The Destruction oft he Medieval Chinese Aristocracy Nicolas Tackett ])S 74q,35 ,13_3 ;)D!Lf Published by the Harvard University Asia Center Distributed by Harvard University Press Cambridge (Massachusetts) and London 2m4 © 2014 by ,he President and Fellows of Harvard College Printed in 1he Unit<:d States of America 1·he Harvard-Ycnching lnstirn,e, founded in 1928, is an independen, founda1ion dedicated to rhe advancement of higher education in the humanities and social sciences in Asia. Headquartered on 1he campus of Harvard Universi1y, the Institute provides fellowships for advanced research, for Liu Kan rraining, and graduate studies ar Harvard by compcritively selected faculty and graduarc studcn1s from Asia. 11,e lnsrirure also supports a range of academic activiries at its fifty parmer universi ties and research institutes across Asia. At Harvard, the Institute promotes East Asian studies rhrough annual contributions to the Harvard-Yenching Library and publication of rhe Hl/lwtrd }011mfll ofA sit1tic Swdies and the Harvard-Yenching I nstirme Monograph Series. Publicarion of this book was partially u11derwri11en by the Mr. and Mrs. Stephen C. M. King Publishing and Communicarions Fund, esiablished by Stephen C. M. King to further the cause of international understanding and coopcrarion, especially between China and the United States, by enhancing cross-cultural education and the exchange of ideas across national boundaries through publications of the Harvard University Asia Center. Library of Congress Caraloging-in-Publication Data Tacken, Nicolas, 1975- lhe destruction of the medieval Chinese aristocracy/ by Nicolas Tackett. pages cm. - (Harvard-Ycnching Institute monograph series; 93) Includes bibliographical references and indexes. ISBN 978-0-674-49205-9 (hardcover: acid-free paper) I. China-History-Tang dynasry. 618-907. 2. China- Hisrory-Tang dynasty, GI 8-907-Sourccs. 3. Aristocracy (Social class) China-Hisrory-To 1500. 4. Elite (Social scienccs)-China- Historr-To 1500. 5. Power (Social sciences)-China-Hiscory-To 1500. 6. Social change-China-History-To 1500. 7. Adjustment (Psychology)-Social aspects-China-History-To I 500. 8. China-Social life and customs-221 B.C.-960 A.D. 9. China-Social conditions-221 B.C.-960 A.D. I. Tide. DS749.35.T33 2014 951 '.O I 70862 l- dc23 2013038080 Index by rhc author €) Printed on acid-free paper Lasr figure below indicates year of this printing 20 19 18 '7 16 15 14 Contents List of Figures 1x Acknowledgments x, Conventions x111 Map of Tang China xv Introduction r The Transformation of Medieval Elites 3 Tomb Epitaphs as a Historical Source r3 r 71,e Bureaucratic Aristocracy ofM edieval China 27 Clan Lists and the Classification of the Great Clans 29 The Demographic Expansion of the Medieval Aristocracy 36 The Geographic Dispersal of Great Clan Descendants 44 Bureaucratized Aristocrats 61 Conclusion 67 2 7he Geography ofP ower 70 Localizing Elites 72 Capital Elites 82 National Elites in the Provinces 88 Ocher Elite Migratory Pathways 98 Conclusion 105 3 7he Capital Elite Marriage Network 107 Reconstructing Pacrilines 108 Localizing Patrilines 113 Geographic Distribution and Size of the Late Tang Political Elite 119 The Social Landscape of the Capitals 122 Marriage Networks and Social Capital 129 Conclusion 141 4 7he Late Tang Provinces 146 The Lace Tang Provincial System and the Hebei Autonomous Provinces 149 Recentralization after the Xianzong Restoration 155 Figures The Tang Policical Oligarchy and the Provinces 160 Social Mobility in Provincial Governments 170 Provincial Cultures 178 Conclusion 185 5 Huang Chao and the Destruction oft he Medieval Aristocracy 187 Chang'an under Huang Chao 191 0.1. Epitaph of Li Gao 14 Devastation in Luoyang and the Provinces 206 1.1. Forty-four eminent clans nor included in extant clan lists 34 The Demise of the Tang Elite 218 1.2. Frequency of attribution of great clan status in ninth-century The Survivors and rbe New Structure of Power 231 epitaphs (by region) 37 Conclusion 235 1.3. Relative prestige of clan attributions in ninth-century epitaphs (by region) 39 Appendix A: Guide co rhe Accompanying Database 243 1.4. Incidence of elites residing in the same prefecture or province as Appendix B: Estimating the Total Size of the Lare choronym place of clan