphal, but most late Ming and early Qing biographies of Lian accept this version of the events. See, e.g., Li Zhi (I527-I6o2), Xu cangshu, pp. 283-85; and the Ming mingchen yanxinglu I0.4a-7b. For a late Ming collection of different versions of this confrontation, which generally confirm the version presented, see]ianwen chaoye huibian IO.I5a-31a. 3· Huang Ming tongji shuyi 3.66a-66b. See also Li Zhi, Xu cangshu, pp. 287-88; Mingshi jishi benmo, pp. 206-7; and Crawford eta!., "Fang Xiaoru in Light of Early Ming Society," pp. 30 5-7. For unofficiallate-Ming versions, again see ]ianwen chaoye huibian 7.Ia-28b. 4· Given the overlap between the accounts of Lian Zining's and Fang Xiaoru's defiance, it may be that the accounts have been conflated. In the Yupi lidai tongjian jilan (Imperially inscribed collection for the mirror of history), compiled during

Notes to Pages 59-63 the Qianlong reign (I736-95), for instance, the King Cheng story is associated only with Fang Xiaoru; see IOI.2I b-22a. The circumstances surrounding their deaths remain indisputable, however. 5. For an account of these rumors, Zhu Di's unwillingness to quash them, and how the Qing dynasty Ming History project handled them, see Yupi lidai tongjian jilan IOI.2Ia-b; and Kahn, Monarchy in the Emperor's Eyes, pp. I446. The Qianlong emperor wrote that the Yongle emperor had permitted the story that the Jianwen emperor had escaped to flourish to cover up the latter's murder. 6. See Mingshi jishi benmo, pp. 206-I9. 7. See the ]ianwen chaoye huibian 20. 24a. 8. See Chan Hok-lam, "The Rise of Ming T'ai-tsu." See also David Chan, The Usurpation of the Prince ofYan, passim; and Franke, "The Veritable Records of the Ming Dynasty." 9· Sec Zhu Di, Shengxue xinfa, "Xu" Ia-28a. Cf. Huang Ming gongjukao r.82a. For a recent summary of what the author calls Zhu Di's sagely "image building," see Cheuk-yin Lee, "Zhiguo zhi dao," pp. 2II-25; Lee's account glosses over the murders of I4o2. ro. See Hok-lam Chan, "The Chien-wen, Yung-lo, Hung-hsi, and Hsuan-te Reigns, I 399-I43 5," pp. 214-2 I. On the troubled rise of Dao Learning, see the recent accounts by Hoyt Tillman, Confucian Discourse and Chu Hsi's Ascendancy, pp. I-r 8; and Peter Bol, "This Culture of Ours": Intellectual Transitions in Tang and Sung China, pp. I-3 r, which have superseded earlier ahistorical accounts of "Nee-Confucianism" that contained unfortunate "modernist" readings of Sung Confucianism permeated with contemporary teleological notions of Chinese culture and philosophy. Cf. my "Zhongguo wenhuashi de xin fangxiang." rr. See the useful discussion of "orthodoxy" in China as "socioethics" in Kwang-ching Liu, "Socioethics as Orthodoxy," which supersedes the philosophically overdetermined notion of "Nco-Confucian orthodoxy" elaborated in Wm. Theodore de Bary's Neo-Confucian Orthodoxy and the Learning of the Mindand-Heart, pp. I-66. 12. Howard]. Wechsler, "T'ai-tsung (reign 626-49) the Consolidator," pp. I82-87, 214-15. For the Hongwu emperor's interest in Li Shimin, seejia Naiqian, "Cong Mengzi jiewen zhi Qianshu," p. 45· 13. See Romeyn Taylor, "Official and Popular Religion." 14. On the shi, see Bol, "This Culture of Ours," pp. 3 2-7 5, which notes how the evolution of local elites from first aristocrats and then civil officials in Tang and Song times was accompanied by the rise in favor of Dao Learning among these elites. 15. See my "Social, Political, and Cultural Reproduction," pp. 9-23. Cf. Pierre Bourdieu and Monique de Saint-Martin, "Scholastic Values." I 6. Hermann Ooms explores somewhat similar problems in the role of Confucianism in eighteenth-century Japan in his Tokugawa Ideology. I 7. For differing views, see F. W. Mote, "The Growth of Chinese Despotism"; and w. T. deBary, Neo-Confucian Orthodoxy, pp. I s8-68. I 8. See my "Changes in Confucian Civil Service Examinations."

Notes to Pages 64-68

379

19. Dianshi dengkelu, I4oo: 1.14, in Huang Ming dengkelu huibian; and DMB, I5I9-2r. 20. Dianshi dengkelu, I400: I. I-68. 21. Dianshi dengkelu, I4oo: 1.3. On Xie's career, see his biography in Mingshi I47·4II5-23· See also DMB, 555-56. The process of fabrication is detailed in Chan Hok-lam, "The Rise of Ming T'ai-tsu," pp. 688-91. 22. Huang Ming sanyuankao 1.34a-b. 23. Dianshi dengkelu, 1400: I. I I, I 5. 24. There is a story that Lian told Jin that Jin was destined to become a "famous official," whereas Lian himself would be known as a "loyal official." See ]ianwen chaoye huibian IO.I5a-I5b; cf. DMB, p. 9I I. 25. Zhuangyuan tukao I.Ip. 26. See Lin Qingzhang, "Wujing daquan zhi xiuzuan ji qi xiangguan wenti tanjiu," pp. 366-67. 27. Mingshi I47·4II5-I6. 28. Huang Ming gongjukao 1.76a-8Ib; Huang Ming tongji jiyao 6.Iob. See also Qiu Hansheng, "Mingchu Zhuxue de tongzhi diwei," pp. I42-43. 29. Huang Ming gongjukao r.8Ib. See also Shang Zhuan, Yongle huangdi, pp. I40-47; and Zhang Chenshi, "Yongle dadian shihua," pp. I87-92. 30. See Kahn, Monarchy in the Emperor's Eyes, pp. 44-46. 31· Huang Ming gongjukao I.8Ib; Huang Ming tongji jiyao 13.9a. 32. Shang Chuan, Yongle huangdi, p. I47· 33· See Chan Hok-lam, "The Rise of Ming T'ai-tsu," pp. 689-90; DMB, I563; and Zhang Chenshi, "Yongle dadian shihua," p. I88. 34· See Danj6 Hiroshi, "Mindai kakyo kaikaku no seijiteki haikei," pp. 4995 14· Provincial examinations were held in I403 rather than I402 because of the civil war and because they had been held in 1402 under the Jianwen emperor. Similarly the metropolitan examination was delayed to I404 instead of I403. See Huang Ming sanyuankao 2.Ia-2b; and Huang Ming gongjukao 2.47-48a. Beijing was officially designated the Northern Capital for the I403 provincial examination. In addition to provincial examinations, recommended candidates were also selected for appointment in I403. See Shang Chuan, Yongle huangdi, p. I53· 35· See Huang Ming tongji jiyao I3.6a. 36. See ibid., I3.4b. 37· Huang Ming jinshi dengkekao 3.2a. Cf. Ikoma Sho, "Minsho kakyo gokakusha no shushin," p. 48. 38. See Huang Ming tongji jiyao I3.7b. 39· Huang Ming tongji shuyi 4.I2a gives an account of Zhu Di meeting with his seven top advisers, most of whom were from Jiangxi (including Xie ]in, Hu Guang, Yang Rong, and ]in Youzi), and praising them for their support since he took power in 1402. 40. See Huang Ming zhuangyuan quance 2.I8a-44a; and Huang Ming jinshi dengkekao 3. Ia-20b. 41. See Huang Ming zhuangyuan quance 2.I8a-I9b, 36a-b. 42. See ibid., 2.36b-43b, esp. 36b-37a.

Notes to Pages 69-73 43· See my "History in Policy Questions from Southern Provincial Civil Examinations," Table I. 44· See Huang Ming zhuangyuan quance 2. Ia-I7a. Cf. my "Philosophy (I-Ii) Versus Philology (K'ao-cheng)." 4 5. For further examples, see the palace examination of 14 r 2 in Dianshi dengkeiu 1412: 1.275-285, which links the "mind-heart of the emperors and kings" with their political governance, and the palace examination of I52I in ibid., 1521: 6.2993££. Such questions were repeated in the Qing. See my "Changes in Confucian Civil Service Examinations," pp. I23-33· 46. Huang Ming gongjukao r.82a. 4 7. See Zhu Di, Shengxue xinfa 2b-3 a, 3 6a-b. For discussion, see my "Philosophy (I-Ii) Vs. Philology (K'ao-cheng)," pp. I75-222. 48. Zhu Di, Shengxue xinfa, "Xu" 24b-2sa. 49· Gao Yao and Gui served Emperor Shun; HouJi and Xie served Emperor Yao. so. See Huang Ming tongji jiyao 14.11b, 14.18b. For Xie's fall from grace, see Huang Ming tongji shuyi 4.17a-b. 5 r. Chan Hok-lam, "The Chien-wen, Yung-lo, Hung-hsi, and Hsuan-te Reigns," p. 221. 52· See Gu Yanwu, Rizhi Iu 8 58 I 8or. 53· Huang Ming gongjukao r.82a-82b. See also the I78os accounts of the Sishu daquan and Xingii daquan by the editors of the Siku quanshu project in Ji Yun eta!., Siku quanshu zongmu 36.13b-14b, 93.7b-8b; and Mingshi, 282.7222. Cf. Qiu Hansheng, "Mingchu Zhuxue de tongzhi diwei," pp. 147-53; and Lin Qingzhang, "Wujing daquan zhi xiuzuan," pp. 377-81. 54· Zhu Di, "Yuzhi Xingii daquan xu," Ia-3b. 55. See Hu Guang et a!., "Jinshu biao," 3 b. See also Da Ming Taizong Wen huangdi shiiu I 58. 2a-4a. 56. See Huang Ming tongji jiyao 9.5b-6a. 57· See Tu Shan, Mingzheng tongzong 5.IIa; and Mengzi yinde s6!7BII4. 58. Lunyu yinde 23/12/7. See also Mengzi yinde 26/4N2. 59· Huang Ming gongjukao 1.85b. See also Tu Shan, Mingzheng tongzong 5.rra; and Ho Yun-yi, The Ministry of Rites, p. 95· Cf. DMB, pp. 363, 389, 957· 6o. Mingshi I39.398. See also Huang Ming tongji jiyao 9.5b-6a. During the Qing dynasty, Zhu Yizun (I629-I709), among others, thought the story of Zhu Yuanzhang's censorship of Mencius untrue because there were insufficient records of it; see Zhu Yizun, Pushuting ji 69.8b-9b. 6r. See the "Tici" by Liu Sanwuto the Mengzi jiewen 1a. 62. Huang Ming tongji jiyao rr.8a, 12.I7b; Mingshi jishi benmo I6.I93· 63. Huang Ming tongji shuyi 4.8b-9a describes how Zhu Di in I404 had already decided that Mencius should be part of diwang zhi xue (studies of the ruler). 64. Qiu Hansheng, "Mingchu Zhuxue de tongzhi diwei," pp. I44-47· 6 5. For an example of Ming dualism, see notes chosen by Hu Guang for ren

Notes to Pages 73-78 (humanity) in the Lunyu jizhu daquan, r2.4b-5a, in Sishu daquan, comp. Hu Guang eta!. (SKQS ed.). See also my "Criticism as Philosophy." 66. Mengzi jizhu daquan 2.28a-29b (quotation of Hu Bingwen), 8.5b-8b, ro.34a-3 5b. Cf. Yun-yi Ho, "Ideological Implications of Ming Sacrifices in Early Ming," pp. 5 5-67; and my "Imperial Politics and Confucian Societies," pp. 390402. 67. See Hucker, "Confucianism and the Chinese Censorial System." 68. For an interesting exception to this pattern, see the translation of an essay by Ai Nanying (I583-r 646) on the quotation min wei gui ("the people are the most valuable") from the Mencius, chosen by late Ming examiners to express their dissatisfaction with eunuch power, that Ai wrote for his I 624 Jiangxi juren degree, in Andrew Lo, "Four Examination Essays of the Ming Dynasty," pp. 176-78. Because of the nature of his remarks in answering a pointed policy question for the same provincial examination, said to be critical of the powerful eunuch Wei Zhongxian (r568-r627), Ai was barred from taking the metropolitan examinations for three periods. See his biography in Ai Nanying, Tianyongzi ji, p. 49; Mingshi 288.7402; and Eminent Chinese of the Ch'ing Period, p. 4· 69. I base this general conclusion, which includes several exceptions, on a study of over IOO Ming examinations and some 400 Qing examinations in the No. I Historical Archives (Beijing) and the Ming-Qing Archives (Nangang, Taibei) that I have seen to date. 70. See Zhang Yishan, Zhu-Ming wangchao shilun wenji, pp. 280-93. 71. See my "Social, Political, and Cultural Reproduction," pp. 19-23. 72. Jia Naiqian, "Cong Mengzi jiewen zhi Qianshu," pp. 43-44. 73. Ming Shizong shilu (Taibei: Academia Sinica reprint, I965 ), r 19.3 b-4a. See also Huang Ming gongjukao r.88b-89a. Cf. DMB, p. 1095. Zheng Xuan, who would become a "patron saint" of the Han Learning revival in the eighteenth century, had his tablet reinstated in the Confucian Temple in 1724. See Da Qing Shizong Xian (Yongzheng) huangdi shilu (Taibei: Huawen reprint, 1964), 2o.I8b2oa. 74· Mote, "The Growth of Chinese Despotism," pp. I3-I4. See, however, the qualification of Mote's analysis for the Song in Anthony W. Sariti, "Monarchy, Bureaucracy, and Absolutism." The cited slogan also carried a double edge. "To expel barbarians" became a rhetorical element in Han Chinese efforts to eliminate first Mongols and then Manchus from imperial control during the mid-fourteenth and late nineteenth centuries, respectively. See my Classicism, Politics, and Kinship, pp. 147-56. 7 5. See my "Social, Political, and Cultural Reproduction." 76. Hu Juren, ]uyelu 714.36-44. See also DMB, pp. I497-98; and Wingtsit Chan, "The Ch'eng-Chu School of Early Ming," pp. 45-46. My thanks to Adam Schorr for his help on this point. 77· Mingshi 147·412.5. 78. Jiao Hong, Guochao xianzheng lu 20.56a-56b. 79· For the pardon granted Lian's descendants, see the emperor's order, ra3a, in the "Liangong yishi" appended to Lian Zhongcheng gong wenji.

Notes to Pages 78-84 So. See ibid., 4a-5b, for the I573 imperial pardon. 8 I. See ibid., Ia-p, p. 24a (separate pagination from above); and Shang Chuan, Yongle huangdi, pp. 127-31. 82. See Yao's "Ba" (Afterword), Ia-5a, in ]ianwen chaoye huibian. 83. Zhu Lu, Jianwen shufa ni, "Qianbian" 1a-20b, and "Fubian xia" 6b7a (Wanli ed.), in Beijing tushuguan guji zhenben congkan, vol. I I. 84. On the late Ming process of legitimating the Jianwen reign, see ]ianwen chaoye huibian 20.1a-2oa. See also Zhu Lu, ]ianwen shufa ni, "Fubian xia" Ia-Ib, in Beijing tushuguan guji zhenben congkan, vol. II. 85. See Li Zhi, Xu cang shu, pp. 291-92. 86. See the biographies of Fang Xiaoru and Lian Zining, which include both men's confrontations with Zhu Di, in Jiao Hong, Guochao xianzheng lu 20.5 3 a56b, 54.I8a-I9a. 87. See Ai Nanying, Tianyongzi ji I0.14a-I9a. 88. Ibid., IO.I6a-I Sa. Ai perished in I646 after leading Ming loyalist forces in unsuccessful efforts to roll back the Qing armies advancing throughout south China. 89. For Qian Qianyi's elegy, see Zhu Lu, ]ianwen shu fa ni, "Fubian shang" 28a, in Beijing tushuguan guji zhenben congkan, vol. I I. Cf. Eminent Chinese of the Ch'ing Period, pp. I48-5o. 90. Huang Ming gongjukao, 2.4 p. 91. SeeShiRunzhang, "Xu" (Preface) Ib-2a,inLianZhongcheng]inchuanji. 92. Manchus and Mongols were initially appointed based on other criteria and occupied important positions in the central bureaucracy. See Kessler, "Chinese Scholars and the Early Manchu State." 93. Eminent Chinese of the Ch'ing Period, pp. 26o-6r. 94- See the Kangxi "Xu" (Preface), in SKQS 7IO: I-2. 9 5. Kahn, Monarchy in the Emperor's Eyes, p. 46. Chapter 3 Portions of this discussion have also appeared, in different form, in my article "The Chinese Poetic Canon and Its Boundaries," in John Hay, ed., Boundaries in China (London: Reaktion Books, I994). I. For a critical exploration of this and other commonplaces, see Owen, The Great Age of Chinese Poetry. 2. The editors of the Complete Poems of the Song (Beijing daxue, Guwenxian yanjiu suo, Quan Song shi), of which the first volume appeared in I99I, note that the number of poets eventually to be included in the series "is no fewer than nine thousand"; they do not mention the total number of poems. See the "Editorial Explanation," I: IO. The formidable magnitude of late imperial poetic production led Yoshikawa Kojiro to state, as recently as three decades ago, that "as for a Complete Poems of the Ming Dynasty or Complete Poems of the Ch'ing Dynasty, there is little possibility of their appearing, even at some time in the future" (Five Hundred Years of Chinese Poetry, rrso-r6so, p. 7). It should be noted, however, that the

Notes to Pages 84-86 first volumes of a projected Complete Poems of the Ming (Quan Ming shi), edited by a national committee of scholars, nonetheless began appearing from the Shanghai Classics Publishing House (Shanghai guji chubanshe) in I990. 3. For another discussion of issues pertaining to canon formation, with specific reference to anthologizing, see my "Poems in Their Place." 4· Kermode, "Canon and Period," in idem, History and Value, p. II5· 5· Weimann, "Shakespeare (De)Canonized," p. 71. 6. Tompkins, "'But Is It Any Good?': The Institutionalization of Literary Value," in idem, Sensational Designs, pp. I88-89. 7· Raymond Williams, Problems in Materialism and Culture (London, I98o), p. I6; as cited by Chandler in "The Pope Controversy," p. 481. 8. Frye, Anatomy of Criticism, pp. 2I-22. 9· Although it is not known precisely when the writing of regulated verse first came to be tested, there are poems included in the Song dynasty anthology Wenyuan yinghua, compiled in 987, that are clearly examination pieces from as early as the Kaiyuan and Tianbao reign periods, i.e., 7I3-55-the High Tang. According to Shang Yanliu (Qing dai keju kaoshi shulu, p. 249), 458 poems are included there. See also Wang Yunxi, "Shi He yue yingling ji xu lun sheng Tang shi ge," I: 29, nr. IO. In McMullen, State and Scholars in T'ang China, pp. 225ff. I I. Before 1070, for example, the Song dynasty palace examination (the last of three steps on the route to the jinshi degree) "had included only a poetry and rhyme-prose question; after I07o, however, the palace examination was changed to a single policy question administered by the emperor to rank the candidates for official appointments. This format remained intact until the twentieth century" (Elman, "The Changing Role of Historical Knowledge in Southern Provincial Civil Examinations," p. 282). I am grateful to Professor Elman for sharing with me his extensive bibliography on the civil service examinations and, in particular, this article. 12. Benjamin Elman ("The Changing Role of Historical Knowledge in Southern Provincial Civil Examinations") notes that tests of poetry and rhyme-prose composition were eliminated from sessions of the civil examination during the Yuan "because of their frivolity" (p. 283), and that when Ming emperor Taizu promulgated a decree in 1370 re-establishing the examination process of selection, he based "the examination curriculum on the Yuan dynasty examinations because of their more concrete-less literary-requirements" (p. 3 I 2, ni 2). During the Qing, however, debates revived over the testing of poetic composition on various levels of the civil service examination; in 1679, for example, a special examination required candidates to compose poetic works that the emperor himself would evaluate. See Deng Siyu, Zhongguo kaoshi zhidu shi, p. 275. It was not until I7 57, however, that all students sitting for the district-level examinations had to compose a pentasyllabic poem in eight couplets, a requirement that was then extended to other levels of the exam. 13. As E. A. Kracke, Jr., has pointed out, however, the memorization and explication questions that prevailed on versions of the exam other than the one leading to the jinshi degree could be graded even more mechanically than could

Notes to Pages 86-88 original compositions and were thus presumably less susceptible to the subjective evaluation of the examiner; see Civil Service in Early Sung China, 96o-Io67, p. 6414. From the section "Criticisms of Poetry" ("Shi ping") of Canglang shihua; in Guo Shaoyu, ed., Canglang shihua jiaoshi, p. 136. 15. These comments from Yang Shen's Sheng'an shihua and others are cited by Guo Shaoyu, Canglang shihua jiaoshi, pp. r 3 6-3 7. r 6. Particularly influential was Yan Yu's charting of the Tang into five "period styles": (1) Early Tang style (Tang chu ti); (2) High Tang style (Sheng Tang ti); (3) Dali (reign period, 766-835) style (Dali ti); (4) Yonghe (reign period, 8o62o) style (Yonghe ti); and (5) Late Tang style (Wan Tang ti); see Guo Shaoyu, Canglang shihua jiaoshi, p. 48. 17. From a letter to his maternal uncle Wu Ling, included in Guo Shaoyu, Canglang shihua jiaoshi, p. 23 4, ·as translated by Richard John Lynn in his review of John Timothy Wixted's Poems on Poetry, p. 710. 18. Siku quanshu, ji bu, 307: 1368fr75. 19. Ibid., 13681r76. 20. First, the Four Bravos have been grouped together without further subcategorization because, since their works collectively represent the first transformation of the previous Six Dynasties style, they are essentially all of a prosodic piece, despite variations in line length. The "orthodox sounds," by contrast, have been labeled and grouped by form (into ancient verse [gu shi], regulated verse [lu shi], quatrains [jueju], and lines of five or seven syllables), because even though there are qualitative and stylistic differences among them, the prosodic resemblances prevail; these groupings, he added, are "for the convenience of the reader." As for the "echoes handed down," they have not been subclassified because the works included are so diverse in length and prosodic form; they have simply been selected according to their merits (ibid.). 21. Yang was criticized almost immediately by Su Boheng (1329-92?) for having created divisions based on formal stylistic categories rather than historicist ones; see his Gu shi xuan Tang xu, a preface to an anthology compiled by his friend Lin Yuzhi, in Su Pingzhong wenji, 4.2oa. Su Boheng's critique is cited, and disputed, by Qian Zhongshu in the opening of his Tan yi lu, pp. 1-2. 22. Yang was not, however, the first or only editor to do so. It is well known that these canonical figures were not well represented in collections compiled during the Tang itself. They were also often not included in later anthologies with differing critical agenda, such as Zhao Shixiu's southern Song Zhongmiao ji, which favored a style associated with the ninth-century poet Yao He, and Wang Shizhen's Qing dynasty Tang xian sanmei ji (discussed below), which valorized the Wang Wei "school." In addition, the Tang bai jia shi xuan attributed to Wang Anshi (ro2r-86), but more likely compiled by Song Minqiu (1019-79), appears, like Yang's volume, intentionally to have omitted poems by major figures like Li Bo and Du Fu whose collected works enjoyed wide contemporary circulation. See Shi Zhicun, Tang shi bai hua, no. roo, pp. 72-773. I am grateful to Professor Kang-i Sun Chang for reminding me of this essay. Much later, Jin Shengtan omitted Du Fu's many examples of heptasyllabic regulated verse from his Tang caizi ji because he was compiling a separate edition

Notes to Pages 88-92 of the poet's works. Such decisions suggest the real complexity and contingency of anthological choices. 23. Gao Bing begins his elucidation of the general principles (fanli) of his collection by stating that he took his lead from Lin Hong, the acknowledged leader of the group, who argued, against prevailing taste, for the superior merits of High Tang poetry, which opinion was then confirmed in Gao's mind by his reading of Yan's work; see Gao's Tang shi pinhui, p. 1424. The dates given here reflect a refinement of Gao Bing's original determinations by the late Ming scholar Shen Qi in the preface to his Shi timing bian, because Gao's Early Tang did not include the reign of the founding emperor, and his Middle Tang period was too short; as noted by Zhu Ziqing in Shi yan zhi bian, p. 179· For an account of many of these issues in English, see Robertson, "Periodiza tion in the Arts." 25. Gao Bing, Tang shi pinhui, p. 9· The entire preface has been translated by Richard Lynn in an unpublished manuscript, "The Canon of Tang Poetry: Gao Bing's Tangshi pinhui." 26. An appendix to the first printing added 954 poems and 6r new poets and rounded out the original 90 juan to the 100 of the current edition. 27. For a discussion of this wide-ranging tradition, see Wixted, "The Nature of Evaluation in the Shih-p'in," pp. 227-28. 28. I am using, with his permission, Richard John Lynn's extremely apt translations of these terms from "The Canon of Tang Poetry." 29. From Gumbrecht's discussion of Voltaire's Temple du goat, first published in 1733, in "'Phoenix from the Ashes,"' p. 147. 3 o. Lynn, "Alternate Routes to Self-Realization in Ming Theories of Poetry," p. 320. 3 r. As noted by Lynn (ibid.); and in Ji Yun et a!., Siku quanshu zongmu tiyao 188.4189. In addition to the Tang shi pin hui, Gao Bing compiled another shorter and more selective volume entitled Orthodox Sounds of Tang Poetry (Tang shi zheng yin), which included a much higher proportion of High Tang to Late Tang verse. See Shi Zhicun, Tang shi bai hua, p. 775· 3 2. Gerald Graff traces these developments in detail in his Professing Literature. 3 3. Indeed, although almost all the Former and Seven Latter Masters were holders of the jinshi degree, many of them hailed from families of merchant or even farming background, a point Yoshikawa Kojiro emphasizes in his discussion of the guwen ci movement; see his Five Hundred Years of Chinese Poetry, pp. 137-76. 34· As Wai-kam Ho observes, since the majority of these figures-including Li Panlong-were officials in the Ministry of Justice, they constituted a virtually literal "high court for literary orthodoxy"; see his "Tung Ch'i-ch'ang's New Orthodoxy," p. II6. 3 5. This was the case with most other anthologies of the later Ming and Qing as well; almost every scholar who produced a collection of Tang poetry compiled a pre-Tang volume, and the two were often printed together. 36. The seven most amply represented Tang poets are Du Fu, 82 poems; Li Bo, 70; Wang Wei (7or-6r), 47; Gao Shi (716-65), 40; Cen Shen (715-70), 39;

