The Chinese State in Ming Society (Critical Asian Scholarship)

The Chinese State in Ming Society The Ming dynasty (1368–1644), a period of commercial expansion and cultural innovation, fashioned the relationship between the present-day state and society in China. In this unique collection of reworked and illustrated essays, one of the leading scholars of Chinese history re-examines this relationship and argues that, contrary to previous scholarship, which emphasized the heavy hand of the state, it was radical responses within society to changes in commercial relations and social networks that led to a stable but dynamic “constitution” during the Ming dynasty. This imaginative reconsideration of existing scholarship also includes two essays first published here and a substantial introduction, and will be fascinating reading for scholars and students interested in China’s development. Timothy Book is Principal of St. John’s College, University of British Colombia.

Critical Asian Scholarship Edited by Mark Selden, Binghamton and Cornell Universities, USA The series is intended to showcase the most important individual contributions to scholarship in Asian Studies. Each of the volumes presents a leading Asian scholar addressing themes that are central to his or her most significant and lasting contribution to Asian studies. The series is committed to the rich variety of research and writing on Asia, and is not restricted to any particular discipline, theoretical approach or geographical expertise. Southeast Asia A testament George McT.Kahin Women and the Family in Chinese History Patricia Buckley Ebrey China Unbound Evolving perspectives on the Chinese past Paul A.Cohen China’s Past, China’s Future Energy, food, environment Vaclav Smil The Chinese State in Ming Society Timothy Brook

The Chinese State in Ming Society Timothy Brook

LONDON AND NEW YORK

First published 2005 by RoutledgeCurzon 2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN Simultaneously published in the USA and Canada by RoutledgeCurzon 270 Madison Ave., New York, NY 10016 RoutledgeCurzon is an imprint of the Taylor & Francis Group This edition published in the Taylor & Francis e-Library, 2005. “ To purchase your own copy of this or any of Taylor & Francis or Routledge’s collection of thousands of eBooks please go to http://www.ebookstore.tandf.co.uk/.” © 2005 Timothy Brook All reserved. No part of this book may be reprinted or reproduced or utilized in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloguing-in-Publication Data A catalog record for this book has been requested ISBN 0-203-31133-7 Master e-book ISBN

ISBN 0-415-34506-5 (hbk) ISBN 0-415-34507-3 (pbk)

Contents List of illustrations

vii

Acknowledgements

x

Abbreviations Introduction: a grave in Nanchang PART I Space

xii 1 16

1 The spatial organization of subcounty administration

17

2 The gazetteer cartography of Ye Chunji

42

PART II Fields

60

3 Taxing polders on the Yangzi Delta

61

4 Growing rice in North Zhili

79

PART III Books 5 Building school libraries in the mid-Ming 6 State censorship and the book trade PART IV Monasteries

97 98 114 131

7 At the margin of public authority: the Ming state and Buddhism

132

8 Buddhism in the Chinese constitution: recording monasteries in North Zhili

150

Conclusions: states of the field

172

Notes

181

Bibliography

213

Index

232

Illustrations Cover A monk-demon is untouched by the arrows the official has ordered his soldiers to fire; illustration from the 1592 Hangzhou edition of the popular story by Luo Guanzhong, Sansui pingyao zhuan (The Three Sui quell the demons’ revolt). One could speculate on the tensions between civil and monastic establishments buried within this picture (see Chapter 8).

Maps

1

Ming China

xii

2

The Lower Yangzi region (Jiangnan)

64

Figures

1.1 Administrative map of Wujiang county, Suzhou

23

1.2 Administrative map of Taicang subprefecture

24

1.3 Map of coastal defences of Jiading county

25

2.1 Map of Huian county

49

2.2 Cartographic legend on the map of Huian county

50

2.3 Map of the Huian county seat

51

2.4 Map of Huian Township 2

52

2.5 Map of Zengcheng county, Guangdong

56

2.6 Map of the waterway network of Tongxiang county, Zhejiang

57

3.1 Small polders along the Yangzi River

63

3.2 The division of Changshu county by sector

71

3.3 Sector 34 of Changshu county

72

4.1 The paddy fields watered by Rice Canal

93

7.1 The Lamaist monastic complex on Wutai Mountain

140

8.1 Map of the seat of Hejian prefecture

167

Tables

1.1 Registered population of Anqui county, Shandong

21

1.2 Households per subcounty unit in five prefectures

22

1.3 Distribution of subcounty administrative units in a standard structure

26

1.4 Distribution of subcounty administrative units in an elaborated structure

26

1.5 The subcounty system in Huian county

27

1.6 Hundreds (li) per county, by province

34

1.7 Subcounty administrative units in the Ming

38

4.1 Number of counties and subprefectures for which rice is first reported as growing in North Zhili

84

4.2 Percentage of rice in Ming grain and hay taxes

88

5.1 Core texts in Ming school libraries

105

Acknowledgements To Mark Selden belongs the credit of helping me discover the book that lurked within these essays and encouraging me to fashion the parts into a coherent whole; the volume would not exist but for his prompting. I would like to name two others who have contributed to the thinking that went into this project. Ever since our graduate student days together, Bin Wong has read and criticized my work with the precision of a scholar and the patience of a friend, and did so once again on this occasion. Michael Szonyi has been a more recent influence, but his Practicing Kinship in particular helped me rethink some of the issues this volume addresses. For their more practical contributions, I wish to thank Trish McAlaster for drawing the maps, Nick Hawkins for editing the manuscript, and my research assistant Kevin Lu for being so patient and enthusiastic through the long process of getting from essays to book.

Abbreviations jr.

year in which juren degree was conferred

js.

year in which jinshi degree was conferred

r.

reigned

Map 1 Ming China

Introduction A grave in Nanchang By the time the dossier reached the Hongzhi emperor’s desk on 29 November 1499, the case had become complicated—which is what it had to be in order to get there. Wang Zhen owned a piece of land in the hills outside the city of Nanchang, the capital of Jiangxi province. In this hilly region south of the Yangzi River, population was dense, land scarce, and the locals often on the move elsewhere looking for work or land. “The hills are many and the fields few,” as a Nanchang county author noted by way of explaining why the local people were so lean.1 Even the hills, used for graves rather than fields, were at a premium. The most coveted bits of upland topography were those spots where professional geomancers judged that the lines of energy (qi) streaming through the landscape converged propitiously. Bury an ancestor within such an energy field and the deceased’s spirit will radiate fortune to his descendants. Jiangxi lineages competed for the best tomb sites and resorted to tricks and violence in their struggle to improve their fortunes at others’ expense. Grave land feuds were endemic to the province through the Ming and Qing dynasties. The case that went up to the Hongzhi emperor started because Zhang Yingqi buried a body on Wang Zhen’s grave land without his permission. Zhang was a student on stipend at the Nanchang government school. An aspirant to higher elite status, he was positioned to rise into the upper gentry should fortune, education, and wealth continue to conspire in his favour. Whom Zhang buried on Wang’s land, and why he had chosen to bury that person on land that was not his, are not stated in the surviving case summary that appears in the court digest, the Veritable Records of the Hongzhi Reign (Xiaozong shilu).2 Zhang appears not to have been driven by the usual goad of poverty. Geomancers must have declared this bit of hill as top grade for burial purposes, a place where Zhang might entomb his ancestor so splendidly that good fortune could not help but rain down upon the living, including himself. Wang Zhen, the owner of the land, was not a student, nor did he possess any token of official status. Yet even a commoner could take his case to court, if he were willing to deal with the exactions and interferences of the lesser functionaries standing between him and the presiding judge. This is what Wang did, filing a lawsuit with the prefectural government. Given the high costs of pursuing a case through the court and the impossibility of controlling the outcome, only the truly desperate surrendered their conflicts to official arbitration. But Nanchang people seem to have breathed a different judicial culture. The compiler of the earliest Ming-period gazetteer for the prefecture, produced in 1378 in response to the request from the Hongwu emperor (r. 1368–98) that local gazetteers be submitted to the court, praised the ardent passion for virtue and diligence animating the hearts of the local people. The chronicler also noted, though, that these ardent feelings—animated by the same energy (qi) that the geomancers detected in the land—could go to excess, breeding intolerance and propelling people into lawsuits. This was not news to the emperor, who in his final instructions to the people in 1398 singled out Jiangxi natives for being “prone to litigation” and complained that they “cannot endure even minor matters, and go directly to the capital to bring suits.”3 Nor