origin (by region) 47 Tang Capital Elite 248 1.5. Choronym places of origin of elites from four select Appendix C: Sources of Ninth-Century Excavated Epitaphs 250 prefecrures 48 Bibliography 253 r.6. Incidence of elites residing in the same prefecrure or province as Personal Name Index 265 choronym place of clan origin (by select prefectures) 53 General Index 275 1.7. Core versus periphery localization of native and nonnative clans in the Lower Yangzi region 54 2.1. Percentage of individuals who died in prefecture/province of burial (by region) (800-880 CE) 77 2.2. Sire of death of individuals buried in the capital region (800-880 CE) 78 2.3. Length of epitaph rexrs in differem regions of China 83 2.4. Family traditions of officeholding a,mong elites (by region) (800-880 CE) 85 2.5. National versus local prominence of officeholding elite (by region) (800-880 CE) 86 2.6. Private residences in the provinces belonging to Luoyang or Chang'an elites 89 2.7. Places of provincial burial of individuals with national ancestries 92 3.1. Places of burial of members of select parrilines 115 3.2. Suburbs of burial of branches of two Luoyang-based patrilines 117 3 .3. Places of bu rial of the cop seventy-five officeholding parrilines 120 Acknowledgments 3.4. Major types of capital elites (Chang'an versus Luoyang) 121 3.5. Percentage of burials for kinsmen or kinswomen of chief minis- ters (by region) 121 3.6. Marriage network oflate Tang elite families 123 3.7. Composition of capital-based marriage cliques 126 3.8. Elite residency patterns in Chang'an 128 First and foremost, I would like to express my thanks to my gradu 3.9. Top ninth-century officials by patriline home base and marriage ate advisor, Robert Hymes, for his extensive help and encouragement network 134 during the decade I spent writing my dissertation and its prequel, this 4.1. Office rotations of two ninth-century Youzhou officials 153 present study. During my student years, I also had the good fortune of 4.2. Methods by which governors of select provinces acceded to office attending seminars taught by many others who have also made a mark (by year and province) 163 on my development as a historian, and co whom I am graceful: Robert 4.3. Backgrounds (civilian or military) of governors of select provinces Harrist, Li Feng, Martin Kern, Ellen Neskar, Richard Bullier, Zhang (by year and province) 164 Xiqing, Sarah Schneewind, Hal Kahn, Valerie Hansen, Christian de 4.4. Governors with family history of bureaucratic service (by year Pee, and Christian Lamouroux. Thanks also to Paul Smith for partici and province) 165 pating in my dissertation defense. For his insightful comments to my 4.5. Governors based in the capital or with ties to the capital marriage presema_ti~ns at conferences over the past few years and for inspiring all network (by year and province) 166 of us doing GIS and prosopography, I extend my gratitude to Peter BoL 4.6. Types of service in provincial bureaucracies (by home region and My research rime in China was made possible by a Columbia class of officials) 177 University Traveling Fellowship and a grant from rhe Fulbright-Hays 4.7. Places of county and prefectural service of capital-based Doctoral Dissertation Research Abroad Program. I also benefited from officials 180 a year at rhe Geery Research Institute and in the Stanford Introduction 4.8. Places of county and prefectural service of provincial officials to tl:e Humanities program. Thanks to all who made those fellowships serving in their home provinces 181 possible. Charles Salas and Ellen Woods were particularly supportive. p . Consolidation of post-Tang regimes in North China (875-920 I would also like to thank all who helped me obtain access to research CE) (by province and year) 216 material: Wang Xiaomeng and Li Jugang of the Shaanxi Provincial p . Consolidation of post-Tang regimes in the Middle/Lower Yangzi Archaeological Institute; Li Chaoyuan of the Shanghai Museum; Meng (875- 920 CE) (by prefecture and year) 217 Fanfeng of the Hebei Institute ofA rchaeology; as well as the staffs of the 5.3. Number of excavated epitaphs in the capital region and in Hebei/ Changshu Museum, the Changzhou Museum, the Yangzhou Museum, Hedong (by decade) 225 the Shanghai Library, the Shandong Museum of Stone Carvings, and 