Notes to Pages 92-94 Wang Changling {ca. 69o-ca. 756), 3I; Wei Yingwu {737-ca. 792), I6; and Shen Quanqi {ca. 6 50-7I 3 ), I 6. Yang Songnian conveniently tabulates these and other numbers in "Li Panlong," pp. 40-44. 37. Li Panlong's only preface in the anthology precedes the selection of Tang poetry and begins with a statement that elicited scores of outraged responses: "Tang lacks pentasyllabic ancient verse [wu yan gu shi] but has its ancient verse"; see his Xuan Tang shi xu, in Gu jin shi shan 244: IO.Ia. 38. Zhong Xing {I574-I624) and Tan Yuanchun's {I585-I637) Tang shi gui, for example, published together with a pre-Tang collection, Gu shi gui, as a collective Shi gui in I6I7, favored ancient-style poetry and what was regarded by some as a rather eccentric style. It was widely popular for several decades but then fell from grace in the Qing. And Lu Shiyong, a "senior licentiate" {gong sheng) of I63 3, also compiled two anthologies, Gu shi jing and Tang shi jing, in an attempt to promote a style of evocative, spontaneous "inspired resonance" {shen yun) associated primarily with the poetry of Wang Wei and the theories of Wang Shizhen. 39· See Shi Zhicun, Tang shi bai hua, pp. 775-78. 40. Ling Ziliu, "Tang shi xuanben Li Bo shi caixuan tongji"; idem, "Tang shi xuanben Du Fu shi caixuan tongji"; idem, "Tang shi xuanben Wang Wei shi caixuan tongji." Unfortunately, since Ling does not use the same group of anthologies for all three studies, there is no real statistical control here, but his figures provide a good general indication of the overall frequency of selection. 41. Wang Wei's poems are included in five out of nine fully extant Tang anthologies; Li Bo's works appear in two of them and also in the fragments of a tenth unearthed at Dunhuang; and Du Fu's are collected only in one anthology, compiled in 900. 42. The single exceptions are, for Wang Wei, "Song of Peach Blossom Spring" {"Tao yuan xing"); for Li Bo, "Changgan Song" {"Changgan xing"); and for Du Fu, "Meeting Li Guinian in Jiangnan" ("Jiangnan feng Li Guinian"). Although each of these poems has now become a staple of the respective poet's collection, the omissions are not utterly surprising in view of the fact that Li Panlong's selection, as noted above, is heavily weighted toward regulated verse {which the Wang Wei and Li Bo poems are not), and that more than one critic {even Shen Deqian) disparaged the relative quality of Du Fu's heptasyllabic regulated quatrains. 4 3. Shen is referring, on the one hand, to the Airs of the state of Zheng in the Classic of Poetry, which were deemed to be licentious in nature, and, on the other, to another group of poems in the Classic that were felt to embody values of the aristocracy and of the state. 44· From Shen Deqian, Tang shi bie cai ji, p. I. 4 5. Ibid., pp. I, 3. More than one-tenth of the almost 2,ooo poems in the volume are by Du Fu. 46. Wang Shizhen's preface to the anthology makes both the principles of selection and their critical heritage clear: Yan Canglang in discussing poetry said: "The people of the High Tang were only concerned with inspired interest. They were antelopes who hung by their

Notes to Pages 94-97 horns leaving no traces by which they could be found. Their marvelousness lies in being as clear as crystal and free from obstruction. Like a sound in the void, color in appearances, like the moon's reflection in water or an image in a mirror-their words come to an end, but their ideas are inexhaustible." Sikong Biaosheng [Tu] in discussing poetry said: "It is the flavor beyond the sour and the salty." At the end of the wuchen year of Kangxi [r688], I picked up the poetry written in the Kaiyuan and Tianbao eras [i.e., the High Tang] every day and read them, and then I had a different flash of insight into what these two masters had said. I set down those poems which especially possessed intriguing interest and transcendent meaning and came up with 42 poets, from Wang Youcheng [Wei] on down, as the Samadhi Collection of Tang Worthies. From the Tang xian sanmei ji jianzhu, 1883 woodblock ed., annotated by Huang Peifang, p. 1. Trans. (with minor revisions by me) by Richard]. Lynn, "Orthodoxy and Enlightenment," p. 240. The citation from Yan Yu can be found in Guo Shaoyu, Canglang shihua jiaoshi, p. 24. Sikong Tu's comment comes from "A Letter to Student Li Discussing Poetry" ("Yu Li sheng lun shishu"), in Zu Baoquan, Sikong Tu Shi pin zhushi ji yiwen, p. 68. 4 7. It is likely, in fact, that Sun Zhu, the editor of this volume, modeled his selection heavily-although not completely-on that of Shen Deqian. See Wang Zhong, "Lun Tang shi san bai shou." 48. See Wu Hongyi, Qingdai shixue chutan, pp. I 29-34. The Feng brothers were also among the many scholars who prepared newly annotated editions of Fang Hui's Yuan dynasty anthology, the Ying kui Iii sui, discussed below. 49· Cited in part by Wu Hongyi, Qingdai shixue chutan, p. 123. 50. For a translation of sections of the preface, see James J. Y. Liu, Chinese Theories of Literature, pp. 82-83. 5 r. See Jin Shengtan, Shengtan xuanpi Tang caizi shi, p. 3. 52. Gu Youxiao, Tang shi ying hua r: I 2b. 53· Ibid., n.p. 54· For discussion of these influential Ming and Qing celebrations of individual expression, see Lynn, "Orthodoxy and Enlightenment"; idem, "Alternate Routes to Self-Realization in Ming Theories of Poetry"; and Jonathan Chaves, "The Panoply of Images." 55. See Elman, Classicism, Politics, and Kinship, esp. pp. 290ff. 56. Lii's preface begins by clearly articulating the immediate impetus for this enterprise: "Ever since the Jia[jing] and Long[qing reign periods], scholars discussing poetry have esteemed the Tang and disparaged the Song. Song poets' collections have covered saucejars and papered walls and been discarded repeatedly, so that today it is extremely difficult to seek them out and purchase them. Those who disparage Song poetry call it 'rotten,' but this is from not yet having seen Song poetry." At the same time, however, he takes care to disavow any canonizing intentions: "I am not esteeming the Song over the Tang, but wish that all those who have disparaged the Song will be able to see in what way Song poetry is Song"; see Lii LiuJiang, Song shi chao, I: 3-4. 57· Wang Shizhen, Tang xian sanmei ji, p. 58. The persistence of this construction is evident even in challenges to it. The twentieth-century critic Qian Zhongshu, for example, disputes the traditional

Notes to Page 97 historicist linkage of poetic style to political history (Tang vs. Song) but nonetheless retains a dichotomy between two "Tang" and "Song" styles that are construed simply as aesthetic temperaments; see his Tan yi lu, passim; and Huters, Qian Zhongshu, pp. 40-48. Qian's work is in effect a monumental and ultimately quite systematic argument against the valorization of the Tang, one that he highlights by ironically using as the title for his work one that had been used by one of the Former Seven Masters, Xu Zhenqing ( I4 79-I 51 I), who of course supported the Ming archaistic program to canonize the High Tang. 59· See the introduction to the recent reprinting of the anthology with collected critical commentaries edited by Li Qingjia, Ying kui Iii sui hui ping, I: 9· In addition to being one of the few anthologies collecting poetry of both the Tang and the Song, the Ying kui lU sui is organized, as befits a work concerned with providing concrete models for composition, first by prosodic form, and then by topic, and not by chronological order. 6o. Ibid., I: I of main text. 6I. Repeated comments throughout the collection refer to Du Fu as the "progenitor" (zu) of a school of which Huang Tingjian, Chen Shidao (I052-II02), and Chen Yuyi (I090-II39) are the disciples or patriarchs (zong). Although these three Song poets are not among the ones most favored by partisans like the Yuan brothers, Fang Hui contributed to the Tang-Song debate with numerous remarks praising his Song examples, albeit sometimes in strangely backhanded ways. Of Chao Juncheng's "Deng Duojing lou," for example, he wrote that "not a single word or a single line of this poem is not skillful: who says that Song poetry is not Tang poetry?" (Li Qingjia, Ying kui lU sui hui ping, I: 23). 62. Here we have a good example of how anthologies are often overlooked in critical discourse on Chinese poetic theory. Scholars who have traced the history of interest in this pair of terms usually focus on the critical utterances of Ming and Qing writers like Xie Zhen (I495-I575), one of the Seven Latter Masters, or Wang Fuzhi (I 6 I 9-9 2) (see, e.g., James J. Y. Liu's Chinese Theories of Literature, pp. 40-43), with an occasional nod to harbingers like Wang Changling's "Discussion of Literature and Meaning" ("Lun wen yi") or Sikong Tu's letters on poetry. Fang Hui's comments throughout his anthology, by contrast, must have served as a much more systematic and widely read vehicle for promoting the ideal of a necessary counterpoise of the two elements. A slightly earlier Southern Song anthology of Tang poetry compiled by Zhou Bi (fl. I24o) had been concerned with a similar sort of balance, both syntactic and semantic. Titled Three Forms of Tang Poetry (San ti Tang shi), Zhou's anthology was compiled explicitly for the study of regulated verse and contained examples of only three forms: heptasyllabic quatrains and penta- and heptasyllabic regulated verse. He divided each prosodic form into six or seven subcategories, primarily according to how the "empty" (emotional) and "full" (scenic) couplets or halves are relatively positioned; the poems are then arranged according to his subcategories, which makes the anthology virtually unusable for most purposes. 6 3. David Rolston briefly discusses the development of such interlineal critical aids in How To Read the Chinese Novel, pp. I I-I 3. Wu Hongyi, citing a remark

Notes to Pages 97-99 by Ye Dehui in his Shu lin qing hua, notes that the anthologies by Zhong Xing and Tan Yuanchun were among the first books published to take advantage of techniques just developed in the late Ming of printing in more than one color: black for the poem itself, and red for the comments and marks of emphasis; see his Qingdai shixue chutan, p. 26. 64. Essays discussing this revival were presented at a 1990 conference on ci sponsored by the Joint Committee on Chinese Studies of the American Council of Learned Societies and the Social Science Research Council and have been collected in my Voices of the Song Lyric in China. 6 s. See the chapter by Kang-i Sun Chang in this volume. 66. Benjamin Elman ("The Changing Role of Historical Knowledge in Southern Provincial Civil Examinations," pp. 279-280) provides some telling statistics from south China that speak to this impressive rise in numbers and its effect on the competitiveness of the examinations. During the Ming dynasty in Yingtian prefecture, for example, the number of candidates soared from 2,300 in 14 74 to 7,soo in 1630; because the number of graduates increased only slightly-from I 3 s to I so-the percentage of successful candidates dropped from s. 9 percent to 2.0 percent. In Jiangnan during the Qing, Io,ooo candidates attempted the exams in I 684 and I7 ,ooo in I 893; although the number of graduates doubledfrom 73 to qs-the likelihood of success remained infinitesimal: 0.7 percent in I684 and o.8 percent in I893· In other words, the odds against passing the Yingtian examinations in the late Ming were so to I and increased to over IOO to I in Jiangnan during most of the Qing. In another article, "Changes in Confucian Civil Service Examinations" (p. I I7 ), Elman documents the purposeful lowering of quotas for successful candidates in the local and provincial examinations during the Qing and notes that by I 8 so the chances of succeeding nationwide had dropped to o.oi percent. 67. From Shen Deqian, "General Principles," in idem, Tang shi bie cai ji, p. 3. 68. From the "Original Preface by the Recluse of Hengtang" ("Hengtang tuishi yuan xu"), i.e., Sun Zhu, in idem, Tang shi san bai shou, p. 3· 69. At the end of his "General Principles," however, Wang Yaoqu did take care to praise the accomplishments of various Song dynasty poets and explained: "Poetry after the Tang has many varieties, which cannot be discussed in a single summary. Today, since our age still values the Tang, my explications have therefore stopped with the Tang. It is not that [those afterward] are especially diminished or vulgar to one's vision and hearing but rather that their beauties are too numerous to gather together" (Gu Tang shi he jie, p. 2). 70. Wang Yaoqu printed his two separately prepared volumes of ancient and Tang dynasty verse together because, as he explained at the beginning of his "General Principles," he "wanted to select from the original sources of poetry up through its mainstream and thereby attain its totality." He then provided the standard litany of evolutionary transformations that evolved from the Classic of Poetry to Tang regulated verse; see his Gu Tang shi he jie, p. r. However, this edition now contains only the selections from the Tang, although the original title and prefatory material to both have been retained.

390

Notes to Pages 99-roo

Any one set of comments on any poem could provide evidence of Wang's rather rudimentary faith in his reader's imaginative and inductive powers, but this is especially striking in the case of familiar quatrains like Li Bo's "Thoughts on a Quiet Night" ("Jing ye si"); see, e.g., 4h3a. 71. See the recent reprinting of this volume edited by Yang Yerong, Tang shi he xuan. 72. Throughout his study of Jin, Yuan, and Ming dynasty poetry, Yoshikawa Kojiro refers to a new class of "townsman versifiers" whose existence can be documented by the thousands of extant poems whose authors were not officials in the civil bureaucracy. A poetic competition held in 1286 in Zhejiang, for example, attracted over 2,700 contestants, of whom, with but few exceptions, "the great majority of entrants are otherwise unknown and can be assumed to have been ordinary townsmen" (Five Hundred Years of Chinese Poetry, p. 68). Earlier Yoshikawa writes that these townsmen "formed a newly ascendant social stratum (comprising a different percentage in each dynasty) that sought to express its vitality first of all in poetry" (p. 12). Although it is true that aristocratic pedigree no longer played the social and political role it had enjoyed before the Song and that some of these poets may well have come from merchant families, it is important to remember that the key distinguishing feature between these poets and those who held official office was the lack of a degree, not social class. Given the statistics on the examinations and the high rate of failure cited earlier (see note 66), it is small wonder that degree-holders per se came to constitute an increasingly smaller proportion of a larger poetry-writing population. From a different perspective Benjamin Elman ("The Changing Role of Historical Knowledge in Southern Provincial Civil Examinations," p. 278) reminds us not to isolate the small percentage who survived the civil service selection process from the much larger pool of candidates, for the examinations created "a broad class of classically literate and historically knowledgeable males that encompassed all who competed in the selection process, including the failures." In an "Afterword" to the English translation of Yoshikawa's work, William S. Atwell perceptively suggests that Yoshikawa's use of the term "townsman" (shimin) may perhaps best be understood as an attempt to read a distinctive element of Japanese history onto the Chinese landscape: "That urban developments in China during the late imperial period were of great importance both culturally and economically cannot be denied, but one wonders to what extent Yoshikawa's deep interest in what might be called the 'townsman culture' (chemin no bunka) ofTokugawa Japan (1603-1867), an interest that is revealed (among other places) in the footnotes to this work, colored his understanding of Chinese social, intellectual, and cultural history" (Yoshikawa Kojiro, Five Hundred Years of Chinese Poetry, p. 193). 73· Yoshikawa Kojiro, Five Hundred Years of Chinese Poetry; Wu Hongyi, Qingdai shixue chutan, pp. 17-18. 74· Yoshikawa Kojiro, Five Hundred Years of Chinese Poetry, pp. 67-69. 7 5. For a brief summary of the history and features of bagu wen, see the entry by Andrew Plaks in William H. Nienhauser, ed., The Indiana Companion to Traditional Chinese Literature (Bloomington: Indiana University Press, 1986),

Notes to Pages roo-roz

391

pp. 641-4 3. The structural similarities between the eight-legged essay and the examination poem of eight couplets have been pointed out by scholars since the Min g. The Manchu rulers of the Qing briefly suspended the bagu wen requirement on the civil service examinations from r663 to r667, but powerful objections from the Han Chinese forced them to rescind this change; see Elman, "Changes in Confucian Civil Service Examinations," p. I 19. 76. Here we may see one basis of the arguments advanced by the Ming archaiclanguage school that mastery of past models and rules was to be taught not for the purposes of simple mechanical imitation but rather as a necessary means of acquiring the tools that anyone needed to speak. See, e.g., Li Mengyang's ( 14 721529) famous analogy between employing a carpenter's square or compass to draw squares or circles and following the formal regulations of the ancients without plagiarizing their words (Kongtongzi ji 62. 7a-b; trans. James J. Y. Liu, Chinese Theories of Literature, p. 91). 77· From Plaks's entry on bagu wen, in Nienhauser, Indiana Companion, p. 64!. 78. See, e.g., Rolston, How To Read the Chinese Novel, p. 23. 79· See Shang Yanliu, Qing dai keju kaoshi shulu, pp. 2pff. Jin Shengtan, for example, insisted on the similarities between the two forms-he compiled anthologies of both-throughout his collected correspondence. A letter to Xu Xuelong opens by asserting that what is "regulated" (Iii) in "regulated verse" is method (fa), and not rhyme (yun), which of course establishes the basis for the comparison. "Before the Tang," he went on, "there was nothing with this name. It was solely because the people of the Tang wished to use poetry to select officials that they developed a new concept and created a poetic form: two [lines] beginning, two continuing, two turning, and two uniting-restricted to eight lines, it was called regulated verse." He then noted how, shortly after the Ming reinstituted the civil service examination, its government also "developed a new concept," the eight-legged essay, and the two forms play the same role in their respective contexts (Shengtan xuanpi Tang caizi shi, pp. 3 r 5-1 6). Another letter shown to Gu Songwen and Han Weiyun simply states: "Although poetry and [examination] essay are two different kinds of literary forms, they nevertheless share one rule or method. This one rule is qi cheng zhuan he. Without qi cheng zhuan he there would be no rules for the essay; without qi cheng zhuan he there would be no rules for poetry" (ibid., p. 323 ). For an example and analysis of a typical bagu essay, see Tu Ching-i, "The Chinese Examination Essay." So. This is a section of his Shi fa jia shu, included in He Wenhuan, Lidai shihua, 2: 726-37. Sr. Jin Shengtan, Shengtan xuanpi Tang caizi shi, p. 2. 82. Responding to an unnamed correspondent, ibid., p. 314. 83. Letter to Xu Zhijian, ibid., p. 317. 84. Anthologists employing the qi cheng zhuan he terminology, with or without the bipartite division, include Fang Hui, Feng Shu, ]in Shengtan, Wang Yaoqu, and Xu Ceng, who compiled the Er'an shuo Tang shi, preface dated 1662; see

392

Notes to Pages roz-ro7

Shi Zhicun, Tang shi bai hua, p. 778. In addition, Huang Peifang's nineteenthcentury edition of Wang Shizhen's Tang xian sanmei ji explicitly announces its intention of explicating the poems as simply as possible, largely by means of such analytical terms. 8 5. Shi Zhicun, Tang shi bai hua, p. 729. See Shang Yanliu, Qing dai keju kaoshi shulu, pp. 254-58, for a sample analysis of an examination essay. 86. Rolston, How To Read the Chinese Novel, p. 24. 87. See Shang Yanliu, Qing dai keju kaoshi shulu, pp. 253-54. 88. Shen Deqian, Tang shi hie cai ji, p. 2. 89. Wang Shizhen, Tang xian sanmei ji jianzhu, Ilia. 90. One such text was the Tang lU qingli ji, compiled-by Xu Rilian and Shen Shijun (the grandson of Shen Deqian, who provided a preface) immediately after the change in the examination in I7 57. In addition, Gu An's Tang lU xiaoxia lu, completed in I 7 56, was reprinted with a new set of ping dian annotations by He Wenhuan in I762, most likely in response to the new requirement; see Shi Zhicun, Tang shi bai hua, pp. 78I-82. As Benjamin Elman notes, the inclusion of a question testing poetic composition can be seen as a sign that poetry was once again "a testable measure of cultural attainment." Its revived significance was further enhanced when it was moved from session 2 of the examinations to a more prominent position in session I, in I787; see his "Changes in Confucian Civil Service Examinations," pp. I20-2I. 91. Alexander Woodside, "State, Scholars, and Orthodoxy," p. I73· 92. Ibid. The shock was hardly unprecedented, however. Benjamin Elman observes that the seventeenth-century scholar Gu Yanwu (I6I3-82) also "was chagrined to point out that classical studies during the Ming dynasty had declined to the point that students no longer bothered to read the Classics themselves. They merely picked out phrases and sentence patterns that they needed from compendiums of the latest eight-legged essays published by booksellers throughout South China" ("The Changing Role of Historical Knowledge in Southern Provincial Civil Examinations," p. 2 73 ). 93· See Wu Hongyi, Qingdai shixue chutan, pp. 14-25; Yoshikawa Kojiro, Five Hundred Years of Chinese Poetry, pp. I72-73; Elman, Classicism, Politics, and Kinship, pp. 29I-93· For a discussion of these issues with reference primarily to prose, see Huters, "From Writing to Literature." 94· Weimann, "Shakespeare (De)Canonized," p. 68.

Chapter 4 I. Wei Xi, "Chongjian Pingshantang ji," p. I48. 2. See Van Zoeren, Poetry and Personality, pp. I I I - I 5. 3. Perhaps the primary reason such contextualization is hard to avoid is our significantly greater knowledge of later imperial society. "Greater knowledge" here is not so much quantitative as qualitative: we know about phenomena and points of view not represented in earlier discourse. Thus the immediate act of contextualization that undermines the authority of Wei Xi's passage follows

Notes to Pages ro7-Io8

393

directly from the qualitative extension of the range of representation in the late imperial period, a phenomenon I consider below. There is no reason to suppose that earlier periods, the Tang, for example, actually had a less extensive range of social response than the Ming or Qing, but their representations are far more limited. And it should be stressed that by and large late imperial readers had access to no more pre-Song texts than we do in the modern period. To them, as to us, the Tang was how it represented itself. 4· Qian Qianyi, "Preface to the Poems of Ji Cangwei," in Wu Hongyi and Ye Qingbing, Qingdai wenxue piping ziliao, I: 34· Qian Qianyi is extending an old commonplace on the Classic of Poetry, first articulated by Sima Qian. After dissociating "lust" (haose) from "lewdness" (yin) and "rancor" (yuanfei) from "disruption" (luan), Qian tells us, "Only when there is genuine lust and genuine rancor will there be genuine poetry in the world." 5. Memorized poems tended to be from the Tang and earlier, although there was some memorization of more recent poetry. Writing on a friend who had put together a collection of pastiches (jiju) of Song poetry, Mao Qiling ( r623-I7I 6) observed, "Since the rise of the technique of writing pastiches and its tradition, there have been pastiches of pre-Tang poetry, of Tang poetry, of the Classic of Poetry, of the poems of Tao Qian and Du Fu, and of song lyric [yuefu shiyu changduan ju; i.e., ci], but no one has made pastiches of Song poets. The reason is that there are very few people who remember Song poems. If one doesn't remember them, one can't make a pastiche out of them; and if one doesn't remember them, those who read it [the pastiche] will not recognize the cleverness of the pastiche" (in Wu Hongyi and Ye Qingbing, Qingdai wenxue piping ziliao, I: 2II).

6. In some ways this is a continuation of the nostalgia, the sense of decline, that had been with Chinese civilization at least since the days of Confucius. There are, however, two important and related differences between earlier nostalgia and the late imperial version. First, there had earlier been phases of immense cultural confidence that gave a cyclical notion of cultural history credibility against a model of permanent decline (indeed, Qing literary history very much believed in linear notions of history). Second, there was the position first articulated by SuShi in the Song, that cultural forms were "complete" (bei) with the Tang. For a seventeenth-century version, we might quote Wei Xi's "Letter to Mr. Cai": "As I have said before, the changes in literature are finished now; and given the fact that we cannot depart from the ancients and create a form for ourselves, it is only in the superiority of the power of judgment that we may hope to join the ancients by making some addition" (in Wu Hongyi and Ye Qingbing, Qingdai wenxue piping ziliao, I: I6o). Thus rather than simply a case in which the past is better than the present, we have a profound sense of the end of history in authoritative cultural forms. Many theorists, such as Ye Xie, argued strenuously against this position, while others glorified vernacular literary forms as the natural successors to classical literature. Such arguments, however, were posed again and again over the course of centuries, which shows that their positions were never generally accepted (at least not until the vernacular literature movement of the early twentieth century).