The chinese state in ming society

2

was a taste for lawsuits in Jiangxi uniquely an early-Ming reputation. The compiler of another gazetteer within the prefecture notes that during the Yuan dynasty, Nanchang people had a reputation for “relishing fights, losing their tempers, and enjoying lawsuits.” The subsequent edition of the prefectural gazetteer, in production at the time Wang Zhen’s case was being heard, observed that over-taxation, poverty, and land scarcity were making judicial tempers that much worse.4 Nanchang’s reputation for fractiousness did not dim with time. A gazetteer editor in 1565 complained that local people “enjoy instigating idiotic lawsuits and giving vent to their passions, all for the sake of small profit.”5 A few decades later, another Jiangxi commentator summarized Nanchang people as “avid in work and stingy in giving, indifferent to duty and happy to quarrel; cunning and glib, litigious and libellous.”6 The prefectural judge agreed to hear Wang Zhen’s case and decided in his favour. Presumably he ordered Zhang Yingqi to remove the bones he had buried there. Zhang, however, decided to fight back. He turned to a fellow student at the Nanchang school, Liu Ximeng, who had something Zhang needed: a connection to a higher authority. Liu had got a job tutoring the son of an assistant provincial surveillance commissioner, Wu Qiong (js. 1469). Liu had managed to ingratiate himself with the Wu family by handing out presents, thereby positioning himself as a small-time power broker between Wu Qiong and anyone who might want his help. The connection suited Zhang, for an assistant surveillance commissioner outranked a prefectural judge. Money changed hands, and a word from Liu in someone’s ear in the Wu household succeeded in getting Wang’s case against Zhang overturned. Zhang was unwilling to rest content with this victory. Presumably the disputed corpse was still in the ground and he was still vulnerable to counterattack from Wang. He chose to go on the offensive and put together a lawsuit against Wang. As the prefectural judge had already judged against him, Zhang found another patron, Assistant Education Intendant Su Kui (js. 1487). Su approved admissions to government schools in Jiangxi province and regularly tested the students who were admitted, responsibilities that placed Zhang in the position of being able to communicate with him legitimately. An issue such as this, however, lay outside Su’s proper jurisdiction. Su had a reputation for refusing to act on behalf of private interests, which makes his willingness to help Zhang, allegedly after receiving a bribe, puzzling. It is possible that none of this is true, or that Zhang misled Su into acting on his behalf without full knowledge of what was at stake. Wang Zhen meanwhile, seeing that Zhang had support elsewhere in the Nanchang bureaucracy, looked for a connection to another reservoir of state power, the eunuch establishment. How he got to Grand Defender Dong Rang, a eunuch of the imperial household whom the emperor had sent out to supervise regional security, is not known.7 Presumably the right sum of money could open any doors, so long as one knew which corridors to travel. Wang presented his case to Dong, and Dong obliged him by having both Zhang and Liu thrown in prison, where torturers could persuade them to withdraw Zhang’s claims. Until this point, there was nothing unusual about what the players in this little drama did. Two people caught in a struggle over land had looked up into the lower levels of the state bureaucracy in search of connections to aid them in their struggle, and mid-level provincial officials had been happy to oblige. Money had changed hands on both sides, and the justice system turned into a network for channelling bribes and influence from competing nodes of authority, not for resolving disputes. By acting as they did, the two

Introduction

3

litigants were conforming to the modus operandi of the political system in which they found themselves. This is how the Ming state—which may be defined as a coercive system of territorial authority and communication moving information, resources, and personnel in regular ways designed to ensure the wealth and security of the dynasty— worked. The apex of the communications system was the throne, and the channels along which information, resources, and personnel moved were the lines of authority and reporting through which percolated the throne’s capacity to act via the imperial household, the bureaucracy, and the army. By seeking influence with people embedded in the state’s administrative networks, Zhang Yingqi and Wang Zhen were simply responding to the opportunities available to them on the landscape of power. Their intention was not to send their dispute all the way up to the throne to resolve, of course. It was to send up a counter-flow of influence and information into the lower reaches of the state system, a modest capillary response to the percolation of authority downward, with the expectation of stemming (in Wang’s case) or directing (in Zhang’s case) its flow. The turning point in this communicative process—when a grave dispute between minor families makes its way up through the system to the emperor’s attention and leaves traces in the dynasty’s written records—came when Grand Defender Dong handed over Zhang Yingqi and Liu Ximeng to keen jailers. Dong intended only to intimidate the two students into backing off, but the torturers went too far. The hapless two readily revealed that they had been bribing state officials to back their side of the lawsuit. Once this revelation came out, what had been a local property dispute, best handled in that context, turned into a bureaucratic crime that had to travel up officialdom’s hierarchy to Beijing, first to the censorate, then to the Ministry of Justice, and finally to the throne. Dong Rang had gone too far, and now an emperor was looking down at the situation in Nanchang and demanding that the ministry investigate. The modest capillary action of rational bribery from below (what could be more rational than to appeal effectively to those who could produce a favourable decision?) overwhelmed the percolating gravity of state authority from above, in part because the information it carried was of a sort that interested this particular emperor. Hongzhi (r. 1488–1505), as it turns out, was passionate on the subject of corruption: his reign’s records are filled, more than those of any other Ming emperor (with the exception, as in all things, of the foimder, his great-great-great-greatgrandfather, Emperor Hongwu), with dismissals for official malfeasance and incompetence. The investigation uncovered that Eunuch Dong had already marked Su Kui as his enemy before Wang Zhen ever came to him for assistance. Indeed, the rivalry between them gave the bribes their traction. Dong felt that Su had insulted him over another matter, and so agreed to take Wang Zhen’s side as a way to get at Su. Did Su actually agree to support Zhang Yingqi? Or did Dong embroider the formal connection between them into something that made the commissioner look as though he were improperly involved in a land dispute, when he was doing nothing of the sort? It is impossible to tell. What is clear is that the dispute had spiralled out of the control of the two men who started it, turning into a case that had nothing to do with who had the right to bury his ancestors where, and everything to do with the political tension between the eunuch and civil bureaucracies in one provincial capital. Su returned Dong’s dislike, and was not alone in doing so. Dong’s highhanded activities on behalf of the imperial household inspired other officials elsewhere in the bureaucracy to petition for the man’s removal,

The chinese state in ming society

4

first to the Hongzhi emperor while he was alive, and later to his successor, though never to any effect.8 Dong was in the stronger position and managed to get Su thrown in prison on a corruption charge, probably in connection with a different matter. Students at the Nanchang school were so offended by the eunuch’s attack that a hundred of them stormed the jail and freed their superior. Su was exonerated and later promoted (everyone assumed he was innocent, though that is what they needed to believe), yet Dong was left untouched.9 Matters having gotten so far out of hand, the emperor could have dispensed terrible judgments on those involved, yet he chose not to. His hand may have been stayed by the need to protect his exposed eunuch servant, who after all was supposed to be in Nanchang to look out for imperial (i.e., his) interests. Or he may have wanted to avoid siding with one arm of the state over another, so as to keep his eunuchs and bureaucrats in dynamic tension with each other. Hongzhi explained his decision not to take harsh action by reasoning that no actual damage had been inflicted on any of the parties. He reprimanded Dong Rang and Su Kui for agreeing to adjudicate lawsuits they were not entitled by their positions to entertain, and he reprimanded Su Kui and Wu Qiong for taking payments. The burden of his judgment fell away from his officials, however, landing most heavily on the two students who started the affair. Zhang Yingqi and Liu Ximeng were not beaten or fined or sent into exile, which could have been their fate at the hands of an angrier emperor. Instead, they were stripped of their studentships and stipends and banned from ever again trying to climb that ladder of success—punishment enough in a status environment as competitive as mid-Ming China’s. The storm in Nanchang’s teacup happened to catch an emperor’s eye, and that it did suits my purpose, which is to frame the eight studies in this book by inquiring into the presence and power of the Chinese state in Ming society. The dramatic intervention of an emperor could be taken as a vivid example of the state’s capacity to control society: a demonstration that the Ming court could reach all the way down to the bottom of the realm and pull apart two men fighting over a grave. This is how Ming historians would once have told this story, when the emperor fetish that has long lurked around the field of Ming studies was still strong.10 The fetish is one that Ming historians inherited in the first instance from Ming officials, caught as they were within the operations of a public rhetoric that obliged them to refer to an emperor by the correct euphemism (Sagely Founder, one of Hongwu’s titles, would have done Confucius proud) and to glow when doing so. Our emperor fetish is also the product of a long historiographical tradition in Europe going back at least to the ghost of Georg Hegel, who could not conceive of China other than as a realm in which only the emperor had full individuality and every other person was his slave.11 If the remarkable man who founded the dynasty in 1368 had been the only emperor of the Ming, the fetish might well be justified, given his extraordinary energy and the mythic scale on which he created and destroyed. But he wasn’t, and it isn’t. Another way of understanding the Nanchang burial case is to remain within the state frame but reverse the equation between the emperor’s presence and the operation of the bureaucracy and regard what Hongzhi did as a momentary disruption in the routine functioning of state administration, the exception that breaks the rule of flawless state control rather than the one that proves it. For rarely did an emperor intervene in what went on in the bureaucratic structure beneath him; even less could he see into the social

Introduction

5

networks stretching out beyond the bottommost rungs of that system. As I shall note at several points in this book, an emperor had within his gaze only what his officials brought to his attention. Hongzhi put himself in the way of more information than most of his line. One of his first acts after his enthronement was to dismiss almost the entire staffs of the Ministries of War, Justice, and Personnel12 as a sign that he would not tolerate the corrupt and incompetent. This move prompted the zealous and the ambitious to forward more information to him about what was going on in the field bureaucracy than was usually the case. Even so, the range and depth of his knowledge was limited. As Hongzhi himself admitted in January 1499 in the edict of penitence he issued after the Qingning Palace within the nine-walled Forbidden City burned down, “I live deep within the Nine Walls, and though I stretch my thoughts over the entire realm, there are places my ears and eyes do not reach and where my grace has not been manifested.”13 This was not simply a matter of scale, however. The regular bureaucracy and the parallel intelligence operations of his eunuchs directed information his way, but both could block or distort information as well as transmit it. Dong Rang’s activities would never have come to his eyes or ears through the eunuch channel, for instance, nor were the alleged bribes that Wu Qiong and Su Kui accepted knowledge that the civil officials wanted relayed to the emperor if they could help it. An emperor’s communicative links to society were few, and easily closed when all his subordinates agreed that they should be. But this is not the crux of the problem of reading Ming history from the state side. That problem resides rather in our conceptualization of the Chinese state. Ming people knew they were subject to the authority of the emperor, but that is not how they experienced the state. The state exerted its presence in Ming society less because of what the man at the top did or wanted done—it must have shocked the litigants to have the emperor weigh in on their case—than because of interventions of state representatives further down the communications system. Even these interventions were exceptional, for most people knew the state only by distant proxy in the course of dealing with the systems through which their affairs were administered; specifically, the taxation, education, justice, and military systems that made the state present in society as more than an abstraction. The taxation system, with its regular grain taxes and labour levies, was the common context in which people interacted with the Ming state and its officers, which is why most of the studies in this book address issues arising from the taxation nexus between state and society. The education system, regularized as a bureaucratic operation in 1436 when education intendants such as Su Kui were appointed in the provinces, affected only a minority of young men, though the aspiration to gain a place at a government school was socially pervasive. The justice system held universal sway over the emperor’s subjects, ready to snare anyone who contravened his laws, yet the number of suspects and plaintiffs who went before a magistrate could not have been great. Most people managed to live their lives without getting tangled in the law. Least likely to touch the lives of ordinary people was the military system, which did not conscript civilians but drew its soldiers from hereditary military households. A military designation was a fiscal rather than martial category, though, and those who could escape it and move into the commoner population did so. The most likely way for commoners to encounter soldiers was when the latter were mobilized to deal with banditry or other unrest, which is why a eunuch grand defender such as Dong Rang was assigned to the unsettled province of hilly