5.4. Number of excavated epitaphs from the capital region (by the various libraries at Peking University. period) 226 While writing this book, I have benefited from long conversations 5.5. Military battles and campaigns with more than ten thousand with Miranda Brown, Al Dien, Huang Yijun, Ye Wa, Tim Davis, Alex reported casualties in the period 750 co 919 229 Cook, Linda Feng, Tom Mullaney, Brian Vivier, Tony DeBlasi, Lu Yang, Jessey_C hoo, Zhang Cong, Sukhee Lee, Lewis Mayo, Zhang Tianhong, and l1yama Tomoyasu. Since arriving at Berkeley, I have also enjoyed countless wonderful and inspiring discussions with numerous of my xii Acknowledgments Conventions department colleagues, including several (somewhat grueling) ones on Lomas Cancadas (aka El Toyonal!). Special thanks co David Johnson and Geoffrey Koziol for their particularly detailed comments co my manuscript, and co Kristen Wanner, editor at the Harvard University Asia Center. I have also received invaluable feedback over the years from Rong Xinjiang, Li Hongbin, Patricia Ebrey, Lau Nap-yin, Ch'en Jo-shui, r. The bulk of the primary source material used for this study consists and many ochers. In addition, Liu Kan has consistently provided me of comb epitaphs and ocher funerary biographies. A personal name with the invaluable practical perspective of a peasant turned cultural index at the back of the book provides database epitaph numbers geographer. for all individuals with inscriptions mentioned by name in the text. Finally, I would like to extend special thanks to my father, Timothy In the main text (bur not in footnotes), an asterisk (*) followino Tackett, who first exposed me to the historical profession (in the archives t, a personal name indicates that an inscription exists for the indi- of the French provinces some three decades ago), and who has been a vidual in question. For full citation information, readers are invited constant source of encouragement, as well as of constructive comments to obtain the epitaph number from the personal name index and and criticism of my written work. consult the full Microsoft Access database (or the abridged Microsoft -N.T. Excel spreadsheet). See Appendix A for download instructions. 2. The raw data for most calculations in this study are available in the accompanying Access database. This database is described in appen dix A. It includes information about far more epitaphs (over 3,000) and individuals (over 30,000) than those mentioned in the personal name index. The calculations for most (but not al1 ) of the figures are available in the database. They can be found among the "queries" listed in the navigation pane. TI1ey are clearly labeled; the calcula tions for figure 1.1, for example, appear in che query "Fig u Forty four excluded eminent clans." In addition, a few footnotes refer co additional database queries. For example, note 14 in chapter r invites readers co consult "Fig Lnote14 9th c choronyms" in the database, a query chat can also be found listed in the navigation pane. 3. Years are identified using the Western calendar. In order co mini mize confusion, any single Chinese year is converted inco a single Western year, even though the Chinese new year begins several weeks after the Western new year. Thus, most events dared in chis monograph co the twelfth month of a given year would, in fact, have occurred in January or February of the following year in the Western calendar. 4. In maps of China, the coascline (with che exception of portions of the coast between Shanghai and the Bohai Sea), rivers (with che xiv Conventions exception of the lower reaches of the Yellow River), and latitude and longitude coordinates of most counties in the southeast and some counties elsewhere were obtained from "CHGIS, Version 3:: r--------------------------~ 4" (Cambridge, MA: Harvard-Yenching Institute, January 2007). "C Some additional coordinates of counties were obtained from Ruth I I • i * Moscern and Elijah Meeks, "Digital Gazetteer of Song Dynasty Cinhgi Tnaa nvg.I .eIr"a ( c2o0a1s0t)l.i nMesu cahn do rfi cvheer creomurasiensi,n agc cgoerodgsr wapihthic T daant aQ, iinxcialundg , (eaJ~t),-r: <.::(O:;";,I) 0~5g ' :(a§2"I,_). Zhongguo lishi ditu ji, volume 5. The boundaries of Tang China :0C::, 5a,· "io' ' gQ. 0 3 "' appearing on the