394

Notes to Pages ro8-rro

7· Preface to the Ming wen an (first part), in Wu Hongyi and Ye Qingbing, Qingdai wenxue piping ziliao, I: 88. 8. Qian Qianyi, Xiangguan shuoshu Xu Yuantan shi hou, in Wu Hongyi and Ye Qingbing, Qingdai wenxue piping ziliao, I: 58. 9· Critics sometimes expressed hopefulness about the present, particularly when they were proposing some program for revitalizing poetry. Such optimism must have been hard to sustain considering how generation after generation of poetic reformers had been rejected not only by subsequent generations. The extent and nature of anthologizing are revealing. One recent study, Tang shi xuanben liubaizhong tiyao, gives notices on 6oo premodern anthologies of Tang poetry. There are very few anthologies of Song poetry before the seventeenth century, and then the number is far fewer than anthologies of Tang poetry. There were only a handful of non -contemporary anthologies of Yuan, Ming, and Qing poetry, many of which were compiled with the express purpose of preserving work and writers from oblivion. IO. Cf. Zhu Heling (I 6o6-8 3 ), "Preface to the Poetry Collection of Yu Wushu," in Wu Hongyi and Ye Qingbing, Qingdai wenxue piping ziliao, I: 69. I r. Huang Zongxi even proposes "diatribe" (rna) as the definitive characteristic of contemporary intellectual endeavor (Qi guai, in Wu Hongyi and Ye Qingbing, Qingdai wenxue piping ziliao, I: 93). I 2. The more complicated question, which is too large to fully address here, is the relation between these ironic, contextualizing representations and statesponsored Nco-Confucianism. It is no accident, I believe, that the two appeared on the cultural-historical stage together. It often seems that in earlier periods a person could be contradictory without being hypocritical. Tang Taizong may have killed his brothers, deposed his father, and rewritten history to his own benefit; but he could also assume the role of the ideal Confucian monarch without bad faith. Nco-Confucianism, with its demand for moral purity, authenticity (cheng), and vigilant self-examination, opened hitherto unimagined possibilities of hypocrisy. Nco-Confucianism had its "pure" genres of representation, both in elite writing and in popular genres of exemplary morality. The human beings presented in these forms are, at least on one level of intention, willfully flat. The depth that contradiction made possible in earlier literature was recuperated by a new class of ironic representations, sometimes with genuine anti-heroes and antiheroines, such as Zhao Pan'er in Guan Hanqing's Jiu fengchen or the virtuous bandits of Shuihu zhuan. But even more common was the exposure of falseness; that is, the demand for moral authenticity produced forms of representation based on the exposure of the mere appearance of morality. Such simple ironies were not easy to contain, and once unleashed, they led to many different forms of contextualizing irony. It is not that vernacular literature simply "supplanted" classical literature as a vigorous new growth in some Darwinistic process of literary history; rather, vernacular literature radically changed the field of representation in which classical literature was situated. I 3. The Xixiang ji is the most egregious example because of its destruction of the complicated high cultural figures in Yingying zhuan. I would agree that Yingying zhuan is an ironic text, but it is so in a very different and less obvious mode than Xixiang ji.lt is ironic because of unresolved contradictions, intensified

Notes to Pages

IIO-IIZ

395

to a conflict that leaves the reader nowhere to stand. To be a partisan of Zhang or of Yingying requires the active suppression of an alternative claim on our approval. Such an irony in which two positions mutually undermine each other is very different from the characteristically late imperial mode of irony in Xixiang ji, in which Hongniang constantly undermines, "offers a perspective on," the conventions of high romance. Even though we keep being drawn back into the world of high romance, Hongniang's position is not itself undermined. For a similar example in The Palace of Lasting Life, see the discussion of the lychee scenes: the attempt to restore an innocence to Xuanzong's birthday gift of the lychees to Yang guifei is undermined by the earlier scene of the suffering caused by transporting the lychees; that point of view, however, is not contextualized and undermined. 14. The role of Scholar Chen in Mudan ting is essentially parodic of true Classic of Poetry interpretation. The parodic scene as the guarantor of some genuine parallel scene was a basic device of chuanqi; for example, "Taohua shan" baitan (Scene XXXII) is a parodic mourning for the Chongzhen emperor, a failed fulfillment of the earlier vow made by Zhang Wei, and serves to articulate the proper ceremony of mourning that must be delayed until the final act. 15. No concept of the "poetic" existed in traditional China (at least as the term is used in English). However, since the Song there was a shared belief that "poetry" was a quality of experience that might or might not appear in verse and that could appear in manifestations other than verse. Although China did not have the convenient distinction between "poetry" and "verse," formulations that imply such a distinction were commonplace; for example, Qian Qianyi, "On the Volume of Poetry by Qu Youzhong": "I have always said that in discussing poetry one should not just follow one's whim in discussing whether the poem is beautiful or ugly, skillful or clumsy, but first consider whether there is a poem or not [lun qi you shi wu shi]" (in Wu Hongyi and Ye Qingbing, Qingdai wenxue piping ziliao, 1: 55). 16. Politicized yanzhi poetry was more important in critical theory than in praxis, with the large exception of the yimin (loyalist) poets from the fall of the Ming. 17. I am grossly oversimplifying here; however, the complex play of registers in dramatic language through which character is revealed was dependent on the recognition of "high" registers, the most common of which was the poetic. 18. That is, the poetic exists by virtue of the fact that a competent reader can recognize the "unpoetic" when it appears. Here, of course, we exclude the peculiar literary historical event that is modern poetry, which ever seeks out the unpoetic for the frisson of its inclusion. 19. Here we may extend textual "representation" to its older philosophical sense (clearer in the German Vorstellung), in which the limitation of the textual representation is a way to govern experiential "representations." Note that I make no claim regarding the success of such an intention, only that it can be inferred as a desire. zo. Poetry was linked to social gatherings, parties, which like parties everywhere were governed by strict rules that permitted the suspension of some of the rules operating in the world outside of party-space. Like lyric poetry, the party

Notes to Pages

IIZ-I23

requires exclusions, a temporary freedom from the concerns of the world outside the party. 2r. By saying it is unfree "in its constitution," I mean that entering such a space of putative freedom was driven by social necessity. 22. This, again, is a motif that originated in a much earlier period and took on increasingly radical formulations during the late imperial period. "What other people rush toward, the poet dreads; what other people despise, the poet loves; other people praise something, yet the poet considers it a misery; other people are enraged by something, and the poet considers it a source of delight" (Qian Qianyi, "Preface to the Poems of Feng Dingyuan," in Wu Hongyi and Ye Qingbing, Qingdai wenxue piping ziliao, I: I 9 ). Such a notion of the poet as a figure apart is in direct conflict with the "Great Preface"'s definition of poetry as a capacity common to human beings in general. Differentiating the poet by some degree of sensitivity was one way to reconcile the poet's uniqueness with his common humanity, but this is very different from the opposition of the poet and common humanity that we find in Qian and others. 2 3. Note that this is formally analogous to the need for low figures to offset and mock the high cultural figures of poetic romance in drama. 24. I do not want to suggest that the risible poet is a specifically late imperial theme. It can be traced to Du Fu, but most obviously to the Mid-Tang, where many of the concerns of later imperial culture find their first stirrings. Perhaps the most famous example is Bo Juyi's "Reciting Aloud, Alone in the Mountains" (Quim Tang shi, 22069 ). The genealogy that leads from Bo to Huang Jingren is clear, yet equally clear are profound differences. Where Bo is humorous, Huang Jingren is deadly serious: much more is at stake. Bo acts to suit his nature and without regret: his claim of addiction is undermined by his humor. Huang, on the other hand, needs compulsion. Huang's statement of "regret," although it cannot be taken at face value, still has a force. Bo does not really seem foolish, nor does he really sec himself as such: his drama of the risible poet (leaving society in order to temporarily act foolishly rather than being essentially alienated) is pure pride. Huang Jingren, on the other hand, really does see the "ludicrous" aspect of the alienated poet. 2 5. The peach blossoms on the fan are bloodstains, shed in Xiangj un's struggle to avoid being married off to another. Embellished with leaves to make the splotches of blood resemble peach blossoms, the fan has been sent to Hou Fangyu as a sign of her love for him and of the suffering she has endured on his behalf. 26. Aesthetic absorption (as in standing on a bridge on New Year's Eve staring at a star), unwavering love, and Nco-Confucian "authenticity" (cheng) are all identical in one point: their negation of doubleness and multiplicity of perspective. 2 7· "Union" (tuanyuan) is the conventional dramatic term used for the concluding scene when lovers come together. "Reunion" here should be understood in a Chinese dramatic sense rather than in the conventional Western dramatic usage of the word. 28. Equalizing gender in this case is in part due to the fact that, according to legend, the female Weaver is, in the hierarchical pantheon of deities, the social superior of the Oxherd.

Notes to Pages I29-I48

397

Chapter 5 I. Huagaishan Fuqiu Wang Guo san zhenjun shishi (Daozang 55 6-5 7, Harvard-Yenching catalog no. 777; hereafter cited as HGS), 6:rb-2a. 2. Note that all these are hierarchical relations just as surely as bureaucratic relations are hierarchical. I am not arguing here for a contrast between a hierarchical bureaucratic model on one hand and non- or anti-hierarchical models, perhaps a Ia Victor Turner, on the other. On the contrary: one of the most interesting things about the Huagai Immortals cult is precisely the variety of means it deploys to represent hierarchy between gods and human beings (or between god and god) without resorting to bureaucratic images. 3. Wolf, "Gods, Ghosts, and Ancestors." 4· Hymes, "Personal Relations and Bureaucratic Hierarchy in Chinese Religion." The larger project from which both that paper and this chapter grow is a book in progress, to be entitled Way and Byway: Two Song Dynasty Daoisms in a Local Setting. 5· HGS 5:6b. 6. HGS 5:8a-9a. 7· HGS 5:Iob-IIb. 8. HGS 5:IIb-12b. The Four call him their "sect-leader" or "patriarch" (jiaozu); again there is hierarchy but no bureaucracy here. IO. HGS 6:Ia. 9· HGS 5:Ip-I6a. I2. HGS 6:14b-I5a. II. HGS 6:r3a. I3. Even in the second case it is hard to be sure that a collective community ritual is meant: it is possible that the priests at the Zhaoqing Abbey simply performed a jiao twice a month regardless of community participation. But the regularity at least does remind one of the community events that anthropologists working in Taiwan have described-although these generally recur on a much longer cycle. I4. HGS 5:Iob-IIb. I 5. Schipper, "The Written Memorial." I6. Jiang Shuyu, Wushang huanglu, chap. 52· I7. These appear in Song huiyao jiben, li 20 and 21. Note that those listed in 20: 5oa-56b are specifically Daoist Immortals and Perfecteds; but even of these, few are represented in Jiang Shuyu's list. I8. Jiang Shuyu, Wushang huanglu sz:I8a-b. 20. Ibid., 53:3oa. I9. Ibid., 53:I9a. 22. Lagerwey, Daoist Ritual. 21. Ibid., 56:28b. 24. Ibid., pp. 65-66. 23. Ibid., p. xiii. 26. Ibid., p. 45· 25. Ibid., p. 67. 28. Ibid., p. 46. 27. Ibid., pp. 37-48. 29. Schipper, "The Written Memorial," p. 3 I7. 30. Sangren, History and Magical Power, p. 171. 31. Schipper, p. 3I7. 32. Faure, Structure of Chinese Rural Society, p. 9· 33· Ibid., pp. 8o-82.

Notes to Pages I49-r64 34· Ibid., p. 8o. 3 5. There is, of course, every reason to imagine that, within the Daoists' own part of the celebration, they may continue to represent the villagers' gods as subordinates of their own and of themselves. The point is that the lack of a unified, focused ritual structure like that in Taiwan gives them no opportunity to make this representation stick at the level of the whole population of participants. 36. Hsu, Exorcising the Trouble Makers. 37· Ibid., pp. 91-92. 38. Ibid., pp. 99-102. 39· Ibid., p. 96. 40. Ibid., pp. 94-95. 41. Ibid., pp. 96-97. 42. Ibid., p. 97· 43· Others have certainly argued this as well, without treating the jiao itself in the way I am doing here. See Sangren, History and Magical Power, but also Weller, Unities and Diversities in Chinese Religion, esp. pp. 97-roo. 44· Hsu, Exorcising the Trouble Makers, pp. 48-49. 45· Lagerwey, Daoist Ritual, pp. 52-52. 46. Hsu, Exorcising the Trouble Makers, p. 56. 47· Ibid., p. 6o. 48. Pers. comm. from Theodore Huters. Chapter 6 Abbreviation: JFZ: Ge Yinliang, ed., ]inling fancha zhi (r6o7). Epigraphs:JFZ 2.24a, I.Ib. r. Elements of this argument may be found in Brook, Praying for Power, pp. 22-23, 321-23. 2. One could see in the distinction between state and public authority an analogy to the incomplete distinction in Chinese between guan ("official," "of the state") and gong ("public," "in the public interest"). Arguably, however, guan is narrower than "state," and gong broader than "public authority." Aspects of the guan/gong distinction are considered in Rankin, "The Origins of the Chinese Public Sphere," pp. 40-41. 3· The Hongwu regulations have been surveyed in English in Yii, The Renewal of Buddhism in China, 144-69. The main japanese studies are Tatsuike Kiyoshi: "Mindai no sokan"; "Minsho no jiin"; pp. 9-29; and "Ming taizu de fojiao zhengce," in Zhang Mantao, Ming-Qing fojiao shipian, pp. 1-16. See also Mano Senryu, "Mindai no bukkyo to Mincho," reprinted in his Mindai bunkashi kenkyu,pp.243-323. 4· Chikusa Masaaki, Chugoku bukkyo shakaishi kenkyu. For a useful summary of Chikusa's arguments, see Hansen, Review. 5· Pingshan tang tuzhi (1705), 6.13b, citing a text of the mid-Ming. 6. ]FZ r6.6a; Xu Yikui, Shifeng gao 11.8b; Taizu shilu (reprinted-Nanjing, 1940), 53.3a; Chu Hua, Hucheng beikao (reprinted-1935), 7.2a. 7· Mano Senryii, Mindai bunkashi kenkyu, p. 246. Song Lian's biographies of some Buddhist masters who took part in these convocations appear in ]FZ 16.27a-29a.

399 8. There was one other significant edict on religion promulgated in I 3 73: an order that there should be only one Buddhist and one Daoist monastery per county, and that all Buddhist and Daoist monks should be made to reside in these two monasteries (Ming huiyao, p. 694). This unusually stringent edict was applicable only to the six prefectures nearest Nanjing, and even there seems not to have been enforced; see Yi.i, The Renewal of Buddhism, pp. 145-46. After the establishment of registries in r 3 8 I, the edict was interpreted to mean that one Buddhist and one Daoist monastery should be designated in each county to handle religious affairs. 9· For the history of ordination certificates through the Tang and Song, see Chikusa Masaaki, Chugoku bukkyo shakaishi kenkyu, I7ff. ro. Regarding the purge ofHu Weiyong, see Hucker, The Ming Dynasty, 4rff; and Dreyer, Early Ming China, chap. 4· No text I have found links the purge and the change in Hongwu's attitude to Buddhism directly, yet the restrictive turn is consistent with the other shifts in policy following the destruction of Hu's faction. II. Two prefectural Buddhist registries were set up in Hangzhou in I37I and "early Hongwu" respectively (Wulin fanzhi r. r b, 20b). A subprefectural Buddhist registry was set up in Quanzhou, Guangxi, in I 3 78 (Xiangshan zhi 3 .2ob ). County Buddhist and Daoist registrars from the Yuan dynasty are mentioned in Jiaxing xianzhi (I685), 5.I9b-20a; Liu Dunzhen, Beiping huguo si canji, p. 2 5; and Cai Meibiao, Yuandai baihuabei jilu, p. 3 I. See also the conversation between the Hongwu emperor and a county registrar before the founding of the Ming included in Zhou Tiandu, Tongsu bian (I75Il, 20.2b. I2. In I382, county seats that also served as prefectural seats were excused from setting up a separate county Buddhist registry, their Buddhist affairs being delegated to the prefectural Buddhist registry UFZ 2.7a). I3. E.g., the subprefectural registry of Kaizhou, North Zhili, was founded in I382 (Kaizhou zhi [I534], 2.4b); the county registry of Yanshi, Henan, in I383 (Yanshi xianzhi [1504], shusi). !4· E.g., Yichuan county in eastern Yan'an, Shaanxi, never had a Buddhist or Daoist registry (Yichuan xianzhi [I753], 2.3b). 15. E.g., Caoxi tongzhi (1672), 3.13b. · I6. Dinghai xianzhi (I7II), 6.3a-b. I 7. A case of a county registrar fixing land grades for taxation is noted for I385 in]FZ 2.9b. I8. Chen Yuan, Mingji Dian-Qian fojiao kao, p. 1I9. I9. ]FZ 2.6b-7a. A second Hongwu edict on sect distinctions in I39I substituted the term yujia (yoga, or tantrism) for jiao, although the content of the two categories remained the same (ibid., 2.I5a-I6b). The same text, however, expressly forbade esoteric yogic teachings. Vinaya (lit) monasteries were included in the Teaching sect; see, e.g., Dinghu shanzhi, r. 10a. 20. Da Ming huidian (I588), I04.4a;]FZ 2.5b. 21. Yi.i, The Renewal of Buddhism in China, p. I 68; Marro Senryii, Mindai bunkashi kenkyu, pp. 270-7 I; see also ]FZ 2.22b, 24b;]ingci sizhi (I888), 27 .9bIoa. The term zhenji first appears in an imperial edict in I 3 72 requiring the land

400

Notes to Pages r67-I70

belonging to Nanjing's Jiangshan (Linggu) Monastery to be recorded in a "foundation register" (zhenji pu); see JFZ 2.1 b. 22. E.g., the prominent Chan monastery in Ningbo, Tiantong Si (Tiantong sizhi [r8n], 2.14b). 23. The gazetteer of Hangzhou's Bianli Cloister makes a point of saying that the cloister was unique among the more important monasteries of Hangzhou in not having been amalgamated under the I39I order (Bianli yuanzhi, Wu's preface, Ia). 24. E.g., Zhangzhou fuzhi (r6r3), 34.3b. 2 5. Parallels between early Ming and early Communist controls on Buddhism are striking. Both created monk-staffed bureaucracies to supervise Buddhism, intervened in landholding, closed monasteries and put them to other uses, concentrated monks in a few large centers, limited ordination, and restricted the free movement of monks; see Holmes Welch, Buddhism Under Mao (Cambridge, Mass.: Harvard University Press, I972), pp. 29-3 I, 42-45,73, 8o-8r, I I7-I8, 124-26. 26. In a study of early Ming community schools (shexue), Matsumoto Yoshimi has estimated that 9 5 percent of these new schools were housed in converted monasteries ( Chugoku sonraku seido no shiteki kenkyu, p. 475 ). Most of these were monasteries closed in the amalgamation campaign. Matsumoto's observation is attested in Hejian fuzhi (1540), 5.Iob, which notes that rural schools had been made over from Buddhist and Daoist properties. 27. Da Ming huidian 104.6b. 28. Yii (The Renewal of Buddhism in China, p. 146) reads the expression jiu e in the 1391 edict as "old quota" and argues that the emperor was banning monasteries that did not fall within the quota of one per county laid down in a 1373 edict. I read the expression as "old name plaque" and understand the Hongwu emperor to be banning recently constructed monasteries that did not have plaques; for corroboration, see Shaxian zhi (1701 ), 7.49a. For the bestowal of name plaques in the Song, see Chikusa Masaaki, Sodai bukkyo shakaishi kenkyu, chap. 2. 29. Ming lu jijie fuli 4.7a-b. 30. Xu Xueju, Guochao dianhui 134.2ob. 3 I. JFZ 2.24a-27b; the edict includes yet other restrictions besides those mentioned. 32· Guoxian zhi (I566), 8.9b. A similar observation from Fujian is made by the editor of Shaxian zhi (170I), 7.49a. 33· Da Ming huidian 104.2a, 4b-5a. This regulation has been misinterpreted in the secondary literature variously as a quota on ordinations or as a quota on the total number of ordained monks in a county. It applied only to novices, who had to be between the ages of I4 and 20 and have parental permission for their novitiates. 34· The quotas (4o/ 3oho) were identical; see Dreyer, Early Ming China, p. r 3 3. 3 5. Hongwu in I 3 87 restricted ordination to those under twenty years of age; in 1395, he defrocked all monks who failed a national examination (Da Ming huidian 104.4a, b). 3 6. Yu Jideng, Diangu jiwen, p. 228.

Notes to Pages r7o-r74

40I

37· Ibid., pp. 23I-32. 38. Lin Xiyuan, Lin Ciya xiansheng wenji 2.28b. 39· Da Ming huidian I04.8b. 40. Shimizu Taiji, "Mindai no jiden," pp. 6s-66; Wang Chongwu, "Ming Chengzu yu fojiao," pp. 99-IOI. A Fujian memorialist in the I53os urged that the hundred-mu limit be applied to every monastery in Jiangnan, noting that it was needed to reduce the number of monks (Lin Xiyuan, Lin Ciya xiansheng wenji 2.3rb). 41. Chen Renxi, Huang Ming shifa lu 39.41b-42a. 42. E.g., Darning fuzhi (rso6), 6.46a; Mianyang zhi (1531), 7.9b. 4 3. Unlike the other county officials listed in the gazetteer for Lushan county, Henan, the Buddhist registrar is listed not under the heading guan but among the yuan (Lushan xianzhi [I552], 4.I9a). 44. Qingliang shanzhi (I 66 I), 3. I b. The stipend was originally instituted in 1392; see Mano Senryo., Mindai bunkashi kenkyu, 264. 4 5. E.g., the Buddhist Registry of Haiyan county, Zhejiang, successfully petitioned to have the grounds of Zisheng Monastery restored in 14 3 5 (Haiyan xian tujing 3 .6oa). Registrar Liangjin of Hangzhou prefecture not only restored land that a monastery had lost in r 5 6o but erected a stele of protection outlining the history of its acquisition and its exact location (Yunju shengshui sizhi 2.3a). 46. ]FZ r6.roa. 47· E.g., Hejian fuzhi (r54o), 4.5b. 48. Neixiang xianzhi (1485), 2.19a, 4.7ob. 49· Yingzhou zhi (rsrr), 4.I5a. 50. Tianzhu shanzhi (187 5 ), 8.57b. See also note 45 to this chapter regarding Liangjin. 51. E.g., both Ruian xianzhi (1555), ro.8ob, and Shizhong shanzhi (I883), 3. 7b, refer to the local registry as a long-abandoned site. The list of incumbents in the post of registrar in ]iujiang fuzhi (I592), 6.3oa-33b, ends with the Chenghua era. In Changle xianzhi ( I64I), 5.58b, a record of temple restoration dated I 5 28 observes that as of that time there were no longer any Buddhist or Daoist registrars in the county. 52· E.g., Guangshan xianzhi (I 5 56), 3. 5b, notes that "the county Buddhist registry of old was located in Baoxiang Monastery, but for a long time no official has been appointed." 53· Songjiang fuzhi (I5I2), I8.Ia. 54· Wujiang xianzhi (I56I), I6.2b-26a. 55· Wang Chongwu, "Ming Chengzu yu fojiao," pp. 93, 96. 56. Concerning the reputations of the Tianshun, Chenghua, and Hongzhi emperors with regard to Buddhism, see Noguchi Tetsuro, "Mindai chiiki no bukkyokai," pp. I93-94, 2I2, 225. Regarding Hongzhi, see Meskill, Ch'oe Pu's Diary, p. I46. 57. Qingliang shanzhi (I 66 I), 4· I 2a. 58. See, e.g., the complaints of Supervising Secretary Zhang Gu (Zhengtong era) and of Minister of War Yu Qian and Censor Ye Luan (Jingtai era), in Yu Jideng, Diangu jiwen, pp. 209, 2I6, 227.

402

Notes to Pages I74-r8z

59· Da Ming huidian I04.3b. 6o. E.g., Hasebe Yukei, Min-Shin bukk yo shi kenkyft josetsu, p. 34 7. 61. Ping-ti Ho, The Ladder of Success, p. 33· 62. Wang Hongzhuan, Shan zhi (I788), 4.22b. 63. Hu Lao, "Sengdao hudie shu," in Huang Ming jingshi wenbian 19.9a; see also Yingzong shilu, 23 .6b. 64. ]ianning fuzhi (I54I), I9.58b. 6 5. One of the last such diatribes I have encountered appears in a memorial presented to the Jiajing emperor by Minister of Rites Fang Xianfu (d. I544) in the I530s (YuJideng, Diangu jiwen, p. 303). 66. Fuzheng', Hanshan dashi nianpu shuzhu, p. I8. 67. Da Ming huidian I04.7b. For an earlier (I546) expression of concern about mass ordination in Beijing, see Overmyer, Folk Buddhist Religion, p. I7I. 68. To mention a few examples: the great Vinaya monastery in Hangzhou, Zhaoqing Si, started rebuilding in I 573 and resumed ordination in I 5 87 once the project was completed (Da Zhaoqing liisi zhi I. I 5b); the ordination platform at the Chan ordination monastery in Suzhou, Kaiyuan Si, was restored in I598 (Kaiyuan sizhi I4a); restoration of the monasteries on Putuo Island, where lay ordinations were popular, began in I6o2 with the help of a grant from the Wanli emperor (Putuo shanzhi [I6 I I], 2. 7a); and Hangzhou's other great Vinaya monastery, Jietan Yuan (Ordination Platform Cloister), was restored in I6I3 and its ordination platform re-established in I628 (Longxing xiangfu jietan sizhi I .I 3a). 69. Canglang xiaozhi (I696), I.I9a. 70. The first instance of the empress dowager's undertaking of a Buddhist project outside the palace was the sending of 3,ooo artisans to Wutai Mountain in I 579 (Fuzheng, Hans han dashi nianpu shuzhu, p. 4 5 ). 71. Johannes Prip-M0ller, Chinese Buddhist Monasteries (Copenhagen: Gads Forlag, I937), P· 73· 72. Caoxi tongzhi (I672), 4.2oa. 73· Ayuwang shanzhi (I6I9), 4b.8b-Ioa. 74· Li Zhi, Xu fen shu, p. 97· 75· Bochi shanzhi (I92o), 48b, from a stele by Tao Se. 76. Kristofer Schipper, "Vernacular and Classical Rituals in Taoism," journal of Asian Studies 45: I (I985): 45· 77· This argument is laid out more fully in Brook, Praying for Power, 3 I I25. 78. Songjiang fuzhi (I5I2), I8.2oa, although this is an early example. 79· Hejian fuzhi (I 540), 7.6a, interpreting yi literally; Yuanshi xianzhi (I642), 4·24b. Chapter 7 Earlier versions of this paper were presented at the Conference on Canon Formation in Late Imperial China held at Laguna Beach, Calif., June I992, imd at a follow-up workshop on culture and the state in China and Japan, held at the University of California, Irvine, in Aug. I993· My thanks to the conference

Notes to Pages r8z-r84 organizers and particpants, especially to the two discussants for the first version of the paper, Timothy Brook and Bruce Lincoln. Helpful comments were also provided at various points along the way by Maureen Graves, Steven Sangren, Anne Walthall, and R. Bin Wong. Katherine Carlitz and Susan Naquin directed me to important sources I might not otherwise have found. The research was largely conducted during a year's leave funded by a President's Research Fellowship from the University of California and a Chinese Studies Fellowship funded by the Ford Foundation and the National Endowment for the Humanities and administered through the American Council of Learned Societies. 1. See, e.g., Taian xianzhi (1934) 12.19b; Daozang (1607), ro63.2b; Jin Qi, Taishan zhi 10.2ob; Chavannes, Le T'ai Chan, p. 29; Luo Xianglin, "Bixia Yuanjun," p. 61. 2. Quoted in Luo Xianglin, "Bixia Yuanjun," p. 5. 3· For estimates of pilgrimages to the Taishan temple, see below; on local temples in North China, see Arthur Smith, Village Life in China, p. 140; on Bixia Yuanjun temples in Manchuria, see Sawada Mizuho, Shina no minkan kamigami, p. 30. See also Naquin, "Peking Pilgrimage," pp. 3 34-37. 4· Naquin, "Peking Pilgrimage," esp. pp. 345, 354· 5. There is a Bixia Yuanjun cult in Fujian that bears some relation to the one in the North, but the exact relationship is murky. And although there are some temples to the goddess of Taishan mentioned in the gazetteers of big Lower Yangzi cities, the cult was never strong south of the North China macro-region. On the Fujian Bixia Yuanjun, see. e.g., Werner, Dictionary of Chinese Mythology, pp. 373-74; on Jiangnan temples to the Taishan Bixia Yuanjun, see Dudbridge, "A Pilgrimage," p. 242, n.45. 6. The idea of a common set of symbols and practices that serve to paper over disagreements as to the meaning of these surface phenomena has been most influentially developed by James Watson; see particularly "Standardizing the Gods." Many people have disagreed with parts of Watson's formulation of how the symbiosis among state, elite, and popular efforts works. However, virtually all the work we have on popular religion-except that on cults proscribed by the state-has told a "success" story, in which some sort of symbiosis and co-optation occurred, and the cult in question has thrived despite, or even because, there was no universally shared interpretation of its meaning. See, e.g., Wolf, "Gods, Ghosts, and Ancestors"; and Weller, Unities and Diversities in Chinese Religion. Recent work by Prasenjit Duara ("Superscribing Symbols") and especially by Steven Sangren (History and Magical Power) pays more attention to clashes over ideas and conflict within symbiosis, but both are studies of approved cults and so remain within the realm of success stories. One of the few works that does focus on a case in which an extremely popular cult was not originally perceived as "heterodox," but on which symbiosis nonetheless did not take is von Glahn, "The Enchantment of Wealth." 7· For a parallel instance of how rules on official worship could be relaxed during crises, allowing the state to tap kinds of cosmic power it normally eschewed, see David Johnson, "The City God Cults of T'ang and Sung China," pp 444-45. 8. See, e.g., Gu Yanwu, Rizhi lu 8: 77-8o (juan 25).