The chinese state in ming society

6

Jiangxi. These agents of the imperial household were farther removed from the people than the regular bureaucratic systems, and more mysterious for their power because of the independence from bureaucratic oversight. The imperial household took an interest in military affairs early in the fifteenth century, when the Yongle emperor put eunuchs on military assignment. Best known from his reign is the Muslim eunuch Zheng He, who commanded the imperial fleets that sailed to the Indian Ocean. Military eunuchs were useful to an emperor, enabling him to keep an eye on security situations without having to see everything from his bureaucrats’ point of view. They were also an annoyance to those same bureaucrats, who distrusted them for not being accountable to regular standards and resented the relatively free hand they seemed to enjoy extracting resources on behalf of the imperial household—and themselves, of course. A year before the Nanchang grave case, the Ministry of War had gently suggested that Hongzhi reduce the number of full and associate grand defenders in the realm, arguing that their operations were a heavy financial burden on the common people. The ministry was tactful enough to blame Hongzhi’s predecessor, the Chenghua emperor (r. 1465–87), for escalating the scale of eunuch surveillance. Hongzhi turned the request down.14 When, just a few short weeks after the affair with Dong Rang, a county magistrate in the northeast indicted a eunuch grand defender there along with two military officials for misconduct that included the indiscriminate slaughter of border people, the emperor would not support his demand for a full investigation. Let the censor already in the region look into the matter, Hongzhi replied. A year later, he dismissed all charges.15 The eunuchs were simply too important to his strategy of control to leave dangling for civil bureaucrats to attack, however badly they behaved. The tension between them and the regular officials was a feature of the Ming constitution, and one that Ming emperors favoured as a device to retain some control over decision-making and policy implementation within the regular bureaucratic systems. The Nanchang burial case managed to entangle all these systems. Most visibly in play was the justice system, which Wang Zhen elected to engage through the proper channel, and which Zhang Yingqi sought to subvert by bribing the prefectural judge’s superior. Also prominent in this story is the education system, which provided Zhang with his access via Liu Ximeng to the assistant surveillance commissioner, and then with his superior in the educational hierarchy, Su Kui. The military system was not directly involved, though Dong Rang’s appointment to oversee regional security empowered him to weigh in and subvert the procedures for settling land cases. Finally although taxes did not come up in this case, the taxation system framed what was going on. Wang and Zhang were fighting over a plot of land that had not been put under the plough and so “brought onto the registers,” as the assignment of tax liability was phrased.16 It was land on which no tax had to be paid. Zhang’s eagerness to push Wang off it confirms this, as he would not have wanted to take over land for burial purposes were he obliged to pay an agricultural tax on it. The tactic for taking control of someone else’s taxable land was discreet encroachment, not forceful seizure, lest an official take note and transfer the tax burden. Zhang was not being discreet. He was “stealing a burial site,” as the court record phrases it in the language of brigandage. The court historian who summarized the dispute for the Veritable Records does not disentangle these systems, for they flow together in the same channel of imperial authority. He begins his account with Liu Ximeng’s relationship to the assistant

Introduction

7

surveillance commissioner, since this is the context that turned a land dispute, otherwise of no concern to the court, into a corruption case in which the emperor had an interest. This way of telling the story could be used to narrate Ming imperial rule as an autocracy in which the ruler served as the fulcrum on which the state had to pivot in order to function. This version would confirm the paradigm of despotism, the origins of which go back to Montesquieu and Hegel and the effects of which, thanks to Karl Wittfogel’s recycling of the trope of “Oriental despotism,” shaped the field of Ming studies as it came into being during the Cold War.17 Given the number of local players who became involved, however, the incident could sustain a different reading, one which tells the story not as the re-enactment of the paradigmatic relationship between an absolute ruler and his absolute subjects—between jun and chen, in the language of Hongwu18—but as the working through of some of the possibilities and constraints of bureaucratic administration. This is how the field of Ming studies began telling Ming stories in the 1970s, digging out from under the dynasty’s reputation as Oriental despotism’s ideal type by building up knowledge of how Ming government worked in practice.19 Consider, though, another reading, one which begins where I began telling this story: not with student Liu Ximeng’s connection with official Wu Qiong, but with the conflict between landowner Wang Zhen and burier Zhang Yingqi. These were the people whose actions set the story going, not the provincial officials, and certainly not the emperor. How each chose to act depended in part on the state systems accessible to him. Like Chinese in almost any period, both must have kept a weather eye on the state. But the conflict arose and took its shape because of the social networks to which Wang and Zhang were tied. Only when fellow student Liu comes into the story do the available state systems begin to direct the flow of events, bringing Wang and Zhang into the emperor’s view and turning a local land-dispute story into a national corruption narrative. State systems were important in this history, but the social was prior. Accordingly, I would like to propose that what was distinctive about Ming China was less its state than its society, since it was within society that the effects of demographic growth, expanding communication networks, rapid commercialization, and new critical thinking were most keenly felt. After the first highly interventionist Ming reign, the state more or less followed in the wake of these shifts, attempting to manage a realm of unprecedented complexity rather than remake what it found. Even when an activist emperor was able to impose organizational frameworks and limitations on local society, his agents could sustain them only by fitting them to the social networks that predated their imposition. Drawing ward boundaries and shepherding communities into lijia units as elements of a state-making program, as we will see the founding emperor doing in the first chapter of this book, were ambitious in what they were designed to achieve, yet underneath these state-directed interventions runs a very different process that political theorist Roberto Unger has coined “society-making.” “Society-making” denotes the process by which people interact with each other through structured networks and make the conditions of their social exis-tence on the basis of the resources available to them by virtue of their social position. “These resources include governmental power, economic capital, technical expertise, and prestigious ideals or the forms of argument that claim to show implications of these ideals.”20 The state may set or seek to influence how these resources become available and how they may legally be used, but the actual forms of the processes through which

The chinese state in ming society

8

the state intervenes to shape ranks and roles depends on the practices current within everyday life, regardless of its own legitimacy and authority claims. My purpose in invoking “society-making” is not to marginalize state effects, but to insist on an analytical distinction from the more familiar concept of “state-making,” the term we use when the state mobilizes resources to build up its administrative capacities and enhance its security. Unger’s purpose was to field an argument in favour of a radical reorganization of power to enable citizens to constitute a democracy not in thrall to the state: to make society prior to the state. He writes from within the European constitutional tradition dating at least to the eighteenth century, in which the rules of political life guaranteed the security of the person and his social relations from state interference. The foundational assumption of an eighteenth-century constitutional theorist such as William Blackstone was that the security and liberty of the individual were secured against state infringement by the law, and that no state representative had the authority to suspend that liberty except “when the state is in real danger”—with the strong proviso that “it is not left to the executive power to determine when the danger of the state is so great, as to render this measure expedient.”21 With these declarations, published in 1765, Blackstone celebrated the success with which the English had worked out their constitutional arrangements, which he regarded as a unique historical outcome (and which allowed him to direct his condescension not at some distant inadequate other, such as the Chinese, but at a group closer to home, the French). Blackstone’s way of conceiving of the just limitations on the state to protect the body and preserve the personal liberty of the individual was not how Ming law understood the relationship between the individual (and his social relations) and the state, nor how Chinese saw themselves within the state context familiar to them. This difference in the structures of legal meaning does not entitle us, however, to go on and argue that Chinese political culture was indifferent to individuals, or that society-making was alien to the Chinese constitutions, or that China was a state stronger than society. These ideas were popularized by Karl Wittfogel, who adopted the notion from Paul Milukov’s study of imperial Russia. Milukov published his book in Leipzig in 1898 at a time when pressure from elements of the Russian elite to dismantle that regime was growing and the impulse to frame the struggle in stark ideological language appealing.22 The Milukovian imaginary of a state stronger than society expressed the repressive character of late tsarist rule, and it was taken up by Cold War sociology to parse the totalitarian regimes of the mid-twentieth century. Whatever sense the argument made at the time, it says more about the gross abuse of the principles of law in the context for which it was formulated than it could possibly contribute to an analysis of the relationship between state and social power in Ming China. Rather than use as a baseline the concentration of state power Wittfogel feared in his own time and attributed to Ming China, a more productive theoretical approach grasps the fact that state-and society-making occurred in overlapping and interactive ways. A generation ago, Michael Mann argued in favour of getting away from the habit of reifying state and society as unitary and equal entities polarized in theoretical combat with each other. He advocated instead that we approach society as “a diversity of intersecting networks of social interaction,” of which the state was but one “interaction network” among many, and one to which fewer social actors are tied. Mann’s approach endows society with greater conceptual breadth and analytical power than the state,