map of Tang China on the facing page represent 3 a: 0 the approximate limits of direct Tang political control in the ninth "C:' (50' -0 century. The Tang regime did not actually demarcate its borders in 0 !2. < 5· such clear terms. Situating Tang China within the modern borders (l (t) "' of the People's Republic of China is done for reference only in order ro help orient the reader. Introduction Born into one of rhe most famous aristocratic families of ninrh cemury China, Ms. Lu possessed a pedigree of unquestionable eminence. Although her family was originally from Fanyang in the northeast, ir had relocated ro rhe great capital city of Luoyang ar least two centuries earlier. 1l1ere, Ms. Lu's forebears joined a social circle that came ro dominate society and, through its influence over the state bureaucracy, political power. Nor atypical of a woman of her station, Ms. Lu could trace her ancestry through an unbroken line of officehold ers going back ro the Han dynasty, some seven centuries earlier; several hundred of her clansmen had served in rhe governments of the succes sive post-Han dynasties, including rhe Tang (618- 907). When she was fourteen, mindful of the importance of pedigree, Ms. Lu's family found a suitable march for her in a young man from another great Luoyang based clan. Unfortunately, her husband died when she was still quire young. Nevertheless, she maintained her high status in society, looking after her children's education and arranging their marriages. She must have felt enormous pride when her son earned rhe prestigious jinshi civil service examination degree and when, in 878, her son-in-law-scion of yer another eminenr Luoyang family-became chief minister and one of the most powerful men in China. 111ree years later, however, Ms. Lu's life and the whole world she had known came roan abrupt end.1 l. Accounts of the life of Ms. Lu /Ii.I\ (818-81) and her husband. Li Shu $-ff (802-50), can be rcconsrructcd from the three epitaphs discovered in her tomb; for transcriptions of the epitaphs and a description of the romb in question (a tomb referred to as "M91t2" in the report), see Yawhi Xi11gy11n11 Tang 11111, pp. 168-252, 361-69. For a reconstruction of the genealogy of Ms. Lu's branch of the Fanyang Lus, see XTS 73 J:.:2885-2912; her grcat•grandfother appears on p. 2907 of the table. Chapter 3 of 1hc present monograph provides more information on her clan (which I identify as parri- 2 Introduction introduction 3 Ar first, the threat must have seemed remote to the residents of rhe of the first decade of the tenth century, most of the smaller regimes had capital cities of Chang'an and Luoyang. Although a series of uprisings been subsumed into larger states. At chis point, in the fourth month of had broken our in central China in the mid-87os, imperial armies had the year 907, the warlord controlling the Yellow River Basin and much succeeded by the end of the decade in routing the rebels, pushing them of North China ordered the execution of rhe last Tang emperor, by then far back coward the deep south. Suddenly, however, in the seventh month merely his puppet, thus bringing the dynasty to its final demise. What of rhe year 880, the rebel Huang Chao (d. 884) took advantage of a weak emerged from rhe Tang's collapse-first rhe so-called Five Dynasties and ness in the empire's defenses co cross rhe Yangzi River. Ir took him only then the great Song empire (960-1279)-was nor merely a sequence of four more months to capture Luoyang, and, before the end of the year, different political regimes. The tenth century witnessed the coalescing of his army marched into Chang'an. Then began one of the most infamous an entirely new social order. The great medieval families chat had main massacres in the annals of Chinese hiscory. Ms. Lu's daughter and son-in tained their prestige for most of rhe first millennium CE, across multiple law, rhe chief minister, were victims of the bloodshed. She, herself, was a dynastic transitions, vanished entirely from rhe scene. Under the Song bit more fortunate, escaping with her two sons co a family-owned villa in dynasty, a culture of merit came to eclipse rhe aristocratic ethos of earlier the countryside some one hundred kilometers ease of Luoyang. Bur her times, largely precluding any resurgence