Notes to Pages r8s-r89 9· Naquin, "Peking Pilgrimage," pp. 3 3 8-39, suggests that this was not the case at Miaofengshan, although it was perceived to be so. I present evidence below to suggest that this genuinely was the case at Taishan, but if it was not, this might actually strengthen the argument made here: social fears would seem all the more important a part of elite reactions to the pilgrimage if they imagined a female predominance that did not in fact exist. IO. See Duara, "Superscribing Symbols"; and von Glahn, "The Enchantment of Wealth," for recent work that, although very different, does call attention to this flip side of canonization. II. Gu Yanwu, Rizhi lu, 5: I05 (juan 14); 2: 47-48 (juan 4). I2. Nie Jianguang, Taishan daoli ji, p. 48; Taylor, "Official and Popular Religion," pp. 148-49. I3. Gu Yanwu, Rizhi lu, Io: 28 (juan 30); Sawada Mizuho,Jigokuhen, pp. 43-53; Huang Yupian, Poxie xiangbian, p. 92; Ayscough, A Chinese Mirror, p. 361. On the formation of a new notion of the underworld in which Yan Wang was a key figure, see Teiser, "The Growth of Purgatory." The drama Xiao Zhang tu fen er jiu mu zaju does describe such a pilgrimage and probably dates from the thirteenth century, but I know of no earlier reference. I4. Luo Xianglin, "Bixia Yuanjun," p. I I. I5. See, e.g., Taian xia11zhi (I934) I 2. I9b; Nie Jianguang, Taishan daoli ji, p. 76. I6. Jin Qi, Taishan zhi IO.I7b; Sawada Mizuho, Shina 110 minkan kamigami, pp. 31-33· I7. See esp. Daozang 1063.2b-7b; and also Tschepe, Der T'ai-scha11, p. Ioo. I8. Zha Zhilong, Dai shi (hereafter DS), pp. 62, 87, 93, 97, 98, Ioo, I02, I05, IIO, II6, II?, I54· I9. Tschepe, Der T'ai-schan, pp. Io5-6. 20. Sawada Mizuho, Chiigoku 110 mi11kan shi11ko, p. 307; Tschepe, Der T'aischan, p. Ioo. 2 r. See, e.g., Cui Xiuguo and Ju Aiqin, Taidai shiji, pp. I I 6, I 67, I 8 3, I 8 6. 22. On patronage at Beijing area temples, see Thomas Li and Susan Naquin, "The Baoming Temple," esp. pp. I40, I74, I87-88; and also Naquin, "Peking Pilgrimage," esp. pp. 353-54. 23. For descriptions of offerings, see, e.g., Taian zhouzhi 4.2Ia; Sawada Mizuho, Chiigoku no minkan shinko, p. 3o I; Tschepe, Der T'ai-schan, pp. 9 2-9 3. 24. On Huashan, see Huayue zhi (I873); on Junxian, see Donald MacGillvray, "Memoirs, Letters, Articles," pp. I82-83, 2I7, 224; and also letter of Leslie Mcintosh, Jan. 25, I893, all in Box 7, File II3, "Materials Collected in I935 for the Writing of the History of the Honan's 5oth Anniversary," North China Mission, United Church (of Canada) archives, Toronto; on events for pilgrims passing through on their way to Taishan, see, e.g., Linqing xianzhi (I934), jianzhi 22a. 2 5. List (available from author upon request) compiled from approximately IOO North China gazetteers, mostly from the nineteenth and twentieth centuries. 26. On Heze, see Caozhoufu Heze xian xiangtu zhi (I9o8). Overall, I have thus far found I 34 Bixia Yuanjun temples mentioned in the gazetteers of 7 4

Notes to Pages 189-194 North China (mostly Shandong) counties surveyed. Since many of these gazetteers list only "orthodox" temples and/or only those in the xian capital, the real totals must have been far higher. 27. Xuxiu Qingping xianzhi (1936), lisu r8a-r9a. 28. See, e.g., Linqing xianzhi (1934), jianzhi, 2r.b; Boshan xianzhi (1937) 2.3a; Yangxin xianzhi (1926) 8.29a. 29. See, e.g., Dudbridge, "A Pilgrimage," p. 236. 30. Ibid., pp. 236-37, 239-42; Pei-yi Wu, "An Ambivalent Pilgrim," pp. 73-74· 31. Huang Liuhong, Fuhui quanshu 3r.ua; Dudbridge, "A Pilgrimage," pp. 235-39; Hubbard, "The Pilgrims of Taishan," pp. 322-30. 32. Huang Liuhong, Fuhui quanshu 3r.rob-ub; Boshan xianzhi 4.I4a. 33· E.g., Xuxiu Fanxian zhi (193 5) 6.r6b-r7b; Guanxian zhi (r83o) 9·I7bI9a; Yangxin xianzhi (1926) 2.3a; Deping xianzhi (1893), pp. 609-rr. 34· Xuxiu Fanxian zhi (193 5) 6.2ob; Guanxian zhi (r83o) 9.17b-19b; Yanggu xianzhi (1942) 2.p; Yangxin xianzhi (1926) 8.29a. 35· Yangxin xianzhi (1926) 8.29a. 36. Chaocheng xian xiangtu zhi (1920) r.2ra, 26b. 37· Yangxin xianzhi (1926) 8.29a. 3 8. On female deities as refuges for the despised, see Sangren, "Female Gender in Chinese Religious Symbols," pp. 14, r 9-20; on the sociology of city god temples, see Stephen Feuchtwang, "School Temple and City God"; on Bixia Yuanjun and the underworld yamen, see Daozang ro63.2b; and Sawada Mizuho,Jigokuhen pp.48-49. 39· See, e.g., Duara, Power, Culture, and the State, p. 138. 40. See, e.g., Goodrich, The Peking Temple of the Eastern Peak. 41. Although on the Beijing area, see again Naquin, "Peking Pilgrimage," pp. 338-39· 42. For the process of "Confucianization" and the modifications of old myths that it required, see Duara, "Superscribing Symbols," esp. pp. 782-90. 43· DS, pp. 58, 6o, 63, 73,93-94, 106-7, u6, 132, 147. 44· DS, pp. 56, 58, 64, 88, 96, 106, u6, 132, 145. 45· DS, pp. 58, 61, 63, 64, 67, 73, 107, 116. 46. DS, p. 70; Daozang 1063·3a-4a, sa, 7b; Jin Qi, Taishan zhi 10.21a. 47· Daozang 1063, esp. 2b, 7b. 48. Taian xianzhi 12.19b; Jin Qi, Taishan zhi 10:2ob. See also Maspero, Taoism and Chinese Religion, pp. 165, 171. 49· Maspero, Taoism and Chinese Religion, pp. 104, 166, 168. 50. Watson, "Standardizing the Gods," pp. 312, 320. 51. Tao Yang eta!., Taishan minjian gushi daguan, pp. 267-28. 52. Watson, "Standardizing the Gods," pp. 295, 320; Dudbridge, The Legend of Miao-shan, pp. 41-42, 55, 89-92,96. According to a 1984 sample cited in Luo Zhufeng, Religion Under Socialism in China, approximately So percent of the visitors to the Guanyin shrine at Putuoshan were women. 53· F. Dransmann, T'aishan-Kufow Guide, p. 262; Chavannes, Le T'ai Chan,

p. 69.

Notes to Pages I94-I97 54· Sangren, "Female Gender in Chinese Religious Symbols," esp. pp. roI5, 2I, 23-24. 55. For the exceptions, see Gu Yanwu, Shandong kaogu lu (quoting Zhang Hua, Bowu zhi), shang Ioa-rrb (which has her returning to Taishan from an unknown husband in the West); and Werner, Dictionary of Chinese Mythology, p. 404 (a story that she married the legendary hero Mao Ying). 56. Dudbridge, The Legend of Miao-shan, pp. 4I-42, 55, 89-92, 96; Sangren, "Female Gender in Chinese Religious Symbols," p. 8. 57· Nie Jianguang, Taishan daoli ji, p. 43· s8. Chen Qinghao and Wang Qiugui, Shandong minjian gushi ji, pp. II15; Tao Yang eta!., Taishan minjian gushi daguan, pp. 12-13, 31-3 2, I s6, 160, 162, 164, 245, 393-94; Sun Jingxian, Shandong zhi minjian chuanshuo, pp. 2-3· 59· Originally in Zhang Hua, Bowu zhi; quoted in full in Gu Yanwu, Shandong kaogu lu, shang 1oa-b. 6o. Stenz, Beitri:ige zur Volkskunde Sud-Schantungs, p. 37· 6r. Goodrich, The Peking Temple of the Eastern Peak, pp. 62-63; Tschepe, Der T'ai-schan, p. 91; Jin Qi, Taishan zhi 10.21a; Bergen, "A Visit to Tai Shan," p. 545· 62. See, e.g., the photo of a polychrome Bixia Yuanjun at the end of Jacques Van Goidsenhoven, Heros et divinites de la Chine, which has feet smaller than its hands; also Bergen, "A Visit to T'ai-shan," p. 545; Dransmann, T'ai-shanKufow Guide, p. 82. 63. Tao Yang eta!., Taishan minjian gushi daguan, pp. 12-13, 31-32, 156, 160, r62, r64, 245, 393-94; Sun Jingxian, Shandong zhi minjian chuanshuo, pp. 2-3· 64. Tao Yang eta!., Taishan minjian gushi daguan, pp. I2-I3, 3I-32, rs6, r6o, r62, 164, 245, 393-4; Sun Jingxian, Shandong zhi minjian chuanshuo, pp. 2-3· 65. Taian zhouzhi 4.21a; Sawada Mizuho, Chugoku no minkan shinko, p. 30I; Tschepe, Der T'ai-schan, pp. 92-93; Zhaoyuan xianzhi (1846) 3 .I sb (local offerings). 66. Maspero, Taoism and Chinese Religion, p. I7I. 67. Mateer, "T'ai san-Its Temples and Worship," p. 4IO. 68. Goodrich, The Peking Temple of the Eastern Peak, p. 63. 69. Yuan Mei, Zi buyu, p. 9, see also pp. 8I-83, I76-77. See also Maspero, Taoism and Chinese Religion, p. Io5. 70. Yuan Mei, Zi buyu, p. 9· 7!. Ibid, pp. I 24-5 72. See the translation of the Taishan pilgrimage chapters by Glen Dudbridge in "A Pilgrimage," pp. 39-64. For a longer discussion of how Bixia Yuanjun's role as patroness of fox-spirits fits the novel's structure of retribution, see Pomeranz, "Room at the Top?," pp. 49-5473. Plaks, "After the Fall," esp. p. 545; and see pp. 568-75 for the decline of virtue and the Ming empire in the Xingshi yinyuan zhuan. 74· See, e.g., Chavannes, Le T'ai Chan, pp. 3I, 37-8. 0

Notes to Pages I97-zor 7 5. Some of these other figures include King Wen and Zhang Xi an; see, e.g., Changqing xianzhi (1934), ro.5a-b; Goodrich, The Peking Temple of the Eastern Peak, pp. 57-60; Jiang Wenhua, Shandong minsu, p. I74· 76. Jiang Wenhua, Shandong minsu, pp. 174-75; Goodrich, The Peking Temple of the Eastern Peak, pp. 55-56. 77. Goodrich, The Peking Temple of the Eastern Peak, p. 56; Stenz, Beitrage zur Volkskunde Sud-Schantungs, p. 68; Maspero, Taoism and Chinese Religion, pp. r65-r66. 78. Jiang Wenhua, Shandong minsu, pp. 174-75. 79· See, e.g., George Wilder, letter of Apr. ro, 1921, in Wilder papers, Bentley library, University of Michigan. The issue of types of reciprocity in asking for children and rain is also discussed in Pomeranz, "Water to Iron," pp. 83-84. So. Stenz, Beitrage zur Volkskunde Sud-Schantungs, p. 68. 8 I. For a reading of funeral rites as in part a fantasy of "pure" reproduction by a lineage of male agnates, see Watson, "Of Flesh and Bones," esp. pp. r788o; for a somewhat different reading, which sees both this fantasy and a different story that recognizes the indispensability of women from outside the lineage, see Stuart Thompson, "Death, Food, and Fertility," esp. pp. 95, 98, roo. 82. E.g., DS, pp. s8, II6, 132, 139, !45, 147; Changqing xianzhi (!835) 1o.9b; Taian xianzhi (1934) 14.5P-55a. 83. DS, 58, 64, 108, 132, 145; Yuan Mei, Zi buyu, p. 108; Jin Qi, Taishan zhi 10.21a; Isabelle Williamson, Old Highways in China, p. 142. 84. DS, pp. 96, 108, 132, 145. 85. Elvin, "Female Virtue and the State," pp. 123, 127-29. 8 6. Tao Yang eta!., Taishan minjian gushi daguan, pp. 262-64; "Xiao Zhang." My thanks to Katherine Carlitz for mentioning this source to me, and providing a copy of the text. 87. DS, pp. 88, 96, II5, 145· 88. Yuan Mei, Zi buyu, p. ro8. 89. On Miaoshan, see Dudbridge, The Legend of Miao-shan; on women's marriage laments see Martin, "Gender and Ideological Difference," pp. 270-7r. See also numerous folksongs, many collected in Eastern Henan (not far from Taishan), in Liu Jingan, Geyao yu funu, pp. 8, 13, 34, 36-37, 53-55, 62-62, 72-73, 179· 90. See, e.g., the cases cited in Elvin, "Female Virtue and the State in China," pp. 142-43· 9r. Yuan Mei, Zi buyu, p. ro8. 92. Huayuezhi(1873)2.35a. 93· "Xiao Zhang," pp. 719-20, 722-23. 94· DS, pp. 58, 64, 88, 96, 108, II5, 132, 145; Nie Jianguang, Taishan daoli ji, p. 92; Taian zhouzhi 4.2oa-21a; Jin Qi, Taishan zhi 6. 3 ra-b, 19.2a-b. For late Qing and Republican references, see the following note. 95. For later references to this practice, see Geil, The Sacred Five of China, pp. 107-9; Isabelle Williamson, Old Highways in China, p. 142; Bergen, "A Visit to Tai Shan," p. 546; Hubbard, "The Pilgrims of Taishan," p. 3 29. 96. See, e.g., Elvin, "Female Virtue and the State in China," pp. 127-29.

408

Notes to Pages zor-zoS

97· Ibid., p. I23. 98. DS, pp. 96, Io8, 145. 99· Yuan Mei, Zi buyu, pp. 9, 81-83; Wieger, Folk-lore chinois modeme, p. 150. roo. See von Glahn, "The Enchantment of Wealth," pp. 657-60, 685,697. 10I. See generally Sawada Mizuho, Jigokuhen, pp. 43-58. I02. Tao Yang eta!., Taishan minjian gushi daguan, pp. 262-64, 265-67. 103. Huayue zhi (I873) r.3 sa. I04. Yangxin xianzhi (I926) 5.7oa-b. I05. Tao Yang eta!., Taishan minjian gushi daguan, pp. 262-4. Io6. "Xiao Zhang," pp. 722-24. I07. See generally Pomeranz, "Room at the Top?" pp. 47-70; on pilgrimages, see esp. Gu Yanwu, Rizhi lu, 5: 38-39 (juan I3), IO 69 (juan 3 1); Huang Liuhong, Fuhui quanshu 3 I. I I a; and the Taishan pilgrimage chapters of Xingshi yinyuan zhuan translated by Dudbridge in "A Pilgrimage," pp. 39-64. Io8. Quoted in Hart, Western China, p. 267. I09. Watson, "Standardizing the Gods," pp. 299-313.

Chapter 8 I thank Theodore Huters for indispensable comments on earlier drafts of this paper. r. "Houguan xian lienii jian chou" (The martyr of Houguan xian exterminates her foe). Patrick Hanan discusses "The Martyr" in The Chinese Vernacular Story (hereafter CVS) pp.I24, I26. See note 49 below for reference to source. I have retained the titles Hanan uses in CVS for all the huaben discussed in this chapter, as an aid to continuity in the scholarly literature. 2. "Jiang du shi xiao fu tu shen" (A filial daughter-in-law has herself butchered in Yangzhou). See note 5 I below for reference to source. Hanan discusses "Siege" in CVS, pp. 123, I36-37. 3· McMahon, Causality and Containment, pp.12-I4. 4· The only study of Langxian and his works to date is Hanan, CVS, chap. 6 ("Langxian"), pp.I20-39· 5. By "politically engaged fiction" I mean that Langxian takes on, explicitly, some of the terms and issues that structured Ming discourse about the constitution of society and the distribution of power. But a politically informed reading of these stories does not reduce them to a simple narrative of conflict between classes or groups. Rather, I feel that we cannot grasp the achievement of these stories, or even discern their immanent structure, unless we work to recover the likely concerns of the author in the fullness of his historical situation. This sort of "political" reading obviously owes much to Frederic Jameson, most notably The Political Unconscious, chap. I ("On Interpretation: Literature as a Socially Symbolic Act"), pp.I7-I02. See especially his sensitive discussion of Deleuze and Guattari and their impassioned plea against reductionism in interpretation. In its details, my argument is also indebted to Goldmann, The Hidden God; Mitchell Cohen, The Wager of Lucien Goldmann, esp. pp. I 54-2oo; and the

Notes to Pages zo8-zrr discussion of Goldmann and other French critics on Pascal and the Jansenists in Frederic Jameson, The Political Unconscious, pp. 43-46, and idem, Marxism and Form, pp. 376-82. 6. Hanan discusses Qing shi under the title The Anatomy of Love in CVS, pp. 95-97. See Hua-yen Li Mowry, "Ch'ing-shih and Feng Meng-lung," for an analysis of Qing shi and translation of selected passages. 7. For a detailed analysis, see Miller, "Genre as Social Action." 8. See, e.g., the Yuan zaju Pen'er gui (The ghost of the pot), where the protagonist is a young merchant. In the fifteenth-century chantefable version of the story, "Bao Longtu gong'an duan wai wu pen zhuan" (Judge Bao solves the case of the misshapen black pot), the protagonist is a young man who sets out for examinations against his parents' wishes. These fifteenth-century chantefables are collected in Ming Chenghua shuochang cihua congkan shi liu zhang fu Bai tu ji yi zhang (hereafter Shuochang cihua). "The Misshapen Black Pot" is no. 5. Similarly, the Imperial Uncle Cao steals a shopkeeper's wife in the Yuan zaju Lu Zhailang, whereas he steals an examination candidate's wife in the chantefable "Bao Longtu duan Cao guo jiu gong'an zhuan" (Judge Bao solves the case of Imperial Uncle Cao), Shuochang cihua, no. 6. For the discovery and themes of these Chenghua era chantefables, see King, "Discovery and Restoration of the Texts in the Ming Chenghua collection"; McLaren, "Ming Chantefable and the Early Chinese Novel"; and idem, "The Discovery of Chinese Chantefable Narratives." 9· E.g.,the Kongmen rujiao liezhuan (Biographies of Confucian disciples) in the Beijing Library, with half-page illustrations of miraculous deeds. Io. "Shilang fuma zhuan" (The biography of Imperial Son-in-law Shi), Shu achang cihua, no. 2, p. I I b. For a summary of this chantefable, see Carlitz, The Rhetoric of "Chin p'ing mei," PP·33-35· I r. "Bao Longtu duan bai hu jing zhuan," Shuochang cihua, no. 8. I2. "Renzong ren mu zhuan," Shuochang cihua no. 4C. I 3. See Plaks, Four Masterworks; and McMahon, Causality and Containment, passim. 14· Hanan, in CVS, p.125, sees the maid Chunlai in the story "Qu Fengnu" (Rocks, no.4) as "lightly based" on PanJinlian's maid Chunlai. In fact, Langxian used as a link a plot element so well known that it went on to gain a life of its own in popular performing literature. In similar scenes, Qu Fengnu's mother and Ximen Qing's first wife, Wu Yueniang, are startled and discomfited when the former maid appears in elegant dress, followed by maids of her own. (See ]in ping mei, chap. 89.) For a zidishu version of this incident, see "Yong fu si" (Yong fu Temple) in Zidishu cong chao, pp. 46I-73. In "Lu Nan," the son of a Lu bondservant is frightened to death by a cat, as is Ximen Qing's son Guan'ge in chap. 59 of ]in ping mei. Since this is a completely unmotivated detail in "Lu Nan," one can only assume that it was planted to call ]in ping mei to mind. I 5. See Pei-yi Wu, The Confucian's Progress, pp.93-9 5. I6. See his "Zhonglu shuyuan ji" (Essay on the Zhonglu Academy), in Li Kaixian, ]i, p. 668. The Mencian "childlike heart" owes its Ming prominence to the way it dovetails with Wang Yangming's influential teachings.

4IO

Notes to Pages

ZII-zi6

I7. For a splendid example, see Li's long epitaph for Liu Tianmin in Li Kaixian, Ji, p. 377· Liu confides surreptitiously to Li Kaixian his fears that his unorthodox ranking of Song dynasty exegetes could get him into trouble. Li Kaixian clearly felt that literature and thought were instruments of both control and resistance. I 8. While this may seem to contradict the thesis of Peter Bol 's chapter in this volume, it actually demonstrates the hegemony of ru discourse achieved by the initial assent of the shidafu to the ru examination curriculum. Mid- and late Ming intellectuals were steeped in the Confucian canon, and their critique of the Ming curriculum is in exactly the same discursive universe as is the assent of the intellectuals of I 3 I 3. I9. See Feng's preface to Xing shi heng yan (Constant words to awaken the world), and Hanan's discussion of it on CVS, p. 78. 20. For a discussion of the elaboration of this theory, see Lowry, "Excess and Restraint." For late Ming fiction and drama theory, and their debt to sixteenthcentury thinkers, see Hanan, CVS, chap. 4 ("Feng's Life and Ideas"), esp. pp. 75-82, and chap. 7 ("Ling Mengchu"), esp. pp. I45-48. 2I. Epitaph for Liu Xiang, Li Kaixian, Ji, p. 429. 22. Hanan, CVS, p. I44· 23. Ibid., p. 78. 24. "Wang Benli tian ya qiu fu" (Wang Benli goes to the ends of the earth to search for his father), Shi dian tou, no. 3. 25. See Rawski, "Economic and Social Foundations of Late Imperial China." 26. For a particularly good example of this paradox, see the exquisitely printed Ren jing yang qiu (Spring and autumn mirror for humankind), by Wang Tingna (.fl. I 573-I62o), juan 20, Ming edition in Beijing Library. 27. This point, and in fact everything said here about late Ming material culture, can be related to the issues of late Ming connoisseurship analyzed by Craig Clunas in Superfluous Things. 28. See Hanan, CVS, p. I44· 29. Qiu Jun, Daxue yanyi bu, juan no, "Ming fu chou zhi yi" (Clarifying the requirements of righteousness in avenging wrongs). Qiu Jun's solution is to give the state the opportunity to punish the malefactor first, and turn to individual acts of vengeance only after exhausting other remedies. 30. Chapter references are to the I2o-chapter recension of the novel. Plaks (Four Masterworks, p. 3 56) notes how popular Li Kui was with Ming literati. 3 I. See Carlitz, "The Role of Drama in the Chin p'ing mei," pp. 428-3 6, for a scene-by-scene summary of Bao jian ji. 32. See the epitaph for Liu Tianmin cited in note I7 above. 33· Tanaka Issei, "The Social and Historical Context of Ming-Ch'ing Local Drama," pp. I52-59· 34· For the recitation of Shui hu zhuan in gentry households, see the example of Wang Daokun in Hu Shiying, Huaben xiaoshuo gai lun, p. 365. For ritual performances of Five Relationships, see Tanaka Issei, "The Social and Historical Context of Ming-Ch'ing Local Drama," p.I5o, n.29.

Notes to Pages 2r6-227

411

3 5. The fullest discussion of this is in McLaren, "The Discovery of Ming Chantefable Narratives." 36. Mote, "Confucian Eremitism in the Yuan Period," p. 234; and Kang-i Sun Chang, The Late-Ming Poet Ch'en Tzu-lung, pp. s-6. 37· The Shanghai, Beijing, and Harvard-Yenching libraries all hold editions of Pin shi zhuan. 3 8. See Hanan, The Chinese Short Story, pp. 66-70, where Langxian is designated "X"; and idem, CVS, pp. 120-21. 39· It can be and has been argued that late Ming expectations as to what constituted a huaben were different enough from our expectations of the modern short story that we should not expect the "two emphases" above to articulate as they would in modern fiction. See Hanan, CVS, pp. r 28-29. 40. "Xu lao pu yi fen cheng jia" (Old retainer Xu establishes a household to manifest his righteous indignation), Constant Words, no. 3 5; "Guo Tingzhi bang qian ren zi" (Guo Tingzhi acknowledges his son before the examination announcement), Rocks, no. r; "Lu taixue shi jiu ao gong hou" (Scholar Lu, lover of poetry and wine, refuses to humble himself before men of high position), Constant Words, no. 29. Hanan discusses "The Old Retainer" in CVS, pp. 123-24, and "Lu Nan" in CVS, p. 122, and points out that while "Guo Tingzhi" does not share the characteristic style-markers of the rest of Langxian's oeuvre (The Chinese Short Story, p.72), it fits in well thematically (CVS, p. 230, n.5). 41. Hanan, CVS, p. 121. 42. Pfau, "The Pragmatics of Genre." 43· In "The Concept of Moral Mind," Tang Chi.in-i discusses liangzhi in its universal, moral aspect. DeBary's discussion of Wang Gen shows how the emphasis in discussions of liangzhi shifted subtly from the universal to the individual ("Individualism and Humanitarianism," pp. 162-63). Hanan shows how ideas of liangzhi fused with those of qing in the late Ming literary world (CVS, pp. 78-79). 44· See Plaks, Four Masterworks, p. 57, for Wang Shizhen, and p. 3 56, for Li Zhi. 45. "Ren xiao zi lie xing wei shen" (The heroic temper of Ren the filial son transforms him into a god), Stories Old and New, no. 3 8. Hanan discusses "Apotheosis" in CVS, pp. 66-67. 46. See Ling Mengchu's comments on Feng, quoted in Hanan, CVS, p. 144. 47· See Lowry, "Excess and Restraint," p. 107; and Hanan, CVS, pp. 78-79. 48. This celebration of violence is a far cry from Qiu Jun's formal and rhetorically tentative justification of filial vengeance. 49· "Shentu shi," in Qing shi, pp. 8-9. 50. For allusions to Bao jian ji in ]in ping mei, see Garlitz, "The Role of Drama in Chin p'ing mei," pp. 324-56. 51. ]iu Tang shu, "Lieni.i," juan 205. 52· See Li.i Kun's Gui fan (Female exemplars), 1617 Huizhou ed., 3.37a-38a. 53· Tang furen's story can be found in Gui fan 3.19b-2Ia ("Tang shi ru gu"), the anonymous Huitu lienii zhuan (Illustrated biographies of notable

412

Notes to Pages 227-237

women) published in Huizhou in the 162os, 8.14b-16a ("Tang furen"), and Huang Shangwen's Nu fan bian, juan 2. For the 1673 commendation, see the 1673 ]iading xianzhi, juan 17 ("Li xiao fu ru gu"). 54· See Garlitz, "Desire, Danger, and the Body," p. III. 55. In chap. Ioo of fin ping mei, the old monk Pujing "proves" to Wu Yueniang that her son Xiaoge is the reincarnation of his father, although her maid has just seen the monk offer Ximen Qing's spirit a different reincarnation. s6. The story of Cao E is included in Huitu lienu zhuan, 6.I6b-I?b, and Cui fan, 2.I1a-12a. 57· See Meng Yuanlao, Dongjing meng hua lu, p. 163, for a proclamation issued during the Song Zhiping reign period (1064-68 ). This proclamation is quoted in full in "Siege." That Wu Zixu was still a "wronged hero" for many Ming literati is clear from a jueju "Ti Zixu miao" (An inscription for the Wu Zixu Temple) by Tang Yin (14 70-1523 ); see Wang Yingzhi, Ming ren jueju sanshi jia shang ping (Hefei: Huangshan shu she, 1991), p. 88. 58. "Xuc lu shiyu fu deng xian" (Recorder Lu, in the garb of a fish, is shown to be an immortal), Constant Words, no. 26. Hanan aptly describes his experience as a "nightmare" in CVS, p. 126. 59· "Cai Ruihong ren ru bao chou" ("Rainbow" Cai endures violation to execute vengeance), Constant Words, no. 36. See Hanan, CVS, p. 127. 6o. "Shi cheng An Wei su xing he ru" (How will history judge the behavior of An and Wei), in Gui Youguang, Zhen chuan xiansheng ji, p. 707. 61. Quoted in Dennerline, The Chia-ting Loyalists, p. 298. 62. "Yu Li Haoqing shu," in Gui Youguang, Zhen chuanxianshengji, p. I44· 63. Ko, Teachers of the Inner Chambers, p. 26 5. 64. See the "loyalty plays" discussed by Idema in The Dramatic Oeuvre of Chu You-tun, pp. I I I-7 5; and the entry "Jia zhou He E," in Huitu lienu zhuan, I2.39b-4oa. 6 5. This is a major thesis of Plaks's Four Masterworks. Chapter 9 I express my thanks to many friends, teachers, and colleagues who provided important source materials and information-among them Tse-tsung Chow, Benjamin Elman, Charlotte Furth, Zhengguo Kang, Wen-k'ai Kung, Judy Liu, Yu-kung Kao, F. W. Mote, Shi Zhicun, Ay-ling Wang, C. H. Wang, Ellen Widmer, and Ying-shih Yu. I especially appreciate the help of Bruce Lincoln, Susan Mann, Haun Saussy, and Pauline Yu, who offered valuable comments and suggestions on this paper. I. Nivison, Chang Hsueh-ch'eng, p. 264. 2. Mann, "'Fuxue' (Women's Learning) by Zhang Xuecheng," p. 53· 3· See Clara Lau, "Zhongguo chuantong cai." 4· Yao Pinwen, "Qing dai funii shige," pp. 54-55. 5. For information about this anthology, see my "Ming and Ch'ing Anthologies."