Introduction

9

integrating the state more fully with society than earlier theorists had done. Though tangled in regrettable Eurocentrisms,23 he points us in the same direction as Unger’s shift from state-making to society-making as a more complete basis for building social theory. This parallels the theoretical moves I have found myself making over the course of interpreting the rich record of state and social life in Ming China. I have found state networks projecting imperial authority more deeply into society than was the case in Renaissance Europe, yet I have also detected a strong counter-flow of influence working upward in capillary fashion from social networks below that placed the state in a posture more reactive than formative in making Ming China what it was. This perspective was still inchoate in 1984 when I wrote the essay on subcounty administration that, in revised form, opens this volume. In the original version, my attention was divided between tracking the power of society to channel state intervention, and the capacity of the state to remake society, unable to decide between them and neglecting the extent to which both were occurring simultaneously and interactively The same duality of analysis ran through the original version of Chapter 4 on the spread of rice agriculture into north China, written in 1980, in which I found that the work of active dissemination promoted by state representatives succeeded only when linked to existing social, consumption, and economic networks—and failed when it did not. The simultaneity of state-making and society-making perspectives is most explicit in an essay I wrote in 1992 on the resilience of Buddhist monastic institutions in the face of state suppression, for which reason it appears here, as Chapter 7, largely unaltered from the original version. I have grouped the eight chapters in this book into four pairs based on the thing, activity, or social group over which state and social networks competed to assert control. In each case, it is tempting to view state-society interactions from the privileged position of state systems and state actors—to see society as reactive to the state—which is how most of the sources represent them. The import of each of the chapters, however, is to show the extent to which social networks obliged the state to adapt to local practices and institutions: in effect, domesticating the state into becoming yet one more resource which local actors could exploit in their competition for power. The first pair, “Space,” begins with a study of subcounty administrative systems the new dynasty imposed at the beginning. These systems were ambi-tiously designed to locate every community and household in Ming China and make them components of larger entities, yet they largely followed the boundaries that social and economic practices had already drawn. In Chapter 2, similarly, the county survey maps that Magistrate Ye Chunji produced could be taken as demonstrating the state’s capacity to configure space and thereby profitably control local social systems; yet they can also be seen as attempts to bring state vision into line with the socially visible. Ye drew his maps in relation to the larger process known as the Single Whip reforms, by which the state was obliged to revolutionize its fiscal systems in order to adapt to an economy that silver and commodity production had transformed away from the rural simplicity the founding emperor imagined when he imposed his systems. In both these chapters, the state’s organization of territory assisted in the formation of a distinctive Ming hegemony, arising from negotiations between state representatives and those for whom local social practices determined how they organized their lives.

The chinese state in ming society

10

If this all seems sensible, it wasn’t necessarily so to our Hegelian forebears, for whom China suffered a terrifying absence of private property and a vulnerability of individuals to arbitrary state power. Hegel is not the only ghost in the house. From a different perspective, Emile Durkheim worried most of a century later that the structure of rural life resulted not from the state machine, but from a mechanical spontaneity of communal organization that denied the independence and intimacy that for him constituted individual life.24 This fear of the local community was, however, a variation on the same anxiety that moved Hegel and Wittfogel: that the individual, and by extension his capacity to act autonomously, in China was profoundly compromised, whether by society or the state. In fact, the dialogue between state and society and the realm of possible social action that this dialogue in part defined, was far more complex and fluid than earlier observers of China were able to discern. The second pair of chapters, “Fields,” continues to consider the problem of the accessibility of land to state scrutiny, but in relation to the growing of rice. The two chapters are concerned with the dialogic interplay between state and society in two very different settings: the registration of rice-growing polders as fiscal units on the Yangzi Delta in Chapter 3, and the introduction of rice agriculture on the North China Plain in Chapter 4. Rice paddies posed particular challenges on state officials concerned to collect taxes but also to direct investment in agricultural infrastructure. The two cases diverge sharply in terms of how state systems interacted with local economies and societies between south and north. In the south, where conditions naturally favoured riziculture, the state’s challenge was to work effectively with local community leaders, who were better suited to managing local needs and resources in densely structured economies than were outside officials. In the north, the natural environment could sustain rice agriculture only with investments on a scale that was beyond the means of most communities. As Philip Huang has phrased the contrast, in Jiangnan the elite could intervene to ensure its access to levies on land and labour, even to take leadership in crisis situations when elites failed to act. These interventions produced a “frequent three-way interaction among state, elite, and peasant.” In north China, on the other hand, waterworks maintenance was on a scale that favoured state management, since converting dry land to paddy fields was too big for individual households or villages to undertake. Those were the projects that in the Yangzi delta drew the elite, the peasant, and the state into necessarily complex and changing relationships of a sort that did not exist in North China.25 Awareness of this complexity was the mark of a capable bureaucrat, and the means by which the state might not lose all contact with society. The third pair of chapters, “Books,” examines the interaction of state and non-state institutions in terms of another portable commodity, a collectible manufacture rather than a consumable crop, one accessed not primarily through ownership but through education, inclination, and opportunity. The wide availability of books meant that readers could read what they chose, not merely what the state put on the curricula of those seeking public office. At moments of anxiety, the Chinese state has worried about this promiscuity. It could respond by going into publishing and distributing the texts it wanted people to read, or by removing those it disliked from circulation. The Ming did both, and came up

Introduction

11

against social networks that enabled it to distribute certain titles and ban others without, however, ever gaining full control of books—or, for that matter, even conceiving of this as a task appropriate to what the state should regularly do. Chapter 5, which looks at the building of libraries at government schools in the midMing, shows that both state and social networks produced and distributed the books that filled the cabinets in these libraries, but that the latter outdid the former in volume and persistence. The libraries ended up owning texts that were not those the state wanted its students to read, as well as those that it did. Chapter 6, which examines the other side of book control, censorship, deals more with the Qing than the Ming because of the richer documentation for the later period. Two extended examples from the Ming are included, and these indicate, to me at least, that the Ming was not an active censorship state. If the Qing became one, it did so intermittently as a result of particular anxieties of the eighteenth-century emperors. Their somewhat misnamed “literary inquisition” was both driven and confounded, however, by the activities of commercial publishers, whose output was too prolific for them to counter. Officials exploited their bureaucratic networks to destroy books, but the capillary capacities of the networks in which book publishers and collectors operated ensured a counter-influence that saved many of the marked titles. A similar dynamic between state and society operates in the next pair of studies, “Monasteries.” Chapter 7 looks at the stringent regulations that the Hongwu emperor introduced in the second half of his reign to reduce the size of the Buddhist establishment and its power in local society. Much of the new legislation was immediately jettisoned after his death in the face of locally rooted opposition; other aspects were later reworked through subsequent renegotiations between Buddhist monks and their patrons on the one hand and the state and its local representatives on the other. The actual unfolding of the suppression provides a good opportunity to interrogate the social processes through which people, whether officials or not, accumulated and dispersed power in the administrative presence of the state. The draconian controls this emperor imposed could be cited as evidence of the state’s capacity for despotism, yet it was a despotism shortlived. Soon enough, the state had to negotiate with local interests concerned to protect institutions and practices salient to their own strategies of local control. Public authority was not the state’s alone to create or command in the face of dense social networks, such as those which bound many Ming monasteries to patrons of wealth and power. The state was not opposed to the presence of Buddhist institutions in Ming society so long as they did not serve to facilitate the emergence of a competing power base. Conservative state representatives often suspected them of this dangerous capacity, yet as Chapter 8 on the recording of Buddhist monasteries in local gazetteers shows, the legal recognition that Buddhism and its institutions enjoyed constrained Buddhism’s moral and ideological foes from doing anything more radical than condemning or demoting them in print. As I argue in that chapter, Buddhism was sufficiently protected by both custom and constitution that earnest Confucians in the local gentry could not call for its eradication and be listened to by anyone but themselves. No matter how strongly the gentry played the state’s hand, community boundaries, landownership, consumption patterns, literacy, commercial production and exchange, and religious devotion—among other factors—determined local outcomes.

The chinese state in ming society

12

In all these chapters, there appear—if not directly in view of the state, then at the edge of its vision, variously in the form of amalgamated monasteries that did not close, banned books that survived, or crops other than rice being planted—spaces for action the state could not colonize. Seen from below, the state’s dominance of Ming society was never unitary, nor its reach unobstructed. State systems and social forces might be in opposition when revenue or security was at risk, but this opposition does not provide a full analysis of the Ming. Better that we begin with this simple pair of assumptions: that social networks shaped the ways in which most people lived their lives, and that state systems could influence these networks but not, other than at exceptional moments, remake them. Active resistance was an option at both levels, whether to stymie state imperatives or to disrupt social networks, as the frequent outbreaks of low-level violence in the Ming attests,26 but the mutual interaction of these networks of authority was more diffuse: percolation from above with no guarantee of where the moisture of state influence might reach, and capillary action from below with potentially no limit on the capacity of society to remake the state. When Emperor Hongzhi looked down through the state systems beneath him at the fight over grave land in Nanchang, he addressed it as a series of errors that participants in state systems had made. As arbiter of those systems, he had to intervene to get them operating properly again. Where Zhang Yingqi’s ancestor got buried was not the emperor’s particular concern: let propriety follow property. Zhang saw it otherwise: abuse of state systems was not of interest except insofar as he could turn it to his advantage; what he cared about was the property in which he had buried his ancestor’s bones, and the social alignments or cohesions that that burial had served to express and reinforce. Burial was formative in the social processes of acknowledging relations to ancestors and therefore relations to living people, so getting the bones in the right piece of ground was critical to the process of individual and household reproduction. But the ground mattered as much as the bones in the everyday life of ordinary people, and that was what Wang Zhen cared about. The grave was on his property, and property was what the Chinese state, for all the exciting flamboyance and peculiarity of imperial interference, was committed to protect. The Ming regime, like any other, was rooted in the preservation of the inequalities of property and the hierarchies of social status among people who used property rights to reconsolidate their positions with every shift of power. The state nodded to the importance of the proprieties that ritual acted out, but for the most part it addressed property rights primarily, on the understanding that the satisfaction of ritual claims would follow. Once the Hongzhi emperor handed out his punishments, Zhang Yingqi’s attempt to steal Wang Zhen’s land disappeared from his sight, never to trouble him again. The conflict between these two men was not necessarily solved, however. Zhang may have carried on his battle by burrowing along other networks and dragging to his support other institutions to get him at last what he wanted. Wang for his part may have worried that he could not rest content with the emperor’s judgment, and that Zhang would still come back at him, obliging him to search among the same local networks and institutions for protection against the next attack. We have no way of knowing. They did not come to the attention of an emperor or a court historian again, nor did they rise to the notice of the local compilers of Nanchang gazetteers.

Introduction

13

Their heirs might still be at it, for all this Ming historian knows. Enormous political and economic changes have intervened in the last five centuries to alter the patterns of state-society relations Chinese find familiar. Yet it is not impossible that the conditions that led then to an eruption of intracommunity conflict—scarce resources, intense social competition, and a differentially responsive political system—are still present, albeit within vastly different political and economic circumstances, and that the approaches that the actors in this little drama took in 1499 might not be unfamiliar to Chinese caught in conflicts today. Within the particular mesh of state-in-society in China, their ghosts are still in the house. Best that ours not be, lest theirs invite ours in and their world begin to look suspiciously like our own. A note on sources The reports that court historians included in the Veritable Records, important as they are for glimpsing the concerns and controversies that swung into the view of the court, reveal only what an emperor heard, said, or chose to do. These voluminous annals are thus one chronicle in which the interventions of the state at the local level can be detected, though often what they record is limited to what state administrators wanted to do or needed to know. As such, these annals are not the best repository of primary materials for exploring how Ming people experienced the presence of the state in their local communities. They see the social from too great a distance, and mostly from the state’s point of view. Still, without the central state keeping occasional track of such matters as, for example, requests for copies of books published by the court printers (as we will note in Chapter 5), we would have no record that such request even went up to Beijing. More detailed, and more useful for analyzing most aspects of state-society relations in the period, are local gazetteers. These compendia of geographical, administrative, and biographical information were mostly produced under the supervision of the local magistrate. Though not official state publications, county and prefectural gazetteers were expected to chronicle the presence of officially mandated state institutions and those who interacted with them, and state representatives were the final arbiters of what their compilers included. As such, gazetteers reveal much about local reactions to and implementation of central policies coming down from the court. They also provide windows onto social and cultural changes flowing through society at a level beneath the operation of state systems. With the limited exception of Chapter 6, local gazetteers constitute the source base for the research in this volume. I think of gazetteers as particularly a Ming source. Although the genre did not begin with the Ming,27 this is the period when it was busily in formation, with the would-be compiler reading any and every gazetteer he could lay his hands on in order to refine and improve the principles of organization governing his own work.28 This process continued in the Qing, as we will see in Chapter 8 with the controversies surrounding the county gazetteer Lu Longqi produced in 1686. If the genre feels like a Ming production, it is also because this is the period when gazetteers began to be published on a scale that enabled copies to survive in large numbers. Close to 7,000 Ming and Qing gazetteers are extant. Most of these are of administratively defined units such as counties or prefectures, though close to a thousand were produced for institutional sites such as monasteries and

The chinese state in ming society

14

topographical sites such as mountains. Roughly 300 are cited in this book; these appear in a separate bibliography. Gazetteers are ideal for writing local history, which is how I use them in Chapter 2 to write a brief history of Huian county, Fujian. The consistency of format, however, gives their data a comparability across place and time that permits their use to detect countrywide trends, which is how I use them in Chapters 1, 5, and 7. Between these extremes of scale lie the regionally based studies in Chapters 3, 4, and 8, the first of which draws on gazetteers of the two provinces between which the Yangzi Delta was divided, and the latter two of which were written on the basis of the surviving gazetteers of North Zhili. As I have used them, so these gazetteers have used me. By virtue of what they are, what they include, and what interests those who compiled them had, gazetteers have played a large role in determining the historical topics I have chosen, the problems I have set myself, and the approaches I have taken. It is fair to say that this volume would not have been about the Chinese state in Ming society had these books not been about just that.

Part I Space

1 The spatial organization of subcounty administration* State administration in Ming China extended downward in an organizational pyramid whose apex was the court and whose base was all households. The connecting middle level was the county, where the state appointed a magistrate to administer local affairs on its behalf.1 The household-county relationship was not unmediated, however, for between these levels the early Ming state elaborated a complex structure of administrative units. This structure consisted of four distinct but interrelated systems, based on earlier precedents but compounded with new elements that the Ming brought into use. The first of these systems subdivided the territorial space of the county into a hierarchy of cantons, townships, and wards. The second system mapped the social terrain by grouping households for census and fiscal purposes into the lijia or hundred-and-tithing system. The third was the baojia or neighbourhood mutual-watch program. Some areas had a fourth system, the xiangyue or rural covenants. These hierarchies stood in parallel with each other and often overlapped, the boundaries of one set of units replicating those of the others. This replication contributed to making these units into a robust and integrated structure of civil administration. The structure continued, with modifications, into the twentieth century, and not all the old boundaries have been lost, even today. Together, these subcounty systems constituted a pyramid of stepped jurisdictions that funnelled resources to the centre and maintained security and surveillance over the people. These systems not only made local administration possible, but endowed the state with a measure of access and efficiency earlier dynasties could not match. Keeping these subcounty units in order was recognized to be a component of good administration,2 not least because they provided the spatial template for tax collection. The logic of earlier studies of local administration has been to distinguish stateimposed mappings from the “actual” boundaries of village society, to speak of “artificial” versus “natural” communities, of “administrative villages” versus “real, historical, social villages.”3 While the notion of a tension between state and society is intuitively appealing, opposing the artificial/administrative to the natural/social can have the effect of masking their interaction. Administrative organization did in fact have its impact on the shape of local communities, an impact that goes back several millennia. Given that the Chinese state was often successful in impressing its patterns on local communities, one is hard-pressed to identify “natural” peasant communities that have not been in some way absorbed into official hierarchies or at least tagged with the names they were required to use. On the other hand, the influence went the other way as well. Social entities consistently pressed upward and moulded state systems, even becoming them. The “community” (she) in the Yuan dynasty, for example, was recognized as an

The chinese state in ming society

18

administrative unit in the Ming, and the Ming “village” (cun) became an official unit in the Qing. When the state recast its administrative systems, it usually took care not to alter too radically what was in place. Drawing new geographical boundaries, even down at the village level, was a hazardous move.4 The cumulative inertia of administrative systems, units, and boundaries resulted in a “close and continuing interaction between decimal hierarchies and the natural divisions of Chinese society such as village, intervillage association, lineage, and market community,” as Philip Kuhn has noted of a later period.5 Social collectivities shaped administrative boundaries as much as, if not more than, they were shaped by them. The Ming founder, Emperor Hongwu (r.1368–98), consciously designed the lijia system to be an exception to this general rule of conformity to prior spatial organization, though there is little evidence that its implementation at the local level brought about the uniformity he envisioned. For the other subcounty systems, there is even less evidence that “natural” communities were squeezed into “artificial” ones. What at first looks like state imposition, when viewed close up, seems more a matter of the state formalizing the informal and deeming what was already there to be what should be there, allowing the inertia of reality to overwhelm the force of the ideal. However much the Hongwu emperor desired to use his subcounty administrative systems to disrupt old patterns and impose new, the bureaucracy could not afford to push aside the existing social networks and rebuild them into an entirely new structure. The main challenges to the system from the fifteenth century forward were in any case commercialization and urbanization, not an interventionist state, but their influences also tended to run along channels already established by long-standing administrative practices. The essayist Xu Yikui (1318–c.1400) eulogized the administrative work of the Ming founder, his contemporary by praising the setting up of the field administration that passed from the capital through the counties down to the locale and the creation of the lijia system. As Xu phrased it, “once he had achieved great stability throughout the realm, he ordered prefectures and counties [be set up] and established the localregistration system.”6 These two systems together cast a comprehensive net of administration over the entire country; both also implied an interlocking set of spatial units for organizing territory into a virtual map that made the realm accessible to the state and organized the flow of information, resources, and personnel that sustained it. This study focuses principally on these two universal systems, and secondarily on the neighbourhood watch and rural covenant systems, which were implemented regionally in the second half of the dynasty. Ming administrators recognized these various systems as interrelated but typologically distinct.7 Each of these systems had its own principles of spatial organization, though their units’ boundaries tended to coincide such that units at the same level in different systems became spatially indistinguishable. The net effect of these interrelationships was a high degree of integration at the local level. This integration may have facilitated governance for magistrates, but it has made it difficult for historians to see what was going on behind the labels. Our first order of business is to sort these out. In addition to the confusing overlapping of systems in a local area, Ming sources show diversity of terminology across regions. The second order of business is then to show that Ming administrative units, despite local variations, adhered to an identifiable template working within a reasonably consistent administrative vocabulary.