Notes to Pages 237-39 6. Zhang Xuecheng, Zhangshi yishu, in Zhang Xuecheng yishu (hereafter cited as ZXCY), p. 48. 7· Mann, '"Fuxuc' (Women's Learning) by Zhang Xuecheng," p. 55· 8. Waley, Yuan Mei, p. r67. 9· Yuan Mei once said: "The world of poetry is the hardest one to define. Some literati scholars read through ten thousand books until they had exhausted all their energies in old age, and yet they never understood the secret of poetry writing. But there were women who, though uncultivated and superficial in learning, were sometimes able to produce one or two lines that even Li Bai and Du Fu could not outdo. This is why I said the world of poetry is unlimited" (Suiyuan shihua, juan 3 ). Yuan Mei's view was apparently influenced by the Song dynasty critic Yan Yu (fl. I I 8o-I 2 3 5), who said that the poetic spirit "transcends that of the books" (fei guan shu ye). IO. See Yuan Mei, Xiao cang shan fang wenji, juan 3 2, in Suiyuan san shi ba zhang 4.8a. II. See my "Ming and Ch'ing Anthologies." I2. In "Fuxue," Zhang Xuecheng summed up his position on this issue: "Ignorant people have supposed that the poem 'Zhen wei' and others like it were written by the lovers themselves, and so they claim that the young women of ancient times simply opened their mouths and poured out complete verses that are superior to those of later male writers. Such people do not know that this belief is utterly without foundation" (trans. Susan Mann; see Mann, "'Fuxue' [Women's Learning] by Zhang Xuecheng," p. 52). I 3. For example, in a poem, Jin Yi praised her teacher Yuan Mei for "working hard to promote talents till his hair turned gray" (Yuan Mei, Suiyuan niidizi shi xuan, p. 28). Also, in his preface to this modern reprint of Suiyuan niidizi shi xuan, Zhu Weigong says that Yuan Mei "loved talents as though his life depended on them" (ai cai ruo ming). 14· In his foreword to Yuan Mei's Suiyuan niidizi shi xuan, Wang Gu (I 7 54I 8 21) praised Yuan Mei for including works of two maidservants and accepting them as his disciples. I5. See "Shi hua," in ZXCY, p. 44· The idea of "nei yan buchu kun wai" first appeared in the "Qu li" section (juan 2) of Liji zhengyi, p. I240. I 6. From Lii Kun, Shenyin yu, as cited in Clara Lau, "Zhongguo chuangtong cai," p. 99· 17. See ZXCY, pp. 44, 48. Trans. from Mann, "'Fuxue' (Women's Learning) by Zhang Xuecheng," p. 50. I8. Both ChenJiru (I5 58-1639) and Feng Menglong (1574-1646) recorded this saying in their writings. See Clara Lau, "Zhongguo chuangtong cai," p. 109. 19. See Mann, "'Fuxue' (Women's Learning) by Zhang Xuecheng," p. 49·

Zhong Huiling, "Qing dai nii shiren yanjiu," p. 224. Jones, Women in the Eighteenth Century, p. 140. 22. I am indebted to Nancy Armstrong's reference to this point; see Armstrong, Desire and Domestic Fiction, p. 5. 20.

21.

Notes to Pages 239-43 23. For more evidence to substantiate Qing women's extreme concern with their own morality, see Zhou Wanyao, "Qingtai Tongcheng xuezhe." 24. Clara Lau, "Zhongguo chuantong cai," p. I IO. 25. Ibid., pp. 95-96. 26. See "Epitaph on Lady Sun," in Lu You, Weinan wenji, juan 3 5. 27. Ibid. The Chinese original reads "cai zao fei niizi shi ye." 28. Wang Zhenyi, De fengtingchuji,juan IO, 9a; in]inlingcongshu, p. 2I I99· 29. The burning (or attempted burning) of women's poems can be traced back to the Tang (see Kang Zhengguo, Fengsao yu yanqing, p. 326). It did not, however, become a fashionable custom until the Qing. In the case of country women burning their songs written in the "women's script"-a unique written tradition supported by a long-standing female oral tradition-it might be linked to a popular belief that the poems could thus be carried to the underworld after the women's death (see Gong Zhebin, "Introduction," in Niishu, p. 3 3 ). 30. Liang Yizhen, Qing dai funii wenxue shi, p. 283. 3 I. Shi Shuyi, Qing dai guige shiren zheng liie, 22a (p. 83 ). 32· Liang Yizhen, Qing dai funii wenxue shi, pp. 22I-22. 33· Nivison, Chang Hsiieh-ch'eng, p. 274. 34· Qian Mu, Zhongguo sixiang tongsu jianghua, pp. 45-46. 35· See also D. C. Lau, trans., The Analects (New York: Penguin, I979), Bk. I4, p. I24· 36. As cited in Clara Lau, "Zhongguo chuantong cai," p. 98. 37· Dai Zhen, Mengzi ziyi shuzheng, p. 39· See also Ann-ping Chin and Mansfield Freeman, Tai Chen on Mencius, p. I43· 3 8. See Zhu Gui, juan 2 of his "Jin cheng wen'gao," in his Zhizu zhai ji (Xuehai tang ed.), 3a-3b. See also Clara Lau, "Zhongguo chuantong cai," p. I47· I am indebted to Ying-shih Yu (pers. comm, Aug. I6, I992) for identifying the authorship of the important "Essay on Talent and Morality" ("Cai de shuo"), an essay that as far as I know, is no longer extant. 39· Ban Zhao, "Nii jie," in Su Zhecong, Zhongguo lidai funii zuopin xuan, pp. 478-8I. 40. Trans. Carpenter, "Value and Women's Writing from the Han to the Song," p. I. 41. Of course, some modern people might feel that this kind of power offers only "an illusion of power based on sublimation and passive virtue" (see Jones, Women in the Eighteenth Century, p. I 5 ). But I am inclined to agree with Nancy Armstrong (Desire and Domestic Fiction, p. 26), who, in her discussion of female power, explains that such power "does not appear to be powerful" to some people, simply "because it behaves in specifically female ways." 42. See Cao Minggang, "Ban Zhao," in Chen Bangyan, Shi da cainii, p. 9· 43· Trans. adapted from Carpenter, "Value and Women's Writing from the Han to the Song," p. 2. 44· Hu Wenkai, Lidai funii zhuzuo kao, p. 22. 4 5. See juan 7 of Zhou li zhushu, in Shisan jing zhushu (Beijing: Zhonghua shuju, I98o), p. 687.

Notes to Pages 243-45 46. This recalls one of Charles Altieri's ("An Idea and Ideal of a Literary Canon," p. 58) criteria for canonization-that the work should contain great "ethical significance." 4 7. For example, the famous Six Dynasties critic Liu Xie said, "The superior man aims at establishing virtue; only after that does he think of immortalizing his words"; sec also Vincent Yu-chung Shih, trans., The Literary Mind and the Carving of Dragons (Hong Kong: Chinese University of Hong Kong Press, 1983 ), p. 187. 48. Trans. adapted from David R. Knechtges, Wen xuan, or Selections of Refined Literature (Princeton: Princeton University Press, 1987 ), 2: r 77. For the Chinese original, see Su Zhecong, Zhongguo lidai funu zuopin xuan, p. 41. 49· Ban Zhao, "Nii jie," in Su Zhecong, Zhongguo lidai funu zuopin xuan, p. 480.

so. Lin Wen-yiieh, "The Decline and Revival of Feng-ku," in Shuen-fu Lin and Stephen Owen, eds., The Vitality of the Lyric Voice (Princeton: Princeton University Press, 1986), p. r 3 6. 5 r. "On Literature," in Siu Kit Wang, Early Chinese Literary Criticism, pp. 25· 52. Trans. Carpenter, "Value and Women's Writing from the Han to the Song," p. 8. For the Chinese original, see "Lienii zhuan" in fin shu, as cited in Hu Wenkai, Lidai {unu zhuzuo kao, p. 10. 53· Xie Daoyun's collected works, Xie Daoyun ji (in 2 juan), are no longer

21,

extant. Only a few of her poems can be found in modern anthologies. See He Manzi, "Xie Daoyun," in Chen Bangyan, Shi da cainu, p. 39· 54· Zhong Rong, Shipin [zhu], pp. 69-70. See also my Six Dynasties Poetry, p. ro8. 55· Adapted from translation by Richard B. Mather, Shih-shuo Hsin-yu (Minneapolis: University of Minnesota Press, 1976), p. 3 55. See also Nanxiu Qian, "B.eing One's Self: Narrative Art and Taxonomy of Human Nature in the Shih-shuo hsin-yu" (Ph.D. diss., Yale University, 1994). 56. The Seven Sages of the Bamboo Grove were Ruan Ji,Ji Kang, Shan Tao, Xiang Xiu, Liu Ling, Ruan Xian, and Wang Rong. 57· Once Liu Liu, the magistrate, invited Xie Daoyun to have a session of "pure talk" (qingtan), after hearing so much about Xic's "air of the Bamboo Grove." See He Manzi, "Xie Daoyun," in Chen Bangyan, Shi da cainu, p. 51. 58. The Shi shuo xin yu records the following anecdote: "Wang Ningzi's wife, Lady Xie [Xie Daoyun], after going to live in the Wang family, felt a great contempt for Ningzi. On returning for a visit to the Xie household, her mood was most unhappy. Her uncle, Xie An, hoping to comfort and relieve her, said, 'Master Wang is, after alt, the son of Wang Xiihi, and as a person in his own right isn't at all bad. Why do you resent him so much?' She replied, 'Here, in this one distinguished household, for uncles there are Ada and Zhong Lang, and for cousins and brothers I also have Feng, Hu, Jie, and Mo. But who would ever have imagined that between heaven and earth there actually exists a person like Master Wang?'" (Trans. adapted from Mather, Shih-shuo Hsin-yu, p. 354 [see notes s].)

Notes to Pages 245-47 59· However, what has been preserved in the Quan Tang shi might only be a small portion of the actual corpus of writings by these women, because a large number of works were lost, mainly due to a lack of interest in publishing or preserving them. See my "Ming and Ch'ing Anthologies," pp. I-2. See also Robertson, "Voicing the Feminine," p. 64. 6o. Despite the term's religious connotations, the role of a "Daoist nun" in the Tang and Song was considered to be almost identical with that of the courtesan (see Wang Shunu, Zhongguo changji shi, pp. I58-65). This was true especially because some courtesans became Daoist nuns late in their career-Xue Tao was the most notable example. Later, some Southern Song scholars such as Ji Yougong and Sun Guangxian simply called Daoist nuns "prostitutes" (changji), because they disapproved of these women's supposedly "reckless" and "debauched" lifestyles (see Su Zhecong, Guiwei de tanshi, p. 28I). 61. Ibid., pp. 258, 282. Su Zhecong believed that Li Ye was summoned by an emperor other than Xuanzong, since Li Ye did not become famous until the later years of the Dali reign (767-79). 62. Trans. adapted from Kenneth Rexroth and Ling Chung, Women Poets of China (New York: New Directions, I982), p. I9. For the original poem, see Peng Dingqiu et al., Quan Tang shi, II: 9050. 63. See Su Zhecong, Guiwei de tanshi, p. 306. 64. Yu was executed at the age of 23 for beating a maidservant to death, and "in court circles there were many who spoke on her behalf" (see "The Poetess Yii Hsiian-chi," trans. Jeanne Kelly, in Y. W. Ma and JosephS. M. Lau, Traditional Chinese Stories, pp. 305-6). Huang Zhouxing (in his Tang shi kuai) laments that Heaven has mistreated Yu Xuanji (an ideal woman "who combined beauty and talent") in a way that was even more severe than its unfair treatment of many talented male scholars (see Kang Zhengguo, Fengsao yu yanqing, p. 3 20). 65. For example, the Tang anthologist Gao Zhongwu (in his Zhong xing jian qi ji) praised Li Ye for transcending the bounds of the "four virtues" of women (see Kang Zhengguo, Fengsao yu yanqing, p. 3 I8). 66. See Su Zhecong, Guiwei de tanshi, p. 307. See also Kang Zhengguo, Fengsao yu yanqing, p. 3 I?. 67. See Ji Yougong, Tang shi ji shi, as cited in Kang Zhengguo, Fengsao yu yanqing, p. 3 I 5. 68. Jeanne Larsen, "Introduction," to her Brocade River Poems: Selected Works of the Tang Dynasty Courtesan Xue Tao (Princeton: Princeton University Press, I987), p. xiii. 69. Pcng Dingqiu eta!., Quan Tang shi, II: 9033. 70. Trans. adapted from Mary Ellen Kivlen, "Women at the Loom: Weaving Revisions into Double Seventh Poetry" (unpublished research paper, I992), p. 74· For the original poem, see Zhu Shuzhen, Zhu Shuzhen ji, p. I54· For a discussion of Zhu's poem, see Kang Zhengguo, Fengsao yu yanqing, p. 328. 71. Anne Bradstreet, "The Prologue," in Rogers, Meridian Anthology of Early American Women Writers, p. I3. 72. The question of whether Zhu Shuzhen indeed had an extramarital affair has been constantly debated since Ming times-with Yang Shen (I488-I559)

Notes to Pages 247-5I and Ji Yun (I724-ISo5) representing the two opposing views. Some modern scholars have suspected that Zhu Shuzhen killed herself after her extramarital affair had been discovered by her husband or her relatives (see "Appendix," in Ji Qin, Zhu Shuzhen ji zhu, pp. 278-95). 73· Xing Lin, Songdai cainu zhuan, p. S6. 74· See my Late-Ming Poet Ch'en Tzu-lung, pp. ro-r6. 75· There have been conflicting views as to whether Zhang Hongqiao was indeed a courtesan, but many relevant source materials suggest that she was (see Zhou Zongsheng, Zhongguo cainu, p. 239). 76. Zhou Hui, "Qi Jingyun," in his Xu Jinling suo shi, juan 2; in]inling suo shi 157b-5Sa. 77· The most notable examples were Xue Tao, Yu Xuanji, Li Ye, Wang Funiang, Gao Wenji, and Duan Dongmei (See Su Zhecong, Guiwei de tanshi, pp. 26S-69, 303-4, 2S5, 246, 243-44). 7S. Ellen Widmer ("Poems Saved from Burning: Xiaoqing's Literary Legacy and the Place of the Woman Writer in Late Imperial China," Late Imperial China, I3.I Dune I992]: 120) observes that one distinguishing feature about the caizi jiaren heroines is not only that they have talent but "this talent leads them to a man." 79· For example, the idea of two or three matching couples can be found in stories such as Lu mudan (The green peonies) and Ping shan !eng yan (see Hessney, "Beautiful, Talented, and Brave," p. I67). So. Hessney, "Beyond Beauty and Talent," p. 239. Sr. See Xi Peilan's poem "The Renovation of Liu Shi's Tomb by Magistrate Chen Wenshu," in her Changzhen ge shiji (IS I 2: reprinted-Shanghai: Saoye shanfang, 1920), 6.ua. S2. Hu Wenkai, Lidai funu zhuzuo kao, p. SS. S3. Ibid. S4. See Zou Siyi, Shiyuan ba ming jia ji (preface dated I 6 55), preface to the section on Liu Shi, I a. S5. Zhong Xing, Ming yuan shi gui (ca. I62o), Ir.6b. S6. See Liu Shi, "Run Ji," in Qian Qianyi, Leichao shiji xiaozhuan, rev. ed. (Shanghai: Guji chubanshe, r9S3), 2: 760, 773· S7. Ibid., 76o, 774· SS. Ibid., 774· S9. Ibid., 76o. 90. See my Late-Ming Poet Ch'en Tzu-lung, pp. r6-r7. 91. See my "Ming and Ch'ing Anthologies." 92. Widmer, "Epistolary World of Female Talent." 93· Ibid., p. IO. 94· Handlin, "Lii K'un's New Audience," p. 2S. 95· ZXCY, pp. 44, 47· 96. Hu Wenkai, Lidai funu zhuzuo kao, p. 176. 97. Zou Siyi, foreword to his section on Wu Qi, in Shiyuan shi ming jia ji I a. 9S. Hu Wenkai, Lidai funu zhuzuo kao, p. ro6. 99· Chen Weisong, Furen ji, in Zhao dai congshu (n.p., rS33-44), vol. 74, 33b.

Notes to Pages 252-57 IOO. Yeh Shaoyuan, Wumeng tang quan ji, pp. 3-4. IOI. Sherr Quan, in his foreword to Zou Siyi, Shiyuan ba ming jia ji 2a. I02. See Hu Wenkai, Lidai funii zhuzuo kao, pp. 887, 889. I03. Trans. Haun Saussy, "Female Scribes, Ancient and Modern," in Kang-i Sun Chang and Haun Saussy, eds., Chinese Women Poets: An Anthology of Poetry and Criticism from Ancient Times to I9II (Stanford: Stanford University Press, forthcoming). I04. David Hawkes, trans., The Story of the Stone (New York: Penguin, I973), chap. 2, I: 77-78. I05. See Zhong Xing, Ming yuan shi gui 33.I6a; trans. Mary Ellen Kivlen, "Women at the Loom: Weaving Revisions into Double Seventh Poetry" (unpublished research paper, I992), p. 36. Io6. See my "Ming and Ch'ing Anthologies." I07. Clara Lau, "Zhongguo chuantong cai," pp. III-I2. Io8. Liang Yizhen, Qing dai funii wenxue shi, p. II. I09. Their husbands were Sun Yuanxiang, Xu Shanmin, and Chen Zhushi, respectively. For discussions of these ideal matches and others, see Kang Zhengguo, Fengsao yu yanqing, p. 34I; and Liang Yizhen, Qing dai funii wenxue shi, pp. 8I-84,75-77, I38,28o. I IO. See my "Liu Shih and Hsii Ts'an: Feminine or Feminist?" in Pauline Yu, Voices of the Song Lyric in China, p. I 87. I II. See David Hawkes, The Story of the Stone (New York: Penguin, I973), 2: 3 3 3. For a discussion of Xue Baochai as representing the most committed orthodox female Confucian in the novel, see Anthony C. Yu, "Literature and the Conflict of Desire in The Story of the Stone" (manuscript, I992), pp. 3 3-3 5. I 12. See Hu Wenkai, Lidai funii zhuzuo kao, pp. IOI2-I3, IOI?, I02I-22. I I3. Vivien Jones (Women in the Eighteenth Century, pp. 142-43) also talked about the same kind of "playing down of intellectual achievement" as a kind of strategy used by eighteenth-century English women writers. IJ4. See Liang Yizhen, Qing dai funii wenxue shi, p. I95· Yun Zhu, who might be a member of a Han Chinese Banner, was married into a Manchu family. She was the mother of the distinguished Manchu scholar Wanyan Linqing. I am indebted to Susan Mann and F. W. Mote for information about Yun Zhu's background. II5. See my "Ming and Ch'ing Anthologies." n6. For this claim, see Mann, '"Fuxue' (Women's Learning) by Zhang Xuecheng," p. I9. II?. Yao Pinwen, "Qing dai funii shige," p. 58. I r 8. For more Qing women poets who died in their teens, see Liang Yizhen, Qing dai funii wenxue shi, pp. 223, 295. II9. Yuan Mei, Suiyuan niidizi shi xuan, 2: 31. I 20. See Yuan Mei, "Jin Xianxian niishi muzhiming," in his Xiao cang shan fang wenji, juan 32, in Suiyuan san shi ba zhang 4.7b-8a. Translations of this essay are adapted from Mark Borer, "Epitaph for Jin Xianxian," in Kang-i Sun Chang and Haun Saussy, eds., Chinese Women Poets: An Anthology of Poetry and Criticism from Ancient Times to I9I I (Stanford: Stanford University Press, forthcoming).

Notes to Pages 257-6r I2L Ye Shaoyuan, "Xu qie wen," pp. I I-I?, in Wumeng tang quan ji. I 22. I was delighted to know that Dorothy Ko-whose fine essay on talent and virtue became available to me (thanks to Charlotte Furth) only after I had delivered this paper at Irvine, California in June I992-has come to this very same conclusion; see Dorothy Ko, "Pursuing Talent and Virtue." Chapter ro r. Hu Shi, "Wu Jingzi zhuan" (Biography of Wu Jingzi), in Li Hanqiu, Rulin waishi yanjiu ziliao, p. 66. 2. Lu Xun, A Brief History of Chinese Fiction, p. 2 74· 3. Xianzhai laoren (Old man of the Leisure Study) was the first to give this interpretation of the book: "The book takes this passage concerning career, fame, wealth, and rank as its theme .... Ultimately it regards those who reject these things as of the highest character, as standing out like pillars against the current" (see "Xianzhai laoren xu" in Li Hanqiu, Rulin waishi yanjiu ziliao, p. 99 ). This preface accompanied the earliest extant edition of Rulin waishi, the I8o3 Woxian caotang edition. The identity of Xianzhai laoren remains unknown. 4· Zhang Wenhu (Tianmu shan qiao), in his I885 "Rulin waishi ping" (Review of Rulin waishi), stated, "This book is particularly concerned with pointing out the errors of mingshi" (in Li Hanqiu, Rulin waishi yanjiu ziliao, p. I 37 ). He also specifically objects to Wang Mian's criticism of the examination system in the first chapter: "From ancient times pursuit of glory took precedence over real scholarship; this in not simply the fault of the bagu essay" (quoted in the preface to Li Hanqiu, Rutin waishi yanjiu ziliao, p. 9 ). 5. The complete passage reads: "The sacrifice to Tai Bo is the first great culmination of the book. In composing a large book, the author is like the builder of a palace and must first have a blueprint in mind: only after he carefully plans where he will place the central hall, the bedrooms, the study, the kitchen, and so forth can he begin building. In this book, the sacrifice to Tai Bo is the central hall" (see Li Hanqiu, Rutin waishi yanjiu ziliao, p. I I9 ). 6. Li Hanqiu, Rulin waishi yanjiu ziliao, p. I IO. 7· Jin He, "Jin He ba" (Postscript by Jin He), in Li Hanqiu, Rulin waishi yanjiu ziliao, p. 130. 8. Hu Shi, "Jianshe de wenxue geming lun" (On a constructive literary revolution), in Hu Shi wencun, I: 92. 9· Hu Shi, "Wushi nianlai Zhongguo zhi wenxue" (Chinese literature of the last fifty years), in Hu Shi wencun, 2; 182.. For a more complete discussion of Hu Shi's opinions on the novel, see Shuen-fu Lin, «Ritual and Narrative Structure in Ju-lin wai shih," esp. pp. 2.44-48. IO. Lu Xun, A Brief History of Chinese Fiction, p. 274· I I. Paul S. Ropp's Dissent in Early Modern China is the most substantial examination of the novel's critical social message in English. 12. Hsia, The Classic Chinese Novel, p. 209. 13. Timothy C. Wong, Wu Ching-tzu, p. 90. 14· Shuen-fu Lin, "Ritual and Narrative Structure in Ju-lin wai-shih," esp. pp. 256-59·

420

Notes to Pages 26r-63

I 5. Treating chaps. I and 55 as an independent preface and postface, Huang Binze ("Lun Rulin waishi de changpian yishu jiegou," in Anhui sheng jinian Wu Jingzi dansheng erbai bashi zhounian weiyuanhui, Rulin waishi yanjiu lunwenji) divides the intermediate chapters into four sections: chaps. 2 through 17, which concern aspiring scholars; 17 to 30, which treat mingshi; 3 I to 43, which concern self-styled "heroes" who make unrealistic attempts to reform society; and 44 to 54, which portray "wicked and vulgar" characters whose depravity demonstrates the incurable decline of the feudal social order. Although the fictional world described in Rutin waishi certainly includes characters who fit each category Huang identifies, their appearance is by no means limited to the chapters he cites. Huang's categories may be compared with Chen Meilin's (Wu Jingzi yanjiu, pp. 2 34-3 8) method of dividing the novel. Chen distinguishes three sections: chaps. 2 through 30, which demonstrate the corrupting influence of the examination system on a wide range of scholars; chaps. 3 I through 46, which treat a variety of idealists who eschew the examination system and make independent efforts to reform society; and chaps. 4 7 through 55, which dramatize the failure of the idealists in Part II to have any long-lasting effect on their society. Although more credible than Huang's partitioning of the novel, Chen's division similarly obscures the climactic nature of the sacrifices to Tai Bo in chap. 37· C. T. Hsia's tripartite division of the interior chapters of the novel (2-30, 3I-37, 38-54), a system adopted also by Shuen-fu Lin and discussed later in this essay, remains the most useful. I6. Wells, "An Essay on thefu-lin wai-shih," p. I43· I7. Shuen-fu Lin ("Ritual and Narrative Structure in Ju-lin wai shih," p. 262), for example, is forced to interpret the latter third of the book as a demonstration of the absence of "ritual," the very element that he elsewhere defines as the novel's integrative structural principle: "The apparent diffuseness of Part III is not at all a decline of the author's creative energy. The narrative diffuseness is, on the contrary, designed to indicate the total failure of the ideal vision of a perfectly ritualized world so cherished by the central characters of the book. We might consider, therefore, the entire third part as one vast unit corresponding, if only in an antithetical way, to the sacred ceremony at T'ai-po Temple." I8. Wu Jingzi, Rulin waishi (Taibei ed.; hereafter RLWS), p. 9· In quoting from the novel, I have generally followed the translation by Yang Hsien-yi and Gladys Yang (The Scholars, I957), but the transliteration of names has been altered to bring them in line with standard pinyin. I9. RLWS, p. 54· 20. See also Wang Hui's discussion of dreams in chap. 2, where he eagerly accepts their prophetic nature when they flatter his ambitions but deems them "unreliable" when they favor others. 2 I. Wu Jingzi restricted his use of prophetic dreams and the supernatural to three sections of the novel: the opening chapters, the climactic central chapters immediately preceding and following the sacrifices at the Tai Bo Temple, and the concluding chapters. As such, he clearly intended its interjection to serve the structural purpose of highlighting critical stages in the narrative. Although most of the supernatural episodes are satiric or parodic in nature, there are exceptions,