The spatial organization of subcounty

19

To achieve these goals, I have set aside most of the prescriptive regulations the state published and relied instead on records of actual administration in local gazetteers. Most of the data comes from subsections in the chapters on administration or taxation, variously entitled lijia, xiangli, xiangdu, xiangyue, fangli, or baoli. Most gazetteers were compiled under the name of the local magistrate, and in many cases the magistrate actually took an active role in editing the published version. Their orientation toward social reality involved a mixture of the concrete and the abstract: magistrates relied on local practices while compiling them, yet they were also obliged to shape that data so that county systems would appear to reflect state policy. This double orientation may have introduced some distortion into this study, since a desire to conform to official patterns could induce compilers to observe in their local administrative systems greater formal order and regularity than actually obtained in practice. But that in itself is evidence that state officials were effective in making these systems authoritative, even if administrative practice strayed from the official models. The subcounty administrative system At the end of its third year in existence, the Ming state claimed jurisdiction over 887 counties; by the end of the dynasty, that number had risen to 1,159.8 To an official serving in the central government, these constituted the lowest level in the state bureaucracy. Bureaucratic appointments went down only as far as these counties, and county magistrates alone were answerable to the central government for the implementation of policy at that level. The magistrate in his county seat was in turn perched on the tip of one of anywhere from 887 to 1,159 icebergs of local administration that, from higher governmental levels, were largely submerged from view, but were what he surveyed when he looked out around him. Not all counties were subdivided into the same set of levels and units. The standard structure (see the list below), common in central and south China, had three subcounty levels: cantons, townships, and wards. The abbreviated structure, common in north China and other areas of lower population density, had only two levels: cantons and wards. Certain densely populated parts of central China used an elaborated structure: a subcanton was added between the canton and the township resulting in a four-tiered arrangement. Terminology varied between structures and regions, especially central and south China on the one hand and the north on the other. The units I have identified as “canton” (xiang) and “subcanton” (li) have ancient pedigrees within Chinese administrative practice.9 This usage became so well established that by the Song the collocation xiangli came to mean “one’s native rural area.”10 I use “township” to translate du in rural areas, fang and yu in urban areas, and xianq (xiang in the fourth tone, to distinguish it from xiang in the first tone, meaning “canton”) in suburban areas. At the next level down, which I designate as “ward,” terminology divided between north and south: she and tun used in north China, tu used in the south. The subsections that follow examine each of these units consecutively from higher to lower levels.

The chinese state in ming society

20

Subcounty administrative system Standard structure county (xian) canton (xiang) township (du) ward (tu) widely found in Fujian, Guangdong, Jiangxi, and Shanxi Abbreviated structure county (xian) canton (xiang) ward (tu, she, tun) widely found in Guangxi, Henan, Huguang, North Zhili, Shaanxi, and Shandong Elaborated structure county (xian) canton (xiang) subcanton (li) township (du) ward (tu) commonly found in South Zhili and Zhejiang

Canton The Ming inherited named cantons (xiang)11 directly from Song and Yuan practice. This was the largest territorial subunit within the county. Aside from a few places in central and southeast China where xiang were dropped from official usage at the beginning of the Ming (which is why the final column for Zhangzhou prefecture in Table 1.2 is blank),12 this unit was universal. A county could have as many as twenty cantons, though the average was about eight. In the Tang, the xiang had been a demographic rather than territorial unit, rated at 500 households,13 but that standard collapsed over the Song and Yuan, when the number of xiang per county declined and forced cantonal populations up.14 By the Ming, cantons were strictly territorial, though they were still expected to have roughly equal populations (as they do in the case shown in Table 1.1). Between counties, the size of cantons varied widely. Table 1.2 shows a range from under 800 fiscal households (Changsha) to over 15,000 (Songjiang). Urban areas were usually left outside cantonal jurisdictions. The cantons were the countryside, hence the use of terms

The spatial organization of subcounty

21

Table 1.1 Registered population by canton of Anqiu county, Shandong, 1589 Canton

Urban area

Number of households

Registered population

Number of wards

Average population per ward

788

2,297

5

459

Wenshui xiang

2,630

6,608

24

275

Linhuai xiang

2,867

6,823

26

262

Anlie xiang

3,121

7,497

29

259

Renshun xiang

3,380

7,609

30

254

Paoquan xiang

3,420

8,510

31

275

Guangzong xiang

2,968

7,423

29

256

19, 174

46, 767

174

269

3,064

7,412

28

Totals Average per rural canton

Source: Anqiu xianzhi (1589), 8.55b–57a. Note: The gazetteer compiler warns the reader that the records have lost track of real population; these numbers are at least a century out of date, more likely two. Contrary to the impression given by the data in the flrst two columns, Shandong in the late Ming was said to have large households: one source suggests eight as the average number of members per household.

like xiangmin (“people of the cantons”) for country folk and xiangsu (“customs of the cantons”) for rural practices and attitudes. Cantons were mostly vestigial markers. Their only official use was to provide boundaries when new counties were formed,15 although cantons often served as the first unit of fiscal record below the county. In one Fujian county, we know that tax registers were kept by cantonal clerks (xiangshu), since they were discovered falsifying them.16 Cantons might also be used informally to coordinate administrative tasks, such as forming militia17 or organizing irrigation work.18 Cantons could be meaningful social units as well. In the gazetteer he wrote for Shouning county (Fujian) in 1637, for instance, story-writer Feng Menglong (1574–1646) notes that the annual mid-autumn parades for the leading local deity, Goddess Ma, were organized by canton, for he reports that “each canton has its own sheshou” or head of the local religious procession association (yingxian she) in charge of overseeing the event.19 Subcantons The subcanton (li), sometimes found between the canton and the township in the elaborated subcounty system, was even more vestigial than the canton. Since the Qin, but consistently from the early Tang, the li was the chief subdivision of the xiang, but it began to lose importance in the Southern Song, when in some areas the number of li per xiang sank to one.20 As lower-level units assumed greater administrative importance, the

The chinese state in ming society

22

li was dropped in favour of the township at the next level down, or pushed up to the canton level, with which it merged.21 As a result, some cantons came to be known by the old li names.22 In some areas of Jiangnan and the southeast, the xiang and li units

Table 1.2 Households per subcounty unit in five prefectures, 1492–1612 Prefecture

Date of Registered Households Households Households Households per per canton statistics households per ward per township subcanton

Zhangzhou (Fujian)

1612

34,917

139.6

529.0

N/A

Ghangsha (Huguang)

1532

63,801

191.6

N/A

N/A

787.7

Jianning (Fujian)

1492

124,932

134.7

886.0

N/A

3,203.4

Raozhou (Jiangxi)

1502

162,074

141.2

613.9

N/A

2,532.4

Songjiang (S.Zhili)

1512

203,826

144.0

1,772.4

4,076.5

15,678.9

Suzhou (S.Zhili)

1506

582,000

147.0

2,425.0

N/A

7,864.9

Sources: Zhangzhou fuzhi (1613), 8.15b–16b; Changsha fuzhi (1532), 3.3a–23a; Jianning fuzhi (1473), 7.4a–26b, 9.2b–6b; Raozhou fuzhi (1511), 1.7b–24a; Songjiang fuzhi (1512), 9.14b–119b; Gusu zhi (1506), 12.1 a–2a, 18.1 a–25b; Da Ming yitong zhi, 8.1 b–2a.

continued to be distinguished as they had in the Song, but both were increasingly regarded as vestiges. A sixteenth-century text from southeastern Zhejiang thus contrasts “the old names of canton and subcanton” and “the townships and wards in current use.”23 Other Zhejiang and South Zhili gazetteers also distinguish “the old li” as a redundant category, in contrast to “the current townships.”24 Subcantons became important only in the absence of cantons; that is, when they took the place of cantons.25 Township The township (du) was the middle level in the standard three-level structure of the subcounty system.26 The du came into administrative use in the 1070s as a part of Wang Anshi’s baojia system.27 By the twelfth century it had been integrated into the xiang-li system in south China, frequently replacing the li as the chief subunit within the canton.28 Many local gazetteers do not record the use of du, however, until the Yuan.29 By the Ming, the township was to be found throughout central and south China, where it served as the main unit in the subcounty administration. The township appeared only rarely in the north, and, where it did, often existed only on paper.30 Townships were numbered rather than named, following Yuan practice.31 Figures 1.1–1.3 are rare examples of maps on which numbered townships have been marked.

The spatial organization of subcounty

23

Figure 1.1 Administrative map of Wujiang county, Suzhou, prepared for the county gazetteer in 1548. The numbered du throughout the countryside are the rural townships. The numbering of urban townships (here called bao) within the county seat in the upper left is unusual. Source: Wujiang xianzhi (1561), zongtu 1b–2a. A canton could have anywhere from one to a dozen townships (the average is about three), a county anywhere from ten to eighty (the average might be about forty). This variation is conveyed in Tables 1.3 and 1.4. Turning back to Table 1.2 we see that there could be from over 500 to close to 2,000 fiscal households per township. The three lower figures—between 500 and 900—describe the more usual range, suggesting a population per township on the order of 4,000 to 5,000 people. In areas that experienced little population growth after the Song and in which townships were not regularly subdivided into wards, the population of a township could dip well below a thousand (see Table 1.5).32 The township was a significant administrative unit right from the start of the Ming, when the expression xiangdu began to compete with the older xiangli to mean “the countryside.”33 This term came into wide use in the south,34 though it can even be found in North Zhili in the late Ming, where townships rarely appeared, and

The chinese state in ming society

24

Figure 1.2 Administrative map of Taicang subprefecture, showing townships numbered from 1 to 29. Source: From the 1629 reprint of Taicang zhi (1548), tukao 3b–4a.

Note: It is even more unusual for township boundaries to be marked on county maps.

The spatial organization of subcounty

25

Figure 1.3 Map of coastal defences of Jiading county, printed in the 1558 gazetteer. Townships numbers appear in ovals, larger settlement names in rectangles. This map is unusually oriented with southeast at the top; the Yangzi Estuary on the left side is labelled “the great sea” (dahai). Source: Jiading xianzhi (1558), courtesy of the Nanjing University Library.

The chinese state in ming society

26

Table 1.3 Distribution of subcounty administrative units in a standard structure: Raozhou prefecture (Jiangxi), 1502 County

Poyang Yugan Leping Fuliang

Dexing Anren

Totals

Cantons (xiang)

20

13

11

10

6

4

64

Townships (du)

70

36

42

56

36

24

264

337

208

295

102

118

78

1,138

Registered households

47,289

30,182

41,560

17,660

11,891

13,492

162,074

Households per ward

140.3

145.1

140.9

173.1

100.8

173.0

142.4

Wards (tu)

wards per township

Averages:

4

wards per canton

18

wards per county

190

townships per county

44

cantons per county

11

Sources: Raozhou fuzhi (1511), 1.7b–24a; Da Ming yitong zhi, 50. 1b–2b.

Table 1.4 Distribution of subcounty administrative units in an elaborated structure: Taizhou prefecture (Zhejiang), mid-seventeenth century County

Linhai Huangyan Taiping Ninghai Tiantai Xianju Totals

Urban wards (tu, fang)

24

26

13

11

8

10

92

Urban townships (yu)

7

6

2

2

2

2

21

125

61

55

92

48

?

?

Townships (du)

69

45

26

53

37

45

275

Subcantons (li)

42

28

16

19

12

22

139

Cantons (xiang)

15

95

6

4

6

45

Wards (tu)

Averages:

wards per county

93

rural wards per canton

10

rural wards per township townships per county

2 49

rural townships per canton

6

rural cantons per county

8

Source: Taizhou fuzhi (1722), 3.60b–71b.

The spatial organization of subcounty

27

Table 1.5 The subcounty system in Huian county, 1573 Canton (xiang)

Subcanton name (li)

Township (du) number

Ward (tu) number

Households (hu)

Population (kou)

Cultivated land (mu)

Urban

1

(fang)

2

139

916

5,486

3

120

689

8,155

1

133

821

6,175

2

125

902

4,425

Chongde

5

125

661

10,163

Xiangfu

28

133

1,020

5,702

29

139

1,050

5,981

30

157

1,020

4,105

31

147

982

2,571

Wenling

32

125

1,054

2,283

Anren

33

124

907

7,265

1

140

1,025

6,628

2

151

960

5,465

140

772

4,815

21

124

1,310

6,160

22

147

1,248

2,671

23

143

1,359

2,093

24

128

1,162

4,430

25

128

1,160

4,589

26

135

1,201

5,423

133

647

8,321

210

696

8,426

172

1,088

3,937

Wenzhi

Pingkang

Yanshou

34 Xingman

Anmin

19

Chang’an

20

Taikang

Shoujie

27 Zhongshu

Deyin

6 7

1 2

8

1 2

The chinese state in ming society

Guangde

28

9

124

879

2,435

10

152

987

5,950

157

860

13,105

137

702

10,240

144

782

11,293

130

954

3,483

4,062

28,864

171,775

140

995

5,923

Daixian

11

Minsu

12 13

Xinyi

Guihua

3 4

Zunxian

14 15

Tongxin

16 17

Lidian Totals Averages per ward

18/16

18 34/27

40/29

Note: A blank in the three right-hand columns indicates that the subcanton, township, or ward on that line had been in use earlier in the Ming, but was not as of 1573. Split totals give the formal number of units first, then the actual number. Sources: Ye Chunji, Huian zhengshu, 4.2b–6.31a; Huian xianzhi (1936), 1.20b.

never as du.35 Townships were sufficiently in common parlance that the late-Ming traveller Xu Hongzu used them (“township number x of such-and-such a county”) in his diary to indicate where he was on his journey.36 The township was the principal unit by which a county magistrate organized the administration of taxes. When National Academy students were dispatched in 1387 to survey agricultural land and record the results in what came to be called Fish-Scale Registers, so named because of the visual appearance of the maps that appeared at the front of each, they did so on the basis of townships.37 The practice of organizing county fiscal records by township is reflected in the numbering system for registering land. Plots of land within a township had registration numbers or characters in a series that did not repeat within the township; the series then began anew in the next township.38 According to a description of the registration system in Haiyan county, Zhejiang, by the scholar Wang Wenlu (1503–86), the registration numbers assigned to cultivated land within the county were grouped by township, not only to simplify bookkeeping in the magistrate’s office but also to identify (and, it was hoped, discourage) people who bought land outside their own townships as absentee landlords. According to the procedure in Haiyan, someone buying land within his own township could have the registration for the land shifted from the vendor’s household to his own, thereby eliminating ambiguities of ownership. Buying outside the township in which he resided meant that he had to register himself as a dependant under the vendor’s household. The lack of clear title implied by

The spatial organization of subcounty

29

this registration status would have made some landlords think twice before extending their holdings beyond their home township. Of course, Wang adds, this rule was constantly flouted.39 The dynastic founder identified the township as the appropriate unit for setting up community granaries (shecang) in central and south China, and some counties did indeed build them at this rate.40 He also required that every township construct altars for conducting state-cult sacrifices.41 In the heavily irrigated Yangzi Delta, townships could serve as units for coordinating water control: when the governor of South Zhili at the end of the fourteenth century issued regulations for the use of treadle pumps, he required that village officers make them available for emergencies anywhere within their township.42 Landlords within a township might similarly rally together on the basis of their common residence. In Cixi county to the west of the city of Ningbo, for example, “the rural people of Township 5” jointly protested in 1587 against the occupation of land along the edge of a lake, which was being turned into paddy fields and adversely affecting their water supply.43 Lest the township become a sphere of mobilization to engage in more than local infrastructural projects, early-Ming regulations stipulated that village officers not go outside their home townships. The power of townships to shape local elite identity is suggested in Yongan county, Fujian. In 1403, the forty-four townships established in the Yuan period were reduced to thirty-five. The nine discontinued townships, all in peripheral hilly areas, were merged with other townships within their cantons. Sixty-seven years later, however, two mountain townships were reinstated in response to local demand. It seems that elites in the hills found it advantageous to resurrect their earlier administrative status in order to remain distinct from lowland elites, rather than be subordinated under them as junior partners.44 Ward The ward was the smallest unit and lowest level in the subcounty administrative system.45 In central and south China, where townships were in use, the ward was known as a tu.46 In the north, where the subcounty structure was usually abbreviated, it could be called either a she or a tun.47 Because the boundaries of the ward were identical with the lijia system (see below), wards in the north and the south were often referred to colloquially as li, here meaning “hundreds,” but hundreds and wards were components of separate systems, a distinction contemporaries recognized and gazetteer editors followed, at least through the sixteenth century. The standard term for ward, tu, meaning “map,” was adopted for administrative units in the Southern Song from the practice of mapping all taxable land and inserting these maps in the land register (tuce) for each du or bao.48 Thus one could speak of being in such-and-such a register as a way of indicating spatial location. The character is often written without the enclosing three-stroke box radical, as a way of distinguishing this usage from that for map.49 Consistent with its Southern Song origins, tu was not used in the north.50 There, she denoted townships populated by indigenous residents, and tun, those formed after the Hongwu era in areas where government migrants had settled.51 She had been the major administrative unit in north China in the Yuan, nominally flfty households though not held strictly to that number.52 Tun had been a unit for military and

The chinese state in ming society

30

civilian agricultural colonies since at least the Han. By the late Ming, the origin of the she/tun distinction was understood but significant only in one way: urban wards were never known as tun, always as she (e.g., fangshi she, “urban ward”), whereas rural wards could go by either designation. A county could have as few as a dozen wards, but most had several hundred. Exceptional in size are the counties of Songjiang prefecture: Huating in the mid-Ming had 801 wards and Shanghai had 614.53 The variable determining the original number of wards in a county was of course population, since ward boundaries followed lijia boundaries. The grouping of wards into higher units, however, followed no set pattern. A canton might have ten, twenty, or more, although some cantons in sparsely populated areas in north China had only one.54 So too at the next level down, the number of wards per township varied considerably. In Fujian and Huguang provinces, many had only a single ward,55 whereas some prefectures in South Zhili had close to twenty wards per township.56 Three or four per township was more the norm. Since the ward was contiguous with the hundreds in the lijia system north and south, it should have had something over the mandated size in that system of 110 households, yielding a population approaching a thousand.57 The averages per ward in Table 1.2, with the exception of Changsha (its figure of 192 may have been due to the frontier demography of Huguang), confirm this, showing only minor variation between 135 households in Jianning and 147 in Suzhou. The one case for which we have reasonably good census data from the late-sixteenth century, Huian, Fujian (the subject of Chapter 2), shows a range of population figures between 647 and 1,359 people per ward. The same county shows that wards, at least in coastal Fujian, could embrace from 2,000 to over 13,000 mu of agricultural land (see Table 1.5). In coastal Zhejiang, on the other hand, Haiyan county in the same period had an average of just under 1,600 mu.58 At less than 15 mu per household, compared with 42 mu in Huian, Haiyan peasants were living on the border of subsistence.59 Fish-Scale Registers for two wards in Changzhou county, Suzhou, show larger acreages per ward of 2,868 mu and 3,000 mu respectively.60 Roughly, then, we may think of a ward in south China as having a population of about a thousand people and a cultivated area of several thousand mu. Most counties had fewer wards at the end of the dynasty than at the beginning. The equivalence between the ward and the lijia hundred meant that a decrease in hundreds due to registration evasion from the mid-fifteenth century forward produced a decrease in wards. There is practically no county that did not have its wards reorganized at least once.61 Since wards were usually numbered consecutively within townships and not renumbered when changes were made, it is possible to reconstruct the process by which wards were combined in many counties. The reduction of hundreds could, however, throw the alignment of hundreds and wards out of whack, as appears to be the case in Linzhang county, Henan, which at the turn of the sixteenth century had seventeen hundreds, on the one hand, but twenty-two wards (eight she and fourteen tun) on the other.62 The ward was the lowest-level unit for registering land. Each ward had its own FishScale Register, in which were recorded all parcels of land within the ward, identified by sequential registration numbers. The landownership certificates and land tax receipts issued by the county office followed the same system of identification by township, ward, and registration number. As the author of a seventeenth-century handbook for county