Notes to Pages 263-265

421

the most significant of which are Zhuang Shaoguang's discovery of a scorpion during his audience with the emperor (to be discussed below) and Zhuang's later encounter with two "walking corpses." The courtesan Pinniang's dream in chap. 53 would appear to be another exception, in that it is thematically significant and correctly foretells her eventual retreat to a nunnery, but the conjunction of her story with the decidedly satirical portraits of the venal abbess who "exorcises" her evil spirits and the two squabbling fortune-tellers who mistakenly predict her marriage to a high official cannot but serve to deflate our interest in the dream as prognostication; it is primarily introduced, I would suggest, for the image it presents of the psychological devastation wrought by earthly ambition. Considering all the relevant episodes together, one senses that Wu Jingzi would have agreed with the Lou brothers' observation about the Wang Hui incident: "Coming events cast their shadows before, and a fortune-teller is sensitive to these almost imperceptible signs. It is nonsense to talk of ghosts and fairies" (RLWS, p. 74). 22. It was, I believe, Roman Jakobson who first isolated metonymy and metaphor as the two basic forces organizing the operation of language. See, e.g., his "Two Aspects of Language and Two Types of Aphasic Disturbances" and "Randbemerkungen zur Prosa der Dichters Pasternak." My discussion here also owes a great deal to Peter Brooks, in particular the following passage. from Reading for the Plot, (p. 91): "Narrative operates as metaphor \nits affirmation of resemblance, in that it brings into relation different actions, combines them through perceived similarities ... appropriates them to a common plot, which implies the rejection of merely contingent (or unassimilable) incident or actions. Plot is the structure of action in closed and legible whole; it thus must use metaphor as the trope of its achieved interrelations, and it must be metaphoric insofar as it is totalizing. Yet it is equally apparent that the key figures of narrative must in some sense be not metaphor but metonymy: the figure of contiguity and combination, of the syntagmatic relation. The description of narrative needs metonymy as the figure of linkage in the signifying change: precedence and consequence, the movement from one detail to another, the movement toward totalization under the mandate of desire." 23. See de Man, Allegories of Reading, p. 14: "The preference [for figural or metaphorical language] is expressed by means of a distinction that corresponds to the difference between metaphor and metonymy, necessity and chance being a legitimate way to distinguish between analogy and continuity. The inference of identity and totality that is constitutive of metaphor is lacking in the purely relational metonymic contact: an element of truth is involved in taking Achilles for a lion but none in taking Mr. Ford for a motor car." 24. See Brooks, Reading for the Plot, p. 91. 25. See Kernan, The Plot of Satire, esp. pp. 94-104. 26. Cf. Hsia, The Classic Chinese Novel, p. 229: "It is Wu Ching-tzu's regular practice to juxtapose stories whose heroes offer obvious points for comparison and contrast." 27. See Shuen-fu Lin, "Ritual and Narrative in]u-lin wai shih," pp. 260-63. 28. For example, the grandson of Fan Jin's secretary, Qu Gongsun, makes his

422

Notes to Pages z65-7I

reputation by publishing a lost volume of poems by the poet Gao Qi under his own name; his own later poems are described as "full of words and phrases pillaged from earlier poets, a line here resembling Qu Yuan's Lisao and a line there reminiscent of the early philosophers" (RLWS, p. 8o). Later in the novel, Kuang Chaoren discovers how easy it is to compose poetry to the standards of the group of mingshi with whom he associates {he has only to prepare himself by briefly studying a versification primer) and notices that his acquaintances' poems are "full of terms used in the bagu essays and phrases cribbed from old commentaries" {p. 136). 29. RLWS, p. 129. 30. RLWS,p.II4. 31. RLWS, p. 145· 32· As C. T. Hsia (The Classic Chinese Novel, p. 23 r) points out, this is also true in the case of Kuang Chaoren. But the primary interest in Kuang's story is his moral decline {that is, the change that occurs in his personal behavior over the course of the several chapters that take him as protagonist). Niu Pulang, on the other hand, is a "flat" character whose character seems fixed in early childhood; he can serve only as a foil to expose the behavior of the wide range of social types he encounters as his story develops. 33· RLWS, p. 250. 34· When Bao Wenqing's friend Prefect Xiang recruits him and his son to supervise an examination, the Baos are scrupulous in policing the many cheating examinees (RLWS, pp. 189-90). Their conscientiousness compares favorably with the lax invigilation practices of the Master of Ceremonies at the sacrifice to Tai Bo, Dr. Yu, in chap. 3 7 (p. 279 ). 35· RLWS, p. 250. 36. Although extrinsic to the text of the novel itself, the fact that Tai Bo could be construed as the founder of Wu Jingzi's family line is also of relevance here. For discussion of autobiographical references in the novel, see Hu Shi's seminal biographical essay on Wu Jingzi, as well as the following works: He Zehan, Rulin waishi renwu benshi kaoliie; Chen Ruheng, Wu ]ingzi zhuan; and the biographical articles in Chen Meilin, Wu Jingzi yanjiu. 37· H. C. Chang, Chinese Literature, p. 21. 38. Hsia, The Classic Chinese Novel, p. 237. 39· RLWS, p. 276. 40. This is pointed out by Wei-ming Tu in "Lias Process of Humanization," Philosophy East and West 22.2: 190. 41. Xunzi, "Li lun pian" (A discussion of rites). I have followed the translation by Burton Watson in Hsun Tzu: Basic Writings, p. 89. 42. Ibid., p. ro9. 43· References by Wu Jingzi's contemporaries indicate that the manuscript of Rulin waishi circulating in the author's day contained 50 rather than 55 chapters. Judging by their inferior stylistic quality and generic dissimilarity to the rest of the novel, the chapters most likely to be interpolations would seem to be two or three immediately following the sacrifices to Tai Bo (3 8-39) and those toward the end of novel that deal with the martial arts expert Feng Si (50-52). There

Notes to Pages 27I-8I also exists a clearly apocryphal "chap. 56," whose authenticity was questioned as early as r869 by the editor Jin He. For a discussion of this chapter, see Chen Xin and Du Weimei, "Rulin waishi di wushiliu hui zhen wei bian," in Rulin waishi yanjiu lumuen ji, pp. r 53-64. 44· RLWS, p. 278. 45. In citing Mencius, I have used the translation of James Legge, with slight variations. See The Works of Mencius (reprinted-New York: Dover Publications, 1970), pp. 177-79. Zang Cang objects to Mencius's use of more elaborate ritual implements on the occasion of his second mourning than he had used on the first. Mencius defends himself by saying this was simply a reflection of his promotion from scholar to official during the period between the two rites and does not represent an "exceeding" of ritual obligations. 46. Zhuang's reading of the scorpion's message is bolstered by the answer he receives the following day when consulting the hexagrams ("A hermit in the mountains"), which seems to be recommending a life of reclusion; see RLWS, p. 261. 47· RLWS, p. 26r. 48. As has been frequently pointed out, each of the four is skilled at one of the arts traditionally associated with the Confucian gentleman of leisure. 49· RLWS, p. 3· so. RLWS, p. 7· Chapter

II

r. For an account of such discussion from the late Qianlong era through the r 8 3 os, see my "From Writing to Literature." 2. Cao Xueqin, Hong lou meng, pp. 1-3. The English translation is that of David Hawkes, The Story of the Stone, pp. 47-5 r. Note that the text elides any account of the actual act of inscription itself. 3. For a description of the history of wenxue in the Song dynasty, see Bol, "This Culture of Ours." See also Fuller, The Road to East Slope. 4· For some examples, see Ropp, Dissent in Early Modem China, pp. 93roo, r 14-r 6. For an accounting of Wu Jingzi's own critiques of the examination system, see ibid., pp. ror-13. 5. Among other targets in this section, Wu Jingzi sharply mocks literati hatred of the Yongle emperor of the Ming, a focus of shi resentment from ca. I 5oo on and the subject of Benjamin Elman's chapter in this volume. See Wu Jingzi, Rutin waishi (all references in this chapter are to the 1969 Hong Kong ed.), pp. 87, 93; a translation can be found in The Scholars, pp. 143, 147-48. 6. The manuscript was among the effects of Wang Hui, a treacherous and corrupt official who had succeeded Qu's grandfather in office some years before the transaction. Wang gave his few books to young Qu as he was fleeing from authority. See Wu Jingzi, Rulin waishi, p. 86; Scholars, pp. 140-41. 7. For a reading of the function of ritual {li) as depicted in the novel that bears on the difficulty of creating an uncontaminated rhetoric, sec Marston Anderson's chapter in this volume, esp. pp. 267-7 5. Anderson builds on the

Notes to Pages zBr-83 earlier work of Shuen-fu Lin, "Ritual and Narrative Structure in Ju-lin waishih," pp. 244-65. 8. A full description and analysis of the post-r895 intellectual scene has yet to be written. Hao Chang's Chinese Intellectuals in Crisis represents a major effort to cover a field that is in need of more complex theorizing. 9· Thomas Metzger (Escape from Predicament) makes the case for a utopian sense of possibilities in the late Qing. His decided inattention to any of the anxieties brought about by imperialism, however, leaves his thesis one-sidedly optimistic, thereby rendering it easier to blame political figures who came later for not taking advantage of what Metzger sees in the last analysis as a fairly simple opportunity for intellectual liberation. ro. Wu Wenqi, ]in bai nian !aide Zhongguo wenyi sichao, pp. 28-3 r; Qian Zhongshu's withering account of the vulgarity of the result is briefly outlined in Tan yi lu, p. 24. See also Noriko Kamachi, Reform in China, pp. r78-8r. r r. Acknowledgment of the sophisticated intellectual stance of the late imperial fictional canon was pioneered by Patrick Hanan in this country, and he has been followed by Andrew Plaks and David Roy. For the most thorough account of this genre's concern with ideas, see Plaks, Four Masterworks. Lu Xun's hostile remarks about the crudeness of the late Qing novel can be taken as recognition within the Chinese intellectual community of the seriousness of the novels that had come before. The post-1917 commentary on the traditional novel as part of popular culture must therefore be taken as a particular strategy in a broader discourse attempting to rewrite literary history as the manifestation of a longlived "two-line struggle" between elite and popular culture. It need hardly be added that this mode of analysis fit perfectly into the Marxist-inspired scholarly inquiry that gained strength in China from ca. 1930 on. 12. This theme appears in what is now regarded as the first of the post-I 89 5 accounts of the advantages of the novel, Yan Fu and Xia Zengyou's "Guowen baoguan fayin shuobu yuanqi" (Announcing our policy to print a supplementary fiction section), reprinted in Guo Shaoyu and Wang Wensheng, Zhongguo lidai wenlun xuan, 4: 196-205. For a discussion of this important text, see Hsia, "Yen Fu and Liang Ch'i-ch'ao," pp. 227-31. 13. Hsia, "Yen Fu and Liang Ch'i-ch'ao," pp. 23 r-32. Liang Qichao, "Yiyin zhengzhi xiaoshuo xu," in Guo Shaoyu and Wang Wensheng, Zhongguo lidai wenlun xuan, 4: 205-6. 14· The extensive influence of this article was first affirmed by Ah Ying [Qian Xingcun] in his 1937 Wan Qing xiaoshuo shi (History of late Qing fiction), p. 2. I 5. In other words, the category has been virtually devoid of worthwhile works since its inception. Liang's identification of the long-lost Yu chu as the original text in the genre is presumably based on Yan Shigu's (58 r-645) commentary on the Han shu, which cites Zhang Heng's Western Metropolises Rhapsody (Xijing fu) to the effect that "the nine hundred xiaoshuo begin from Yu chu." See Han shu, p. 17 4 5. Zhang Heng's line can be found in "Xijing fu," in Zengbu liu chen zhu wenxuan (Wenxuan with six commentaries and supplement) (Taibei: Huazheng shuju, 1974), 2.23b.

Notes to Pages 283-86 r6. The quotation from Kang Youwei is from his 1897 "Riben shumu zhi zhiyu" (A cognoscenti's notes on A Record of a Bibliography of [books in] Japan). See Chen Pingyuan, Ershi shiji Zhongguo xiaoshuo shi: diyi juan, pp. 4-5. 17. See Hsia, "Yen Fu and Liang Ch'i-ch'ao," p. 232, for translation and discussion of the obvious exaggeration that marks his discussion of the novel's power in the West. r8. On Qu Qiubai's determination to see nothing useful in the literary languages of his day and his utopian calls for completely new forms, see Huters, "The Difficult Guest: May Fourth Revisits," CLEAR 6 (1984): 135-49. 19. Hsia poses this question in "Yen Fu and Liang Ch'i-ch'ao," pp. 230-3 r. 20. Di Baoxian, "Lun wenxueshang xiaoshuo zhi weizhi" (On the position of fiction in literature), X in xiaoshuo (New fiction) 7 (1903), in Guo Shaoyu and Wang Wensheng, Zhongguo lidai wenlun xuan, 4: 237. Lu Xun is intent in his "Moluo shili shuo" (On the power of Mara poetry) in drawing attention to the dangers of the utilitarianism he saw dominating Chinese intellectual life in the first decade of the century. It is significant that Lu Xun regarded himself as an isolated voice. 21. Cai Jingkang, "Wan Qing xiaoshuo lilun chulun," p. 415. For a specific instance of this reading of the traditional novel, see Wang Zhongqi, "Zhongguo lidai xiaoshuo shilun" (On the history of the Chinese novel through the ages), reprinted in Guo Shaoyu and Wang Wensheng, Zhongguo lidai wenlun xuan, 4: 259-60. For Yan Fu's thesis on this, see Hsia, "Yen Fu and Liang Ch'i-ch'ao," p. 230. The most famous example of such a reading of a Qing novel is of course the "suoyin" school of The Story of the Stone, in which the family relationships of the novel are taken as a network of political allegory. 22. See my "A New Way of Writing," for a sample of some of these pronouncements. 23. Leo Ou-fan Lee and Andrew Nathan, "The Beginnings of Mass Culture," pp. 387-88. 24. For a full account of the novel's publication history, see Lu Shudu, "Qianyan" (Preface), in Wu Jianren, Wo foshan ren wenji (The writings of Wo foshan ren [i.e., Wu Jianren]) (Guangzhou: Huacheng chubanshe, 1988), r: 2-3. 2 5. In his study of Strange Events, Chen Xinghui ( "Ershi nian mudu zhi guai xianzhuang" yanjiu, pp. 54-55) calculates that, in fact, only fifteen years elapse during the course of the novel. But Chen also allows that Wu seems to have every intention of giving a comprehensive account of the twenty years in question. 26. Wu Jianren, Ershi nian mudu zhi guai xianzhuang, p. 5· This edition, while published in first-list simplified characters, is both well annotated and easily accessible. An abridged English translation of approximately one-third of the ro8-chapter text by Liu Shih Shun is available as Vignettes from the Late Ch'ing. The translation focuses on the early chapters and omits most of the commentary by the characters on the import of the events they are witnessing. It also leaves out many of the personal episodes of the narrator and most of his evaluations. The discussion that follows will cite the chapter numbers and pagination of the Chinese text, with notes referring to the corresponding pages in the

Notes to Pages 286-92 English version. In the event that the English version omits the passage discussed, no note will appear. 27. Lu Xun, Zhongguo xiaoshuo shi/Ue, 9: 282. A different translation can be found in Lu Hsun, A Brief History of Chinese Fiction, p. 3 72. 28. Lu Xun, Shiliie, p. 286; idem, History, p. 3 79· 29. See the curious passage in" Nahan xu" where Lu Xun claims that during his years in Tokyo he was the only Chinese of his generation to be interested in literature (Lu Xun quanji I: 4I7; translation in The Complete Stories of Lu Xun [Bloomington: University of Indiana Press, I98I], p. vii). Only if we read this comment within the context of his manifest discontent with the direction chosen by Chinese intellectuals in the period immediately following I 89 5 does this make sense. This feeling is most clearly expressed in his contemporary essays written in Japan in I907-o8. See "Wenhua pian zhi lun" (On the extremes of culture), in Lu Xun quanji, I: 44-57, esp. p. 56). 30. See, e.g., Ah Ying, Wan Qing xiaoshuo shi, p. I6, where the critic's plain attempt to praise the novel cannot overcome Lu Xun's negative judgments, which are cited at the beginning of Ah Ying's own evaluation of the book. It is worthy of note that when Hu Shi talked of traditional Chinese fiction in his I 9 I 7 article (i.e., written at least three years before Lu Xun's negative evaluations appeared) "Wenxue gailiang chuyi" (Suggestions for literary reform; in Zhao Jiabi, Zhangguo xin wenxue daxi [Shanghai: Liangyou tushu gongsi, I93 5], I: 70), he included Wu Jianren on a short list of three consisting of Wu, Shi Naian (reputed author of Shuihu zhuan), and Cao Xueqin. 3 I. This information is contained in Wu's contribution to Xiaoshuo conghua (Collected words on the novel), which Liang Qichao put together in I903 and published in the first and second issues of his new fiction journal, X in xiaoshuo (New fiction). Xiaoshuo conghua can be found in Ah Ying, Wan Qing wenxue congchao, I: 347· Wu wrote at the same time that had read some 300 Chinese novels, including a hundred of "recent composition and newly translated [from Western works]." 3 2. Liu Shih Shun summarizes these events on p. r of Vignettes. 3 3. The English version of this episode can be found in ibid., pp. I 3-14, 23, 26. Note that here as elsewhere the revelations about Gou are delayed by the admixture of other plot elements, perhaps a simulacrum of the long period required for Jiusi to learn the true face of things. 34· Compare the almost simultaneous fulminations of the revolutionary party against Manchu rule. See, e.g., the description of Zhang Binglin's racial views in Rankin, Early Chinese Revolutionaries, pp. 54-56. 35· Vignettes, pp. 7I-85. 36. Ibid., pp. I47-64. 37· This episode is relatively well represented in ibid., pp. 149-64. Characteristically, however, the personal information about Cai Liisheng is completely omitted. 38. Ibid.,p. I79· 39· Some of the more conspicuous examples include the tale of the owner of the steamship company who tries to take a concubine in Hankou, only to have

Notes to Pages 292-304 his wife in Shanghai get wind of it and commandeer a boat to go upriver to break up the prospective union told in chapters 5I-52 (ibid., pp. 223-35) and the intricate model steamship employed in chapters 29-30 (ibid., pp. I 3 I-37) to point out the corruption in governmental manufacturing enterprises.

40. Vignettes sums all this up in a few paragraphs on pp. 286-87. 41. For the function of these two characters in the text, see Milena DolezelovaVelingerova, The Chinese Novel, pp. 42-45. 42. Vignettes, pp. I9I-95· 43· Ibid., pp. 349-7I, 379-92 contains a quite detailed account of these episodes. 44· See, e.g., Mencius IIIB.9: "When the world declined and the Way fell into obscurity, heresies and violence again rose. There were instances of regicides and parricides. Confucius was apprehensive and composed the Spring and Autumn Annals" (D. C. Lau trans., p. I 14). 4 5. There is interesting parallel here with Wu's most famous short novel, Hen hai (Sea of woe), first published in 1906. Although the novella is the story of the disruption of the prospective marriages of two young couples by the Boxer Rebellion, the proximate cause of the breakups is the sudden disappearance of modern means of communication, in this case the railroad between Beijing and Tianjin. One couple then must rely on carts and canal boats, and they become hopelessly separated from each other. 46. These events are summarized in Vignettes, pp. 405-8. 4 7. Milena Dolezelova-Veiingerova offers a different evaluation of the import of the novel as a whole: "The meditations offered in the non-action episodes suggest, however, some hope: the crisis of values is temporary. Chinese society requires a reexamination of its traditional values. Perhaps, when properly interpreted and adapted to the crisis of the times, these values, enriched by practical Western learning, could survive the crisis and provide a new ethical basis for China" (The Chinese Novel, p. 49). 48. This is a pattern that was to appear frequently in Chinese literature produced after May Fourth. Perhaps the most famous example is the main character of Qian Zhongshu's exceptional Wei cheng (Fortress besieged), Fang Hongjian, for whom knowledge of his own situation always comes too late. See Huters, Qian Zhongshu, chap. 6. Chapter

12

The abbreviation TZ stands for tongzhi (provincialgazetteer) and WXTK for Qingchao wenxian tongkao. I. For example, Zhu Xi's community granaries (shecang) to replace Wang Anshi's green sprout loans (qingmiao fa). 2. This kind of tripartite division appears in many guises in the social science literature, including Amitai Etzioni's taxonomy of three kinds of power in complex organizations (coercive, remunerative, and normative) and Kenneth Boulding's three P's of coordination in human society (prices, policemen, and preachments); see Etzioni, A Comparative Analysis of Complex Organizations; and Boulding,

Notes to Pages 304-313 Ecodynamics. For an important application in the Chinese studies field, see Skinner and Winckler, "Compliance Succession in Rural Communist China." 3. See Brook, "The Spatial Structure of Ming Local Administration." 4· Fei Hsiao-tung, China's Gentry. 5· Skinner, The City in Late Imperial China, pp. 19-20. 6. For examples of excellent American work, see Wakeman and Grant, Conflict and Control in Late Imperial China. 7. Shigeta Atsushi, "The Origins and Structure of Gentry Rule." 8. Mori Masao, "Nihon no Min-Shin jidaishi kenkyii." 9· Rowe, Hankow: Commerce and Society; and idem, Hankow: Conflict and Community. ro. Rankin, Elite Activism and Political Transformation in China. II. Rowe, "The Public Sphere in Modern China"; idem, "The Problem of 'Civil Society' in Late Imperial China"; Rankin, "The Origins of a Chinese Public Sphere"; and idem, "Some Observations on a Chinese Public Sphere." 12. WXTK 32.5143. 13. I analyze these developments in some detail in Will and Wong, Nourish the People, pp. 25-74. 14· WXTK 35·5I8o-sr8r. 15. Will and Wong, Nourish the People, p. 65. r6. Ibid., pp. 66-67, 406. 17. Ibid., p. 66. r8. Ibid., p. 70. 19. ]ifu yicang tu 1.9a-r5b; Will and Wong, Nourish the People, p. 71. 20. State granary disbursals for famine relief often were figured at 0.3 shi/ month for an adult and half that amount for a child. Thus a 2oo,ooo shi supply would feed 20o,ooo adults for more than three months at this rate. 21. Will and Wong, Nourish the People, p. 3 I 5. 22. Ibid., pp. 71-72. 23. Ibid., p. 306. 24. Hoshi Ayao, Chukogu shakai fukushi seisatsu shi no kenkyu, p. 244. 25. Will and Wong, Nourish the People, p. 306. 26. Matsuda Yoshiro, "Shindai koki Koto Koshufu no soko to zento." 27. Will and Wong, Nourish the People, p. 75-92. 28. Igarashi Shoichi, Chugoku kinsei kyoikushi no kenkyu, p 296. 29. Ibid.; Ogawa Yoshiko, "Shindai ni okeru gigaku setsuritsu no kiban." 30. For the lower Yangzi, Angela Leung ("Elementary Education in the Lower Yangtze Region") has demonstrated that the nomenclature "community schools" was more commonly employed in the Ming and "charity schools" in the Qing. She also finds for the lower Yangzi cases a clear shift in the relative importance of official and elite funding and management from Ming reliance on officials to Qing reliance on elites. But for other parts of the empire, such as Guangdong, the community school nomenclature was retained in the Qing dynasty; see the Guangdong TZ (1822) 144·1-9. 31. Qinding Da Qing huidian shili 396.ra-9b.

Notes to Pages 3r3-r8 32. William Rowe, "Education and Empire in Southwest China." 33· Hunan TZ (1885) 68.36a-37a. 34· Sichuan TZ (1815) 34·79: 2-46. 35· Shaanxi TZ (1934) 37.1a-26a; 38.ra-66a. 36. Guangdong TZ (1822) 144.1a-9a. 37· Hubei TZ (1921) 59.53a-66b. 3 8. In the explanation of contents (fanli) in his magistrate's handbook of r 694, Huang Liuhong (A Complete Book Concerning Happiness and Benevolence, pp. 6r-62) shares with the reader his effort to gather materials on education and economic welfare for the topic of "instruction and nourishment" (jiaoyang); granaries are discussed in the next section on famine relief (huangzheng), since food supply management for crises was a large enough subject to be distinguished from other policies for economic welfare. 39· Yamana Hirofumi, "Shinmatsu Kanan no giso ni tsuite." 40. Had the government made much effort to promote the use of lineages for granaries, Zhu Yang would probably have discussed the phenomenon in his major study of lineages and the Qing state, Qingdai zongzufa yanjiu. I am grateful to Harriet Zurndorfer for pointing out Zhu's work to me. 41. Chen Hongmou, Peiyuantang oucungao 14:3 sa. 42. I base this conclusion upon scanning Chen Hongmou's official writings from these posts collected in his Peiyuantang oucungao. 43· Ogawa Yoshiko, "Shindai ni okeru gigaku setsuritsu no kiban"; Taga Akigoro, "Kindai Chiigoku ni okeru zokujuku no seikaku." 44· Woodside, "State, Scholars, and Orthodoxy," p. 182. 45· My stress on a continuum between official and non-official instead of a sharp contrast between the two does not mean that the roles of officials and elites could not be on occasion contested; from Woodside's research on education (ibid.), for instance, the "studies official" was subject to some disagreement. But disagreements did not lead, it seems to me, to the fundamental distinctions drawn between state and society in many parts of Europe. 46. For instance, see Jerry Dennerline's excellent essay on fiscal reform in early Qing Jiangnan, "Fiscal Reform and Local Control"; and James Polachek's important essay on the gentry in Suzhou during the Tongzhi restoration, "Gentry Hegemony." 47· The integration of elite efforts by the state does not mean that elites, for their part, did not resent official intrusions. This possible friction notwithstanding, there was a functional substitutability of many official and elite efforts to promote social order. 48. Wada Sei, Shina chiho jiji hattatsu shi, pp. 51-52. 49· Ibid., pp. 1 19-26; Matsumoto Yoshimi, Chiigoku sonraku seido no shi teki kenkyii, pp. 131-38; Wada Sei, Shina chiho jiji hattatsu shi, pp. 119-26; Shimizu Morimitsu, Chiigoku kyoson shakai ron, pp. 3 3 8-6o; Handlin, Action in Late Ming Thought, pp 47-s1, 198-99. so. Freedman, Chinese Lineage and Society, p. 87. sr. Hsiao Kung-chuan, Rural China, pp. 184-205.