The spatial organization of subcounty

31

magistrates advised, the best way to prevent fraudulent land claims was to “compile registers of the land owned by people in such-and-such a township and such-and-such a ward.”63 A late-sixteenth-century magistrate in Beijing observed, however, that “getting all the land in the county into all the wards of the county” had become the administrator’s impossible dream.64 As well as organizing land, the ward served as the basic unit for locating households.65 By the mid-Ming, when corvée labour was levied, this was done by ward; and when the lijia records were recompiled, it was by ward rather than hundred.66 Urban and suburban units The subcounty administrative system extended to urban as well as rural areas. The hierarchy of units in urban areas tended to be simpler, with at most two levels. The canton, characterized as a rural unit, was usually absent.67 Urban townships could simply be numbered consecutively as one or more du within a county,68 though the more common practice was to set them apart by separate numbering and different terminology. They were usually called fang, a traditional term for urban units in China since at least the sixth century.69 A common alternative term was yu or “quarter”; in the southeast, towns had been divided into four yu in the Song and Yuan, and the term was kept in the Ming.70 The joint expression fangyu was the colloquial term for “county town” in many parts of Ming China;71 another was fangxianq, combining the separate terms for urban and suburban townships respectively. A xianq was a township immediately outside the city walls.72 This term was used in south China in the latter part of the Southern Song, though it first appears in urban administrative contexts in the Tang.73 The term yu was used occasionally in central China for suburban areas, a usage one source dates to the Yuan.74 In principle, an urban or suburban township should have had the same level of registered population as a rural township. This is so in the case of the county seat of Jiaxing, Zhejiang, which in the mid-1580s was divided into nine townships (fang) and had a total registered population of 6,950 households.75 These numbers yield a pertownship population of 773 households, which falls roughly in the middle of the range of population we found for rural townships. But this statistic is naive. Official figures wildly undercounted actual residents, so the population of urban townships must have regularly exceeded a thousand households. The tendency for urban townships to have only one ward may signal the difficulty of maintaining an accurate record of urban residents. Few magistrates attempted to create new urban wards after the turn of the fifteenth century, so that as town populations expanded, so too did ward populations, rather than the number of wards. When they needed more units, south China administrators tended to acknowledge population growth by adding suburban rather than urban wards. In north China, where suburban sprawl was not sufficiently great to require the designation of suburban wards, new wards that on rare occasion had to be formed under the pressure of urban expansion in the north remained for the most part within city walls.76

The chinese state in ming society

32

The lijia system The lijia system located every civilian in a hierarchy of command extending from the court down to the household. Its initial declared purpose was to organize households for service levy The founding emperor held out the hope that these imposed communities might serve as the units within which other social practices could be organized, such as community religious rites and primary education, including training boys to memorize his Grand Pronouncements (Dagao).77 Formally installed throughout almost the entire realm in 1381,78 the system not only distributed fiscal obligations but, by grouping households into permanent registered units, made local society vulnerable, at least in theory, to comprehensive supervision. The decimal principle of lijia organization, which grouped households into interlocking sets of tens, stood in a two-millennia-long tradition of systems built from multiples of five and ten households.79 Wang Anshi’s baojia system in the Song may have been a prototype for Hongwu’s planners, though his own more immediate inspiration was the weisuo system of military guards, battalions, and companies, which he put in place in the early 1360s well before he started regimenting civilians into units of a similar, uniform size.80 Applying such military principles to civilians was natural to a ruler who desired his subjects to be as disciplined and diligent as his soldiers. Hundred and tithing For the two eponymous units of the lijia, English medieval history provides the approximate translations of “hundred” for li and “tithing” for jia.81 The tithing was defined as a group of ten fiscal households, among whom rotated the post of tithing head (jiashou). The larger hundred was a composite of ten jia plus an additional ten households known as hundred captains (lizhang), among whom the duty of leadership rotated, as it did for tithing heads. In urban areas, hundreds were called fang, in suburban areas, xianq, using the language of the subcounty system.82 The affairs of the hundreds were also overseen by men in less well-defined posts, called hundred elders (lilao). The li of the lijia is the same character as the li translated above as “subcanton.” They derive from the same origin, for a li in the Western Jin and Tang was set at 100 households. The Song and Yuan dynasties did not impose demographic limits on the li, allowing its size to grow as its importance in organizing rural society declined. When the Ming founder decreed that a li of 110 households be established, he was ignoring what li had in practice become and setting them back on their Tang foundations. In physical area, hundreds in practice were identical with wards, so that the ward usually served as the unit of fiscal record for lijia obligations. Jia is a term of less venerable ancestry. It first appeared in administrative usage in the eleventh century as a variant unit of ten to thirty households in the baojia system. It continued thereafter to be used intermittently as a small grouping of households, and by the thirteenth century was common in south China as the subdivision of a li or du.83 Although the hundred is the unit on which studies of the lijia system tend to focus, the tithing was its basic component. Contemporaries stressed the impossibility of working with hundreds if the component tithings were not firmly in place: “When taxable individuals flee, the household is burdened; when the households

The spatial organization of subcounty

33

flee, the tithing is burdened; when [the members of] the tithing flee, the hundred is burdened.”84 Tithings were numbered one to ten. Hundreds were also numbered, sometimes in one series for the whole county, more frequently within individual townships, like the wards whose jurisdiction they shared.85 Very occasionally, hundreds were named.86 According to figures published in 1461, China was divided into 64,854.5 lijia hundreds; a late 1660s source gives a total of 68,929.5.87 Unlike counties, which gently proliferated over the Ming as population and administrative complexity grew, hundreds slowly decreased in number, as households escaped tax registration and fiscal assessment shifted from households to land. The distribution of hundreds correlated at the start of the dynasty with population density. Accordingly, the crowded Yangzi Delta, split between southeastern South Zhili and northern Zhejiang, was the region having the greatest number of hundreds (see Table 1.6). The region with the lowest rate of hundreds per county was the far southwest, where Guizhou managed to average only two and a half hundreds per county. Between these two poles stretched a gradient of hundred-to-county densities decreasing concentrically outward from Jiangnan. Despite their 110-household definition, hundreds were not uniform in size. There could be fewer than the statutory 110, but usually there were far more. Extra households were included under the category of “attached households” (daiguanhu), and those too poor to qualify for lijia service were appended as “supernumerary households” (qilinghu). This is in part why wards averaged about 140 households.88 Hundredal jurisdictions were supposed to be resurveyed once a decade to maintain equity among them, and the results written down in what were called Yellow Registers. Every county magistrate was required to submit these registers for all the households in his jurisdiction when the decennial “great compiling” (dazao, short for da zaoce zhi nian, “great register-compiling year”) came round. These registers, compiled first in 1381, could be painstakingly detailed.89 In practice, though, officials did not forward the results of their surveys lest increases in population prompt officials higher in the system to raise county taxes. The state tacitly accepted this practice. Precise records were expected to be kept, yet the costs of regular reassessment were happily traded for the budget security of set quotas. Hundreds were seldom altered, their boundaries rarely redrawn, and their numbers almost never hiked up. Only when population decreased would local officials or elites actively seek to get the number of hundreds changed.90 After Nanjing’s population was halved by the removal of the court to Beijing, for instance, the prefect in 1437 could legitimately petition that the urban hundreds in the city be reconstituted and reduced.91 By the sixteenth century, the lijia system weakened as a method for organizing local social collectivities. Hundreds came eventually to be thought of as territorial rather than demographic units and were superseded in all but official documents by the ward (tu).92 (By the mid-Qing, southern Chinese knew the lijia by the mutated expression tujia.) Even in official discourse, hundreds and tithings became units of fiscal account bearing no relation to real population.93 Redefined by acreage, a tithing could cover from several hundred to several thousand mu, with the size decreasing over time.94 For example, in 1601 a tithing in Jiaxing was defined as 250 mu of cultivated land; by 1641, after several further fiscal reforms, it was fixed at a mere 120 mu.95

The chinese state in ming society

34

Table 1.6 Hundreds (li) per county, by province, 1461 Range of hundreds per county

Province

about 150

Zhejiang

148.9

102–118

South Zhili

117.9

Jiangxi

102.8

Fujian

69.4

Shandong

54.0

Guangdong

51.7

Shanxi

47.2

Henan

27.6

Huguang

26.3

Shaanxi

21.8

North Zhili

20.6

Guangxi

17.0

Sichuan

11.8

Yunnan

7.0

Guizhou

2.4

47–70

17–28

7–12