430

Notes to Pages 318-z4

52· Mair, "Language and Ideology." 53. Another indica tor of political concern for official control over local institutions comes from Chen Hongmou's efforts in Jiangxi province to utilize lineages more formally in local administration. Chen (Peiyuantang oucungao 14.35a) argued that a leadership position should be created in every lineage to be filled by an individual whose responsibility it would be to report to local officials on local order. 54· For examples from famine relief, see Mori Masao, "Juroku-juhachi seiki ni okeru kosei." 55. Renzong shilu 50.24b-26a. 56. Examples of nineteenth-century officials continuing to play a role in local granaries beyond what the Jiaqing emperor anticipated can be found in Hoshi Ayao, Chugoku shakai shifuku seisatsu shi no kenkyu, pp. 297-98. 57· Lest the diminished coordination and monitoring of local government activities be taken as a more general indicator of state weakness, consider two features of state activity. First, the central government was able to expand fiscal extraction dramatically in the late nineteenth century. Revenues leaped from a range of 30-40 million taels in the eighteenth century to an average of 83.5 million for the decade r885-94 (Wei Guangqi, "Qingda houqi zhongyang jiquan," p. 227; Hamashita Takeshi, Chugoku kindai keizaishi kenkyu, p. 79). Second, the government continued to mobilize armies to put down uprisings. Following military victories, officials such as Zeng Guofan and Zhang Zhidong mounted efforts to re-establish social order in the northwest and the southwest. 58. The term "fractal" was invented by the mathematician Benoit Mandelbrot to refer to the replication of certain irregular geometric patterns on different spatial scales with the degree of irregularity remaining constant. Thus, fractals look the same whether so tiny as to be viewed under a microscope or so large as to be visible only from an airplane. See Gleick, Chaos. 59· These issues are addressed more fully in my "Naguere et aujourd'hui." 6o. I look more closely at the public sphere literature in "Great Expectations." 6r. See Tilly, The Formation of National States; and Bendix, Kings or People, for two very different views of state building and modern politics; where Tilly and the contributors to that volume highlight fiscal extraction, war-making, and bureaucracies, Bendix addresses the shift from authority vested in monarchs to representative forms of authority. Some studies on modern European state formation consider non-Western cases to be deformed; see Badie and Birnbaum, The Sociology of the State. On the other hand, a book on the development of the modern state by Poggi, The Development of the Modern State, simply limits itself to European history-what is troubling here is that political sociologists easily mistake this treatment of European state formation for a study of modern states generally; nothing in Poggi's book title would alert the reader to expect a book on European history. 62. See Tilly, Coercion, Capital and European States, for a long-run view of European national state-making joined to an analysis of the development of capitalism. 63. Kuhn, "Local Self-government Under the Republic."

Notes to Pages 324-329

431

64. Duara, Culture, Power and the State; Min Tu-ki, National Polity and Local Power; R. Thompson, "Statecraft and Self-government." 65. For the association between wuguanfang and buzhengtong, see Link et al., Unofficial China, p. 2.. Chapter 13 I. Sun Yat-sen, "Fu Duan Qirui chanshu shanhou huiyi zhuzhang dian" (Telegraphic reply to Duan Qirui's proposal for a reconstruction conference), in Sun Zhongshan xuanji [hereafter SZSXJ], p. 9 9 3; originally published in Yishibao (Social welfare) Jan. :z.o, I92.5, p. :z.. 2. Sun Wen, Zongli quanji [hereafter ZLQJ], I: IOI. For a full translation into English, see Sun Yat-sen, San Min Chu I. 3· Zou Lu, Zhongguo Guomindang shigao, pp. 6os-6. 4· ZLQ], r: 165-66. In discussing the historical character A Dou (z.o7-7I), Sun specifically referred to the fourteenth-century novel San guo yanyi (Romance of the Three Kingdoms) which "practically everyone" is familiar with. 5. Sun's version of the distinction between those who labor with their minds and those who labor with their hands divides people into those who "understand and are aware right from the start," those who "eventually understand and become aware," and those who "never understand or become aware." Most human beings fall into the last category (SZSX], p. 767; Wei Jieting, SunZhongshan, p.2.43)6. This point was made, for example, by Wang Jiaxiang, a Communist political officer, in a 1937 article (Leng and Palmer, Sun Yat-sen and Communism, pp. III-II:Z.). 7· See Chatterjee, "A Response to Taylor's 'Modes of Civil Society,"' on the conflict between "narratives of community" and "narratives of society" in European and Third World thought and history. 8. Schiffrin, Sun Yat-sen, p. 2.. 9· Ibid., p. 3 6 5. Schiffrin argues that Sun's leadership was "entrepreneurial ... in the sense that the leader provides no original ideas, only the means for achieving what everyone feels is required." ro. ZLQJ, r: 94· See his attack on Rousseau from the standpoint of "historical evolution." rr. ZLQ], I: 90 and 91. 12. ZLQJ, I: Io6. In .contrast to Europeans, Chinese had an excess of freedom because of the minimal demands of the imperial regimes. I3. ZLQ], I: 7I4· 14· Wei Jieting, Sun Zhongshan, pp. 2.77-278. I 5. Young-tsu Wong makes this point in his review of a recent work by Sidney H. Chang and Leonard H. D. Gordon (All Under Heaven). Chang and Gordon are convinced that Sun was a great thinker. See Wei Jieting, Sun Zhangshan, pp. 58-75, for a detailed discussion of controversies on Taiwan and the mainland over the meaning of minsheng in relation to socialism and idealist and materialist conceptions of history.

432

Notes to Pages 329-34

I6. Bauer, China and the Search for Happiness, p. 347· I7. For descriptions of the circumstances surrounding the lectures, see Wilbur, Sun Yat-sen, pp. 197-207; and Zou Lu, Huigulu, 1: 169-79. 18. Paul Cohen, Between Tradition and Modenzity, p. 260. 19. Ibid., p. 242-43. 20. ZLQ], I: 55· Both anarchism and communism are described as "old things China had thousands of years ago." 2r. See Dikotter, The Discourse of Race in China, for a discussion of how minzu "integrated both the notion of people (min) and the fiction of descent (zu)" to make "race as nation ... a conceptual extension of race as lineage" (p. 97). 22. ZLQ], I: I03. 23. For a stimulating scholarly perspective on this possibility, see Littrup, "The Un-Oppressive State." 24. Quoted and discussed in Pusey, China and Charles Darwin, p. III. 25. ZLQJ, I: 2. 26. "May Fourth thinking shared widely in a feeling expressed early in the century by Liang Qichao and popularized in Sun Yat-sen's description of China as a 'pan of loose sand' that due to the subversive legacy of traditional Chinese morality and its institutional basis in the family, Chinese lacked the cohesiveness required for a viable society" (Dirlik, The Origins of Chinese Communism, p. I4I). 27. ZLQJ, I: 64. 28. Pusey, China and Charles Darwin, pp. I 59, I89. 29. Discussed in Hans J. Van de Ven, From Friend to Comrade, p. 28. 30. Mao Zedong, ]i, vol. I; originally published in Xiangjiang pinglun, July 2I, July 28, Aug. 4, 1919. 3 r. Dirlik, The Origins of Chinese Communism, p. r So. 32. Chu-yuan Cheng, Behind the Tiananmen Massacre, p. 74· 33· Young people organized such groups as communal experiments (Dirlik, The Origins of Chinese Communism, pp. I8r, 25I). 34· Cited in ibid., pp. I 34-3 5. 35· ZLQJ, r: 64-65. 3 6. Chatterjee, "A Response to Taylor's 'Modes of Civil Society,"' pp. I 2425. 37· Ibid., p. I24. 38. Ibid., p. 125. 39· Anderson, Imagined Communities. 40. ZLQ], I: 77. Sun believed that foreigners, with the exception of Bertrand Russell, miss the greatness of China because they see the individual rather than the family or group. 41. Ibid. 42. Philip C. Huang, Liang Ch'i-ch'ao, p. 149. Liang advised local selfgovernment and the formation of professional organizations "as concrete steps toward developing the organizational abilities of the Chinese people."

Notes to Pages 334-37

433

43· ZLQ], I: 72. 44· Sun's willingness to follow Soviet ideas and contemplate the elimination of merchant middlemen in the interests of ending the exploitation of workers and peasants may indicate a lack of real interest in preserving the communal or informal dimension of Chinese commerce. On the other hand, the absence in Sun's view of a solid juridical or associational basis for capitalism in China must have made erasing such structures easy to imagine. See Wei Jieting, Sun Zhangshan, p. SI· 4 5. For the impact of the May Fourth movement on the Nationalist Party, see Lu Fangshang, Geming shi zaiqi. 46. ZLQ], I: II3. Sun referred to student movements with the derogatory "campus upheaval" [xuechao] and depicted students as taking the unwelcome idea of freedom to the "society outside" the school, where ordinary Chinese reject it. 47· ZLQJ, I: I3I-32. Sun said that he hoped worker groups would follow good leaders but noted with some anger a Hankow labor newspaper's declaration that "we workers don't want leaders who wear long gowns." Such sentiments, according to Sun, failed to understand the true meaning of equality. 48. ZLQ], I: I39· 49· ZLQ], I: I I5. For example, Sun complained in I9I3, when several provinces refused to rise against Yuan Shikai and Yuan held on to power, that this was because he "had a very solid group while our party was a sheet of sand." 50. There are occasional exceptions. For example, Sun explained the "difficulty and pain" of the Chinese peasant by noting that under the Qing, "the government did not permit peasants to unite. If they formed a group, there was a danger of loss of property and life. Therefore, peasant families had no contact with each other and were like a sheet of sand" (ZLQJ, 2: 508). sr. ZLQJ, I: 6 5-66. 52. "Zai Shanghai xinwen jizhe zhaodaihui di yan shuo" (A speecl:l given at a reception for Shanghai reporters), SZSX], pp. 9 56-67. 53· See, e.g., the story he told about Qing official Zuo Zongtang's relations with the Elder Brother Society. Zuo first used and then destroyed this potent social organization, leaving Sun to lament that "when it came time for our revolution, there was no organ [jiguan] to be used" or, one might say, grasped. (ZLQ], I: 35-36). 54· ZLQJ, I: 69. 55. Cited in Wei Jieting, Sun Zhongshan, p. I 84. 56. Wilbur, Sun Yat-sen, p. I3. 57· See LukeS. K. Kwong, A Mosaic of a Hundred Days, for a particularly vivid description of the role played by such affiliations in late imperial elite politics. 58. See Schiffrin, Sun Yat-sen, p. 65, on Sun's "longing for gentry affiliations." 59· Furth, "The Patriarch's Legacy," p. I89, citing Patricia Ebrey's work. 6o. Woodside, "Emperors and the Chinese Political System," p. 12. 6r. Fei Xiaotong, From the Soil, p. 64. 62. Schwarcz, The Chinese Enlightenment, p. 70.

434

Notes to Pages 337-40 63. Wasserstrom, Student Politics in Twentieth Century China, p. 143. 64. Ibid., p. 128.

6 5. Goodman, "New Culture, Old Habits." 66. Wei Jieting, Sun Zhongshan, p. 250. 67. The May Fourth crowds of protesters underlined the importance of "mass psychology" (qunzhong xinli) and an interest in social psychology among intellectuals two decades old. See Lung-Kee Sun, "Social Psychology in the late Qing Period," esp. p. 258 for reference to mass psychology and the May Fourth Movement. 68. Perry ("Collective Violence in China") argues that without capitalism and the "Great Transformation" described by Karl Polyani to undermine local autonomy, both communities and the state gained strength. 69. Marie-Claire Bergere, The Golden Age, p. 138. 70. Fei Xiaotong (From the Soil, p. 76) notes that the "concept of ren is ... only a logical synthesis, a compilation of all the ethical qualities of private, personal relationships." 71. In societies undergoing changes associated with capitalist transformation, such feelings are reunited "at the level of the political community of the nation, borrowing from another narrative the rhetoric of love, duty, welfare and so forth" (Chatterjee, "A Response to Taylor's 'Modes of Civil Society,"' p. I29). 72. Wei Jieting, Sun Zhongshan, p. 28. 73· ZLQJ, I: 83-84. 74· Wei Jieting, Sun Zhongshan, p. 20. James Pusey (China and Charles Darwin, pp. 3 3 5, 34 I-4 5) points out that, despite his large debt to evolutionary thinking, Sun, like other Chinese influenced by Darwin and the Social Darwinists, sought to avoid the deterministic implications. Thus, democracy was the inevitable by-product of the evolutionary process, but Chinese could choose to adopt it in the struggle to survive. 7 5. For the racialist features of Sun's use of minzu, see Pusey, China and Charles Darwin, pp. 3 30-3 3 I. 77· ZLQJ, I: 159· 76. ZLQJ, I: I57· 78. ZLQ], I: I68. 79· ZLQJ, I: !69. So. ZLQJ, I: 70. 8 I. In the process, Sun followed a standard Chinese response to Western claims of individual autonomy. As Andrew Nathan ("Tiananmen and the Cosmos," p. 3 5) has pointed out, "Any Chinese can see that ... people are born into society, that from the beginning they have social values and goals. It remains a real question for Chinese, perhaps especially for those committed to democracy, how it is that human rights can become so important that in Ronald Dworkin's term, they 'trump' other social values." 8 2. Wei Jieting, Sun Zhongshan, p. 5 I. 83. Later interpreters such as Zhu Yixin (Dangyi ABC, pp. 50-51) saw anarchism and the Three People's Principles as similar in their commitment to equality but different in anarchism's applicability only to the small group level. Sun himself (ZLQJ, 1: 137, 143) warned that anarchism would lead to chaos and mob

Notes to Pages 340-43

435

rule. In a milder criticism, made in I923 and close in spirit to Zhu, he asserted that while anarchism was an understandable reaction to tyranny, you "cannot reform human nature and save the nation without government" (ZLQ.J, 2: 293 ). 84. Wei Jicting, Sun Zhongshan, p. 20. 85. ZLQ.J, I: 4S4. 86. ZLQ.J, I: 694. 87. Wei Jieting, Sun Zhongshan, p. I So. 88. ZLQ], I: 38. 89. For example, Joseph Levenson (Liang Ch'i-ch'ao, p. r6o) pointed out that Liang Qichao rejected Sun's notion that "old China had a nationalism which despotism and the principles of monarchy had stifled." 90· ZLQJ, I: 443-541. 91. As Qu Qiubai, on the eve of execution, wrote as the final line of his last testament: "Chinese bean curd is a very tasty thing. The best in the world." 92. ZLQ.J, I: 4S2. 93· ZLQ.J, I: 447· 95· ZLQ.T, I: 454· 94· ZLQ.J, I: 44s. 96. Henriot, Shanghai I927-I937, p. 208. Most resistance was finally overcome by a public education campaign of a kind Sun would have likely approved of. 97· Ibid.,p.4S5. 98. This distinction was commonplace during this period, used by individuals like Kang Youwei to distinguish between "the politics of principle and compassion and the politics of expediency and force" (Hao Chang, Chinese Intellectuals in Crisis, p. z6). 99· Cited in Wei Jieting, Sun Zhongshan, p. I So. IOO. Ibid., p. 157· IOI. Quoted in Schiffrin, Sun Yat-sen, p. 3I7. I02. Fanon, The Wretched of the Earth, p. 2I7. 103. ZLQ.J, I: 34· I04. This type of populism became evident during Sun's brief tenure as acting president in I 9 I 2 when he ensured that the so-called base or mean peoples received full rights of citizenship (Zhang Qicheng and Guo Zhikun, Sun Zhongshan, p. 57)IO 5. Kuhn, "The Development of Local Government," p. 3 6o. Kuhn emphasizes the bureaucratic subversion of Sun's populist and democratic ideals whereby "the very units which were supposed to be vehicles for self-government became units for deeper bureaucratic penetration of local society." I o6. Price, "Constitutional Alternatives," p. 2 5I . Price specifically cites the "sheet of sand"-related ideas of Sun as problematical in this regard. I07. In his·discussionof democratic institutions such as elections, Sun (ZLQ.J, I: I 7 5ff ) has the people acting as the engineer in control of the machinery of government. He seemed willing to grant ordinary citizens more real influence as the sovereign power, but he also stressed safe and expert operation of government and warned against the fear that government could be too powerful. One looks in vain elsewhere in the Sanmin zhuyi for the kind of confidence in the people required for such a political engineering feat. For Sun's ideas of political tutelage,

Notes to Pages 343-45 see the discussion and texts in Julie Lee Wei eta!., Prescriptions for Saving China, pp. xxix, 205-6, 209-IO. 108. Cited as an epigraph in Hans Konig, 1968: A Personal Report. I09. M. Cohen, "Being Chinese." IIO. Schoppa, Chinese Elites and Political Change, p. 6. Schoppa posits a shifting balance of power between primary and composite tendencies but notes that the primary tended to remain "dominant." III. Tu-Ki Min, National Polity and Local Power, p. I03. Min discusses the complex relationship between the junxian school of centralizers and the fengjian ("feudal") localizers. Many fengjian proponents felt that building up local society was the best way to achieve unity in the empire. A junxian position might include faith in the willingness of local elites to cooperate in carrying out the center's policies. At other times the two envisioned the social order at each other's expense. II2. Ibid. I I 3. Duara, Rescuing History from the Nation; see esp. chap. 5, "The Genealogy of Fengjian or Feudalism: Narratives of Civil Society and State." I I4. Strand, "An Early Republican Perspective." II5. In a speech in Nov. I924 Sun described how he encouraged journalists in Shanghai to form a group so that they could join him in a national "citizens assembly" (ZLQ], 2: 567). II6. ZLQJ, 2: I56-57· I I7. Strand, "An Early Republican Perspective"; and Judge, "Public Opinion and the New Politics of Contestation." A recent discussion of the political thought of James Madison argues that Madison "saw both an extant community and the possibility of an extended national community grounded in public spiritedness and buttressed by institutions of learning and participation in shared values. The latter was to be nurtured by a dialogic enterprise-between citizen and citizen, and citizen and government-grounded in attachment to common values of selfgovernment" (Carter and Kobylka, "The Dialogic Community," p. so). The principle of extending the community to the nation and grounding the enterprise in common values is quite close to Sun's depiction of the Chinese republic, but since Sun's citizen is constructed almost entirely of cultural tendencies and psychological characteristics, the sense of agency required for dialogue is missing. II 8. Carter and Kobylka, "The Dialogic Community." II9. Wasserstrom, Student Protests in Twentieth-Century China, pp. 17884. Wasserstrom notes that in the 1930s these "orchestrated meetings between student petitioners and political leaders" as "rites of supplication" sometimes worked to "legitimate Jiang's image as a popular ruler" but, as the protests showed, "students were not always prepared to be so passive" (p. r8r). 120. Shen Tong, Almost a Revolution, p. 170. Shen, a prominent student leader, used the phrase "A Dou ruins the nation" on a leaflet in reference both to Li Peng's failings and to his relationship to Zhou Enlai (imagined in this case as A Dou's father, Liu Bei, rather than Zhuge Liang). This barb also neatly reverses the formulaic "Don't treat the masses as if they were A Dou." For the popular image of Zhou as Zhuge Liang, see Nathan, Chinese Democracy, p. 176.

Notes to Pages 346-so

437

Chapter 14 I. This quote is from an interview with Zhang Yimou with an anonymous interviewer as reported in the China News Digest (Books and Journals Review), Apr. 2.5, I993, on the internet. The dictionary pronunciation is shuofa. 2.. Bhabha, "DissemiNation," p. 2.97. 3· Foucault, "The Subject and Power," p. 2I2. 4· Duara, Rescuing History from the Nation. 5· Mayfair Yang, "Of Gender, State Censorship, and Overseas Capital," p.

2.99·

6. This section and the following one are condensed and updated from an earlier paper. See Anagnost, "Socialist Ethics and the Legal System." 7· Sun Meiyao, Xuanchuan gongzuo shiyong shouce, p. I98. 8. Li Zhenjie and Bai Yukun, Zhongguo baokan ciyu, p. 2. 55. 9· Renmin ribao, Feb. I I, I982., p. I. IO. See Fujian ribao, June 6, and Dec. 30, I982.; and Zhongguo nongminbao, Apr. I, I982, for accounts of this model in Zhongzhen. II. Fujian ribao, June 7, I982.. I2.. In September of that year, the party Central Committee passed a resolution on a guiding policy for the establishment of a socialist spiritual civilization. I 3. All the townships I visited were designated model units, having attained the status of "civilized" (wenming), either at the provincial or national levels. I4· Shu-min Huang (The Spiral Road, p. Io7) mentions village compacts as part of the normal functioning of local government. In the Fujian village studied by Huang, the party secretary himself wrote the local laws and ordinances and in addition presided over litigations and dispensed justice. I5. Du Xichuan eta!., "Are Village Charters the Same as Law?," pp. I-2.. I 6. See, e.g., the report on the county-wide program to draft village compacts in Jiyuan county (Henan) in Renmin ribao, Apr. 7, r 9 8 2, p. I. The measure was used in this case to correct a rapidly deteriorating social order plagued by an increasing number of disputes over water and land, theft, fighting, and other problems. The compacts were reported as especially effective in reducing the number of civil disputes referred upward from the commune level to the county authorities. I 7. Fengcheng county in Liaoning province instituted "village people's socialist ethics appraisal committees" (cunmin jiaoyu huodongzu), reported in Nongcun gongzuo tongxun, July I988, pp. 42-43. These committees are composed of three to five members assigned decimal units of ten households as the object of their educational activities. The group members are elected by the community or are selected by the party branch in consultation with the masses. They tend to fall into six categories of persons: party members, youth league members, retired village cadres, people of high moral character, female "heads of household" (nudangjia), and individuals of known ability (nengren). Among other activities, these small groups carry out propaganda and educational activities, mobilize the masses for public works, resolve conflicts, transmit information, and provide mediation services. The shehui daode pingyi hui were reported for Wei

Notes to Pages 350-55 county, Hebei province, in the Nongmin ribao,July 15,1987. A similar organization in Hunan province named "small group for the public discussion of ethics" (daode gongyi xiaozu) was reported in the Nongmin ribao, May 9, 1987, p. r. Hongbai xishi lishihui was reported in the Nongmin ribao, July 15, 1987, p. 2, and wenming zu in Renmin ribao, May 27, 1986, p. 5· 18. Nongmin ribao, Sept. 12, 1986, p. 3· Many of the cadres I interviewed in 1991 freely admitted the use of fines and the dismantling of houses to ensure enforcement of the birth policy. Indeed such punishments were attributed with providing the "binding force" necessary to ensure compliance. 19. In speaking of physical punishment, the editorial also urged local officials to abolish any "lineage regulations" (zugui) or "family rules" (jiagui) that might incite lineage feeling or represent lineage power. It cited one lineage with the surname of Li that had drafted lineage regulations (zugui gongyue). In addition to fines, there were also stipulated physical punishments (dapigu zhilei). These, the editorial insisted, are bizarre and "very uncivilized" (hen bu wemning). This connection between village compacts and those of lineages and households provides a gratuitous commentary on the "patriarchal" (jiazhangzhi) character of the compacts. 20. The editorial only obliquely addresses the issue of fiscal responsibility in collecting the fines. It cites unfavorably the regulation of one township that stipulated that the money collected was to go to the level of government responsible for levying them. In cases involving more than one level of government, it was to be divided up proportionally. But one wonders how much the fines serve as revenue or as an incentive for official graft. 21. Nongmin ribao, Sept. 12, 1986, p. 3· 22. From the interview with Zhang Yimou in China News Digest (see note r above). 23. Ewald, "Norms," p. 139. 24. Foucault, History of Sexuality, pp. 136-37. 2 5. Certainly the use of abortion to ensure that birth quotas are met fits Foucault's sense of biopower as not the right to take life or let live so much as "the power to foster life or disallow it to the point of death" (ibid., p. 138; italics in the original). 26. Ibid., p. I44· 2 7. It is important to note the historically specific ground of Foucault's theory, his Nietszchian horror of constructing a universalizing history, and his apparent lack of interest in looking at how cultural difference figures into the constitution of the "West" as imperial subject. 28. Ewald, "Norms," p. 141. 29. Foucault, History of Sexuality, pp. 140-41. 30. Ewald, "Norms," p. 140. 3 r. Ibid., p. 14 5. 3 2. This was also true of the categories that made up the class-status system of the Maoist period; see Billeter, "The System of Class-Status." 3 3. To some extent, this new discursive domain within the academy represents an increasing autonomy of intellectual thought from the language of the party,

Notes to Pages 355-6r

439

as I discovered in 1989 when I tried to put my newspaper vocabulary to work in explaining my research goals to Chinese colleagues. They found my language inappropriate for scientific work. Elizabeth Perry (pers. comm.) suggests that Chinese and Western social science constructions of the "norm" confront each other in designating the "typical" unit as the proper object of scholarly study. In Chinese, "typical" translates as dianxing in the sense of corresponding to a type or model, whereas Western social science operates in terms of the typical as "the average," which in Chinese would translate as yiban. 34· See Foucault, Discipline and Punish, p. So. 3 5. Ibid., pp. 8o-82. 3 6. In addition to any historical linkage between post-Mao village regulations and the late imperial xiangyue, their characterization as a form of "democracy" (minzhu) and "self-government" needs to be tracked throughout the successive tides of modernizing governments in China. See Strand's chapter in this volume. 37· For a discussion of "the people" (renmin) as a product of the nationstate, see Lydia Liu, "Translingual Practice," p. 179 et passim. 38. Bhabha, "DissemiNation," p. 292. 39· Anderson, Imagined Communities. For post-colonial historiography, see the work by the Subaltern Studies Group (Guha, "Some Aspects"; Prakash, "Writing Post-Orientalist Histories"), and Young, White Mythologies. The influence of these movements in Chinese studies is represented by some recent work by Duara (cited above) and in the first volume of the journal positions: east asia cultures critique. 40. Anderson, Imagined Communities, pp. I r-12. 41. Duara, Rescuing History from the Nation, p. 29. 42. Ibid., pp. I3-I4· 43· Of course, this characterization is complicated by that fact that China is recognized as having an ancient wenming, which is not just thought of in terms of having a "Great Tradition" but as being in a state of relative "evolutionary advancement" with respect to the rest of the world (primarily Western Europe). The attainment of wenming symbolizes for many the attainment of a comparable position of leadership for China in the near future. 44· Jing Wang, "Heshang and the Paradoxes of Chinese Enlightenment," p. 24. 45· Ibid.,p. 29. 46. For a discussion of the dual meanings of representation as "portrait" and "proxy," see Spivak, "Can the Subaltern Speak?," p. 276 etpassim. 4 7. In one township where I interviewed, one of the principal reasons cited for the 1990 revision of the local compact was the growing concern over socialorder issues with the incorporation of an expanding population of outsiders. These outsiders included a large number of women who had married in and who were now having marital difficulties and contributing to a rapidly inflating divorce rate. 48. See "Yan'an shiqi de cunmin gongyue" (A village people's compact from the Yan'an period), People's Daily, Dec. 9, 1990, p. 5. 49· Rofel, "Liberation Nostalgia and a Yearning for Modernity."

440

Notes to Pages 36r-65

50. This phrase was suggested by Ted Huters. 5 I. Duara, Rescuing History from the Nation, pp. 3 I-3 2. 52. In the post-Mao period this uneasy positioning between the past and future is omnipresent in people's consciousness as well as in the often contradictory policies for policing the cultural realm. The post-Mao period has seen the recuperation of a tradition that re-creates a glorious past that is in some way directly connected to the present. For instance, the fad for building "old towns" (fanggujie) invokes the former glories of China's ancient cities while providing the foci for the transforming development of a commodity culture. Tradition clearly becomes a commodity that constructs the glory of the national past, but one that must be not only "reinvented" in opposition to the Maoist narratives of the unmitigated darkness of the feudal past but also sanitized and channeled for the promotion of internal tourism and foreign investment. 53· Duara, Rescuing History from the Nation, p. 32. 54· Derrida, OfGrammatology, p. 155. 55· Lydia Liu, "Translingual Practice," pp.I84-86. 56. Michael Taussig, Mimesis and Alterity, p. xvii. 57· In suggesting this, I do not mean to imply that this process is "liberatory" in all its implications but merely to point to a proliferation of possible subjectivities, some of which derive from a commodity economy, but which may not be entirely divorced from the state and its exhortations for peasants to aspire to a "reasonable standard of living" (xiaokang shenghuo). 58. That Zhang Yimou himself is aware that The Story of Qiu Ju may be read as an allegory of how the bans on his films were removed in China is clear from the interview in the China News Digest (see note I above). "It does not matter whether I am Qiu Ju or whether her story is like mine since this is a very ordinary story that happens all the time in China .... The films Uudou and Raise the Red Lantern] were never released, and nobody ever gave me a 'shuafa' [sic] about the ban." 59· The village head was found guilty through a technicality. In beating up Qiu Ju's husband, he had broken a rib. This demonstrable injury was legally worthy of redress, but not the beating itself. 6o. Interview with Zhang Yimou in the China News Digest (see note I).

Works Cited

Abbreviations:

CS]C GX]BCS H]AS ]AS SBBY SBCK SKQS

TZ WXTK

Congshu jicheng iH!J~/VG Guoxue jiben congshu ~~£~fi~ Harvard Journal of Asiatic Studies Journal of Asian Studies Sibu beiyao ~m:gB{ii~ Sibu congkan lm:gBJifU Siku quanshu lmM!.~if (1773-1782) Tongzhi ~;&(provincial gazetteer) (Qing chao) wenxian tongkao m¥JBt~~~

Abe Takeo ~:gB~f~. Gendai shi no kenkyu :lt1i;§t:.O)iiff~. Tokyo: Subunsha, 1972. Ah Ying jliif~ (Qian Xingcun ~:frlliB). Wan Qing xiaoshuo shi B$llWf1N5t§t:.. 1955· Hong Kong: Zhonghua, I973· Ah Ying jliif~ (Qian Xingcun ~:frlliB), ed. Xiaoshuo xiqu yanjiu juan ;}~tiE!H m~;g. Shanghai: Zhonghua, 1960. Vol. I of Wan Qing wenxue congchao Jl$llfrl3t~fi~jl.

Ai Nanying X:f¥i~. Tianyongzi ji :;Rffi!F:f~. I699· Taibei: Yiwen, I980. Altieri, Charles. "An Idea and Ideal of a Literary Canon." In Robert von Hallberg, ed., Canons. Chicago: University of Chicago Press, 1980, pp. 41-64. Anagnost, Ann. "Socialist Ethics and the Legal System." In Jeffrey N. Wasserstrom and Elizabeth]. Perry, eds. Popular Protest and Political Culture in Modern China. 1st ed. Boulder, Colo.: Westview, 1992, pp.I77-206. Anderson, Benedict. Imagined Communities: Reflections on the Origin and Spread of Nationalism. London: Verso, I 9 8 3. Anhui sheng jinian Wu Jingzi dansheng erbai bashi zhounian weiyuanhui ~~ ~~c~~{{&;j$~~= Ef )\ Jti]if.~~l!f. Rutin waishi yanjiu lunwenjif$ ft)'} §!:: iiff~fRB3t~. Hefei: Anhui renmin, 1982. Armstrong, Nancy. Desire and Domestic Fiction: A Political History of the Novel. Oxford: Oxford University Press, 1987.

+

442

Works Cited

Asycough, Florence. A Chinese Mirror. Boston: Houghton-Mifflin, 1926. Ayuwang shanzhi. JlilTW±rlJ;E;. 1619. Eadie, Bertrand, and Pierre Birnbaum. The Sociology of the State. Chicago: University of Chicago Press, 1983. Bauer, Wolfgang. China and the Search for Happiness. New York: Seabury Press, 1976. Beijing daxue, Gu wenxian yanjiu suo ::lt:!R:k~1l3tikliff~.PJT, ed. Quan Song shi ~*ffif, vol. r. Beijing: Beijing daxue, 199 r. Bendix, Reinhard. Kings or People. Berkeley: University of California Press, 1978. Bergen, Paul. "A Visit to Tai Shan." Chinese Recorder 29,.12 (Dec. r888): 54146. Bergere, Marie-Claire. The Golden Age of the Chinese Bourgeoisie. New York: Cambridge University Press, 1989. Bhabha, Homi K. "DissemiNation: Time, Narrative, and the Margins of the Modern Nation." In Homi K. Bhabba, ed., Nation and Narration. London: Routledge, 1990, pp. 291-322. Bi Zhongyou $-{rj:l~fjf. Xitai ji 5-EJ~. SQKS. Bianli yuanzhi. m;fiJJlJ'c;E;. I830. Billeter, Jean-Francais. "The System of Class-Status." In Stuart R. Schram, ed., The Scope of State Power in China. London: SOAS, 198 5, pp. 127-69. Bochi shanzhi. ~ng!li;E;. 1920. Bol, Peter K. "The Examination System and the Shi." Asia Major, 3d series, 3.2 (1990). - - - . "Government, Society, and State: On the Political Visions of Sima Guang (IOI9-I086) and Wang An-shi (ro2r-ro86)." In Robert Hymes and Conrad Schirokauer, eds., Ordering the World. Berkeley: University of California Press, 1993, pp. 128-92. ---."Seeking Common Ground: Han Literati Under Jurchen Rule." H]AS 47·2 (1987): 461-538. - - - . "This Culture of Ours": Intellectual Transitions in T'ang and Sung China. Stanford: Stanford University Press, 1992. Bourdieu, Pierre, and Monique de Saint-Martin. "Scholastic Values and the Values of the Educational System." In J. Eggleston, ed., Contemporary Research in the Sociology of the Educational System. London: Metheun, I 974, pp. 3 3 8-7 r. Boshan xianzhi "!WiliW-*i0· 1937. Boulding, Kenneth. Ecodynamics: A New Theory of Societal Evolution. Beverly Hills: Sage Publications, 1978. Brook, Timothy. Praying for Power: Buddhism and the Formation of Gentry Society in Late-Ming China. Cambridge, Mass.: Harvard University, Council on East Asian Studies, I993· - - - . "The Spatial Structure of Ming Local Administration." Late Imperial China 6.r (June 1985): I-55· Brooks, Peter. Reading for the Plot: Design and Intention in Narrative. New York: Vintage Books, 1984. Cai Jingkang ~:W~. "Wan Qing xiaoshuo lilun chulun" B_fuff!f;J\imrJ!ljifli;f]JljifH. Gudai wenxue lilun yanjiu[fJ {-1.:;3t~:FJ![~liffJ'E r (1979 ): 403-21.

Works Cited

443

Cai Meibiao~~~- Yuandai baihuabei jilu JC{-t E3~3Ji!l!~~- Beijing: Kexue, I955· Canglang xiaozhi ~¥$H\;G;. I696. Cao Xueqinl'f~Jf. Hong lou meng ~J:fl~. Singapore: Xingzhou shijie, 1969. - - - . The Story of the Stone. Vol. I. Trans. David Hawkes. Harmondsworth, Eng.: Penguin Books, 1973. Caoxi tongzhi l!/~~;G;. I672. Caozhoufu Heze xian xiangtu zhi l!/1'i'IJfJVilf~\I!Ji;~~±;G;. I9o8. Garlitz, Katherine. "Desire, Danger, and the Body: Stories of Women's Virtue in Late Ming China." In Christina Gilmartin et a!., eds., Engendering China: Women, Culture, and the State. Cambridge, Mass.: Harvard University Press, I994, pp. IOI-24. ---.The Rhetoric of "Chin p'ing mei." Bloomington: Indiana University Press, I986. - - - . "The Role of Drama in the Chin p'ing mei." Ph.D. diss., University of Chicago, I978. Carpenter, Jennifer. "Value and Women's Writing from the Han to the Song." Unpublished seminar paper, I99I. Carter, B. K., and J. F. Kobylka. "The Dialogic Community: Education, Leadership, and Participation in James Madison's Thought." Review of Politics 52 (Winter I990). Chaffee, John W. The Thorny Gates of Learning in Sung China: A Social History of Examinations. New York: Cambridge University Press, I98 5. Chan, David. The Usurpation of the Prince ofYan, I403-I424. Taibei: Chinese Materials Center, I97 5. Chan, Hok-lam. "The Chien-wen, Yung-lo, Hung-hsi, and Hsuan-te Reigns, I399-I43 5·" In Frederic W. Mote and Denis Twitchett, eds., The Cambridge History of China. vol. 7, pt. I, The Ming Dynasty, I3 68-I644· New York: Cambridge University Press, I988, pp. I82-304. - - - . "The Rise of Ming T'ai-tsu (I 3 68-98 ): Facts and Fictions in Early Ming Historiography." journal of the American Oriental Society 9 5 ( I975 ): 6797I5. - - - . "Wei Su." In Dictionary of Ming Biography. New York: Columbia University Press, I976, pp. 1464-67. Chan, Wing-tsit. "The Ch'eng-Chu School of Early Ming." In William T. deBary, ed., Self and Society in Ming Thought. New York: Columbia University Press, I970, pp. 29-51. - - - . "Chu Hsi and Yuan Nee-Confucianism." In Hok-lam Chan and Wm. Theodore deBary, eds., Yuan Thought: Chinese Thought and Religion Under the Mongols. New York: Columbia University Press, I982, pp. I97-23 I. Chandler, James. "The Pope Controversy: Romantic Poetics and the English Canon." Critical Inquiry, I0.3 (I984): 48I-509. Chang, H. C. Chinese Literature: Popular Fiction and Drama. New York: Columbia University Press, I973. Chang, Hao. Chinese Intellectuals in Crisis: Search for Order and Meaning, I 89oI9II. Berkeley: University of California Press, I987.

Works Cited

444

Chang, Kang-I Sun. The Late-Ming Poet Ch'en Tzu-lung: Crises of Love and Loyalism. New Haven: Yale University Press, 1991. - - - . "Liu Shih and Hsi.i Ts'an: Feminine or Feminist?" In Pauline Yu, ed., Voices of the Song Lyric in China. Berkeley: University of California Press, 1994, pp. !69-87.

- - - . "Ming and Qing Anthologies of Women's Poetry and Their Selection Strategies." In Ellen Widmer and Kang-i Sun Chang, eds., Writing Women in Late Imperial China. Stanford: Stanford University Press, 1997, pp. 14770.

---.Six Dynasties Poetry. Princeton: Princeton University Press, 1986. Chang, Kang-i Sun, and Haun Saussy, eds. Chinese Women Poets: An Anthology of Poetry and Criticism. Stanford: Stanford University Press, forthcoming. Changqing xianzhi :Rm\l\h\;0. 1835, 1934. Chao Buzhi S&:fmZ. ]ibei Chao xiansheng file ji ~~cS&$t:1:.~1l:fJ~. SBCK. Chaocheng xian xiangtu zhi lj!J3i:JiX\\lA\MlE±;0. 1920. Chatterjee, Partha. "A Response to Taylor's 'Modes of Civil Society."' Public Culture 3.1 (Fall 1990): 119-32. Chavannes, Edouard. Le T'ai Chan. Paris: Ernest Leroux, 1910. Chaves, Jonathan. "The Panoply of Images: A Reconsideration of the Literary Theory of the Kung-an School." In Susan Bush and Christian Murck, eds., Theories of the Arts in China. Princeton: Princeton University Press, I 9 8 3, pp. 341-64. Chen Bangyan !l*:FG~, ed. Shi da cainu +:fct:tz::. Shanghai: Shanghai guji, 1991.

Chen Guan ~Jfifl. "Siming zun Yao lu" [91!13#~~- In Sheng Song wenxuan quanji ~5f;:3t~~~ 1832. Chen Hongmou pjfi7,R~. Peiyuantang oucungao ±:g~:§t{Mff.ff.lj. 1896. Chen Li.i !l*fl&. Anya tang ji ~~¥:~. SKQS. Chen Meilin pJfi~#. Wu ]ingzi yanjiu ~1Wlf¥liff~. Shanghai: Shanghai guji, I984.

=

Chen Pingyuan pjf[Zj=ffil:. Ershi shiji Zhongguo xiaoshuo shi: diyi juan, r897I9I6 +t!Hcr:J:l ~lj\~9::.:ffi-!§. Beijing: Beijing daxue, 1989. Chen Qinghao pJfii2f¥i1f and Wang Qiugui .±.'f)c,f:E, eds. Shandong minjian gushi ji JlL*~Fs9ti5::~~- Taibei: n.p., 1989. Chen Renxi !l*C~~- Huang Ming shifa lu ¥BJlt±ti:t:~. 1630. Chen Ruheng ~jft~J. Wu Jingzi zhuan ~1WI:;f¥1$. Shanghai: Shanghai wenyi, 1981.

Chen Shan pjfi~. Menshi xinhua i!'n!ll~~li. Reprinted-Shanghai: n.p., 1920. Chen Shidao p!l[fljj]§l[. Houshanji {kJli~. SBBY. Chen Weisong pJfi*fE~. Furenji MILA.~. 1833-44- Zhaodai congshu BtH-\':~iij:. Chen Xiang pjfillf. Guling ji til!~. SKQS. Chen Xin pjfi~ and Du Weimei U*li*. "Rutin waishi di wushiliu hui zhenwei bian" f~#5'1-9::.:ffin+7\@l~~~¥. In Anhui sheng jinian Wu Jingzi dansheng erbai bashi zhounian weiyuanhui, ed. Rulin waishi yanjiu lunwenji. Chen Xinghui pjfi¥;~. "Ershi nian mudu zhi guai xianzhuang" yanjiu +1=f-

=

Works Cited

445

§§if Z.'i£fj.#j\li1f~. National Taiwan University, History and Chinese Literature series, 6I. Taibei, 1983. Chen Yuan ~j)g. Mingji Dian-Qian fojiao kao i:JJ:J:~':~~f~Uj(~. 1940. Reprinted-Beijing: Zhonghua Shuju, 1962. Cheng, Chu-yuan. Behind the Tiananmen Massacre: Social, Political, and Economic Ferment in China. Boulder: Westview Press, 1990. Cheng Duanli fj)iffijff~. Chengshi jiashu dushu fennian richeng fj.B:;~~gft~?t if: B fj. Hefei: Huangshan, ·1992. ---."Wei zhai ji" !':'Bl!f~. Siming congshu 12]1jjj~~Cheng Duanxue fjftffil~. ]izhou ~1B}~. Siming cong shu 12]8Jj~~Cheng Hao fjl'j and Cheng Yi fj~. Er Cheng ji. =W~. Vols. 1-4. Beijing: Zhonghua, 1981. Cheng Yuanmin fj:JGJID[. "San jing xinyi yu Zishuo kechang xianwei lun" .=:J~ jfjf~!H~+il~H4t~~1J)&§jfg. In Qu Wanli xiansheng qizhi rongqing lunwenji llil fltli!_:)t;±. -t:lJ:::~JlllHiflix~. Taibei: Lianjing, 1978, pp. 249-8 5. Chikusa Masaaki ''zqg;. ff~. Chugoku bukk yo shakaishi kenkyii rp ~ f~~ffd: W/"§l:litf~. Kyoto: Tohosha, 1982. Chin, Ann-ping, and Mansfield Freeman. Tai Chen on Mencius: Explorations in Words and Meaning. New Haven: Yale University Press, 1990. Chiping xianzhi tFf-~"*13· 193 5. Chu Hua :fif!ilf. Hucheng beikao ~f]i\\{illl~. 1813. Reprinted-n.p., 1935. Chugoku noson kanko chosakai r:p ~!lf't't.JID'Wlili'!FW/". Chiigoku nason kanko chosa r:j:l~!lf't'tJ:fiW\'iJi'lf. Tokyo: Iwanami, 1953-58. Clunas, Craig. Superfluous Things: Material Culture and Social Status in Early Modern China. Urbana: University of Illinois Press, 1991. Cohen, Mitchell. The Wager of Lucien Goldmann. Princeton: Princeton University Press, 1994. Cohen, Myron. "Being Chinese: The Peripheralization of Traditional Identity." Daedalus 120.2 (Spring 1991): 113-34. Cohen, Paul A. Between Tradition and Modernity: Wang T'ao and Reform in Late Ch'ing China. Cambridge, Mass.: Harvard University Press, 1974· Crawford, Robert, Harry Lamley, and Albert Mann. "Fang Hsiao-ju in Light of Early Ming Society." Monumenta Serica 15 (1956): 303-27. Cui Xiuguo ~j§:~ andJi Aiqin ~~~- Taidai shiji *-111 §l:l&Jli. Ji'nan: Xinhua, 1987.

Da Ming huidian ::f;:ljjjW/" ~- 1588. Da Ming Taizong Wen huangdi shilu j;:Ejj:;t*Y::~ffl'jfif191:1!\~ Jiusiyisheng fL %- :!t Jizhi >ltiz juan ~ jueju ~ii} Jun xian 1tWv% junquan :ll[fff J unxian tifl Wv% junzi t\"T-

:llt*'

Kaiyuan /Jfl5[; Kang Youwei l.lJHJ~ Kangxi f>l ~~ Kanwa ~m ketou l@li'IT[ Kong Shangren fL f.\Hf Kong Wuzhong JL:ltt{9' Kongtongzi ji ~m-T~ Kuang Chaoren gjffi).,. Kuiyu kun :f:EI!

mlit

Langxian 1.BHW Lao Can youji ~9i3l!H2 li (principle) )]! li (ritual) ii\li li (village) 1t!. Li Bangyan *'fllg Li Bo Li chi (Li ji) ff~~c li dang gu wu er )]! if?&~= LiE J,l!l;~~ Li Fu *iii: Li Guinian *~if Li Jingyi ~jl'l:~ LiKui *~ Li Mengyang *~!Wi LiMo *~ Li Panlong *~JJ~ Li Peng '*'llll~

*8

479

*m'

Li Qingzhao ~.s Li Shimin *t!t.B: Li Xiangjun *~'Et li yan ft~ Li Ye *i~ Li Yin Li Zhaoqi *Btlfc

*12§

*""

LiZhi !ian qi se ~tlt 5 Lian Zining liang Liang Qichao W:~Jffi Liang Yin W: Ji[ Liangjin .R>Fs liangzhi .R 9;0 lie ?,!{ Lie chao shi ji :9Ulfifliff~ lienii ?.!ttl: Lienu zhuan ?,!{:ft~ Liji zhengyi ff~~clE~ lijia 1t!. Ef! Lin Chong ## Lin Hong #i~ Lin Huan #lli! Lin Yuzhi #~@.: ling 1!1 Ling Mengchu ~~:f)] Lingbao illl lingxiu zhi qi 1!13§'2.~ Linxia ci xuan #Tiiiil~ linxia fengqi #T !!\~ Liu Anjie ~U*frl Liu Ban l\fU J1l1 Liu E ~U~~ Liu Guan ~U~ Liu Ling ~U #; Liu Sanwu ~U .=:. :ef Liu Shi tPP~ Liu Wenwei liiU::X:m Liu Xie ~U~ Liu Zhi ~U~ lixue )]!* Lixue cuanyan J]!*g~ Lizong )]!* LoBin *IW Longqing ~.ill' Lou brothers ~f®0Tlu cai shi shi junzi dabing R::t~±tl"

m

**r$

-T;km Lu Guangzu

~Jfdt!l.

Character List Lu Jiuyuan fiifL#Ilf LuNan LiM Lu Qingzi fii~PT­ Lu Shiyong Iii~~ Lu Taixue shi jiu ao gong hou

Li::k~

li\flE!jjjj(1}~

Lu Xiujing [ii~JW LuXun ~ill. Lu Zhailang ~ jf R~ Lu, miss ~,j,~Jl Iii ~ Lii Kun g t£11 Lii LiuJiang 8 if~ Lii Nangong g i¥.i0 Iii shi ~\i\1 Lii shi yaofa ~\i\1~1'* Lii Tao g~ Lii Xiaqing g :8r~ Lii Zuqian 8WtEI.~ Juan ~L Lun wen yi mfli3t~ lunqi you shi wu shi mflj;l:!;i'f\i\1~\i\f Luo Qilan .~~iiili rna ~ Ma Chunshang ,%~_t Ma Jiao ,lf!i, tz;t Ma taihou (Empress Ma) ,lf!i,*J§ Mao Pang =§~ Mao Qiling =Biififi Mao Xiang ~ ll! Mao Yuanyi Jf; fi Mao Zedong =§~J![ Mazei ~lltft Mencius ]ETMengzi ]ETMengzi jiewen ]ET-f!J3t Miaoshan ~.':!>~ Min islJ min wei gui ~mit ming 1fP Ming ElJ3 ming dao El)3~ Ming fu chou zhi yi ElJ3{iHiLZ.~ Ming wen an H_ij?{(~ mingshi ~± Mingshi gao El)3.§(:flilj minquan ~;fl minsheng ~~ Minzhong shi caizi isll!:fl+::tT-

'*

minzhu ~::!:: minzu ~1* minzu de da lianhe ~1*1¥J::klll~ir mu iifj,. Mudan ting tf:f:l-=$= muyu

*ffi.i.

Najia »!l{!J[J Nanhai i¥.iw Nanhua Si i¥.i¥il'~ Nanjing i¥.i 5?.: nanzi you de bian shi cai

~T-i'ftf-{Jl!~

::t nei yan pu chu kun wai 17\J~=i'l±iill!l7i­ nengren H~A Ni Yue {Jl-ffi Niannu ~~Jl. Niu Pulang 4'-nU~ Nongcun gongzuo tongxun Jl;f't I f'F:lm

m

Nongmin ribao /!!!;~ B ¥fli Notes of the Yuewei Hermitage ;§t~~t.

nii zhong zhi ru -J;.:. $ niidangjjia -J;.:. &' Niijie -J;.:.~ niizi wu cai bian shi de

*

z {i* P.:.T-~::t{Jl!~

1!! Ouyang Xiu ~~~ Ouyang Xu an ~ ~

z

pailii '111'~ Pan Jinlian $~~ Pen'er gui 1?;:5[1.)fl Pengcheng ji ~:l)ilG~ pin

00\t&ljt

8b

Pin shi zhuan ~±f,\!1 Ping shan /eng yan zrs:LLP~m pingdian 1¥1' !!!ti Pipa ji J]Efi!~t. po ti lilllm:i pudu tff~ pufa jiaoyu tffl':!~fif qi ~ qi cheng zhuan he !Ee:i¥:ftir qian ij!Z qian hou qi zi Ji1if3'tt:;TQian jia shi =f*\i\1

Character List Ruan Xian mJllK; Rudao AJ1l: rujiao {$/M(

Qian Mu f~~ Qian Qianyi f~~~ Qian Tang i~m Qianlong iliZ;jl;1 Qiguai t:d£

Rulin waishi {$#:9~ 5I: Run ji rMJ~

*

Qin qin ~Jl. Qing qing ·~

san bu xiu -=.=t:;J-5 san cong (side) =:=itECJ2:91!D San Iii da fu =:M*x

m ·Jw 51:~~

Qing shi lei lue

Santi Tang shi =:&miff Sanbu daquan .:=.:g~*3': Sanguo Yanyi =.~i~*

frlf,[,,.:EIJ!k:

qing xin yu ying qingmiao fa qiu dao >j(J]t

wffim

tkJm:tH.r PJ

Qiu ]u da guansi Qiu Jun .5:~ Qiuju f)(Jm QixiaSi ~~~ qu E!l3

Qu Fengnu

M~fr.!Z

Qu gongsun )ii0J% Qu Laixun J.i3/Z 1U Qu li E!l3 ift1 Qu Qiubai WHx 8 Qu Shenfu ).i.~;;lg£:i:f!!i ru {$ Ruan Dacheng Ruan Ji ~j[$i

~n*~AX

Sanmin zhuyi .:=..§!..±:* saomei caizi frff I§ ::tTse B semu 8§ senggang fj~ senghui it~ sengzheng it IE Senior Licentiate Yan !l(~:!£ shan ge rlil:!X Shan Tao 0.1m shang di _t1f¥ Shang Jinglan 1i'!l :llt !li!j Shang yi li'!J!$ shanshu ~~ she ffd: shecang ffd: ~ shehui daode pingyi hui ffd:11f31l:i!!~Hm ~ shen jjj$ Shen Deqian tti!!M Shen Jiji ttre£111\' Shen Liao tt~ Shen Qi tt.~ Shen Qiongzhi ttf.i;fi Shen Quan tt~ Shen Quanqi 1:tfrM Shen Shijun 1:t±M: shen yun ffl$ffil. sheng ren zhi dao ~A.z)]t

Sheng Tang ti sheng yuan

~mB

:!£ 7I;

Sheng'an shihua shengshui

7f'itj;jffgg;

~7.)(

Shengxue xinfa

~~,c,,m

shengyuan 1:~ shenquan iii$ ;f:l Shentu Qian $ i'lf ~ Shentu Xiguang $ !'lt:l'i'J:W:

Character List Shenyin yu

O$~lii\

Songzi Niangniang ~rfr!H~ Su Boheng ~18~ Su Che flil

$\ffmc~

*~'*

18&~

Tang shi bie cai ji 18iff5Ja~~ Tang shi gu chui fli.fiffliiZ~ Tang shi gui I@-§ &'Iii Tang shi he xuan 1@-~-;5~ Tang shi ji shi 1@-iff*c~ Tang shi jie !@-iff fW Tang shi kuai 18ifft::ltlf

~EPP:hf:r

yiban -Jill!: yicang :N~ Yichen ~~ yijia zhi yan -~z 13" yili -:£! yili zhi xue :N:f!z~ yimin ~~ yin (appointment procedure) fllj yin (licentious) g yin/yang ~Ill£'!; yinci gijjjj] Ying kui Iii sui ill!i¥1$~ Yingtian !!!.~ Yingying zhuan 1;!;jt~ yinyong xingqing IJ!flli}d~'ll'f yipian [pan ] sansha -it ('ill!: )~¥jl yixue Yang fu si 7.l~ Zhou Bi }ffj31§ Zhou Di }ffJ ill! Zhou Dunyi }ffJ ~ [!Jj Zhou gong (Duke of Zhou) )ffJ0 ZhouJin }ffj~ zhouzhi wence }ffJ rEJ X fffi Zhu Changwen *=:R:X ZhuDi *~ Zhu Heling *=t~m ZhuLu *~ Zhu Shuzhen *~!Jt Zhu Weigong *='1'~0 ZhuXi *~ Zhu Yizun *~~ Zhu Yuanzhang *=7t:f!i!: zhuajin :111\~ Zhuang Shaoguang jif ~ :Yt zhuanye hu W*F Zhufo shizun rulai pus a zunzhe mingcheng gequ mH~ t!t ~ :tm!Z :g: iii~~ o5 fllH!x

ffll Zhuge Liang ~~:rf zhuke ~f-4 zhuyi Zi shuo "¥ ilft zide § 1~ Zirong -T~

.±:*

*

zong zong xu ~I¥ zongcang ~~ zongzu guannian *:IEi