of rhe old order. good luck did not last. Perhaps because of an epidemic chat accompanied The present book seeks to explain chis dramatic societal and cultural the warfare and chaos, both she and one of her sons died of illness less rhan transformation. ft focuses on the final century of Tang rule. Although a month apart in rhe late spring of 881. With rhe surrounding region in generations of scholars have explored the profoundly different character turmoil, it was not until a year and a half lacer that it became safe enough istics of the Tang and Song elites, a vast corpus of new epigraphic mate for her youngest son to return her remains to Luoyang for burial alongside rial now allows us to elucidate the precise sociocultural processes char her husband. By then, four of Ms. Lu's five children had died without led from one social order to the nexr. On rhe basis of a collective biogra heirs, and the mood among surviving family members was grim. Scrawled phy involving tens of thousands of women and men, this volume exam unevenly onto rhe side of one of the epitaph scones buried in her comb was ines both how the bureaucratic ariscocracy of medieval China managed a note written by her nephew: "Another year has passed since the Son of to maintain its influence despite important political and institutional Heaven went to Sichuan. The great bandit Huang Chao has nor yet been developments in the mid-Tang and why rhe great clans disappeared so captured and killed. With the ravages of war overtaking Luoyang and completely with rhe fall of rhe dynasty. In the process, ir explores in great Gong County, the people have no means by which to survive" ~-f-"tlJ, derail the inner workings of an elite society living well over a thousand ~-P¼-~ 4'-, e. ft£ if~, fol ft jt-A~. .l> i~ A Jit., A..fiR-1.1£. 2 years ago. A subsequent study will address rhe cultural transformation Although rhe emperor did, in fact, return from Sichuan to recapture the during the posr-Tang period whereby a new elite self-identity emerged throne, rhe once mighty Tang dynasty lived on in name alone. Imperial char discarded many of the ideals and values held by Ms. Lu and the old legitimacy collapsed as warlords seized control of the provinces, ushering aristocracy ro which she belonged. in nearly three decades of upheaval, during which dozens of independent regimes all across the country barded for preeminence.3 By the middle The Transformation of Medieval Elites In fact, between rhe late Tang and rhe early Song dynasties, China line 5485). The earliesr known clansman buried in Luoyang was Lu Chisong Ai¼--l':t: underwent a series ofd ramatic changes that utterly transformed sociery.4 (569-625); see Mang Luo beizhi sanbai zho11g, p. 67, for an image ot his epitaph. 1hc chief minisrer in question was Cui Hang -/t.i.it (d. 880). 2. Sec epitaphs from Ms. Lu's tomb. For more on che Huang Chao Rebellion, sec norrh atter Huang Chao's rebellion, sec chaprcr 4 as well as Wang Gungwu, Stmcrure chapter 5. ofP o11Jer, pp. 6-84. 3. For a descripcion ot rhc warlords who barded each orhcr for preeminence in the 4. The Japanese journalisr and hisrorian Nairo Torajiro IJ.111,!i,;,t;.til~ (1866-1934) 4 introduction Introduction A commercial revolution brought about a significant monetization of the Bur perhaps the most dram.atic of the changes associated with this economy, the expansion in certain regions of marketing networks deep so-called Tang-Song transition was the transformation in the nature into the countryside, and the development and spread of urban centers.5 and composition of rhe Chinese sociopolirical elite. The twelfth-century Following a concomitant demographic transformation, the center of scholar Zheng Qiao i11:ffl., (1104-62) succinctly described its essence: "Up gravity of China's population shifted to the south, thrusting the Yangzi until the Sui and Tang dynasties, officials had dossiers [identifying the River valley and the sourheast coast to a prominence they had never offices of their ancestors], and families had genealogies. The appoint before held.6 Simultaneously, remarkable technological innovations ment of officials relied upon the dossiers; marriages between families increased agricultural productivity, while the greatly expanded use of the relied upon genealogies. ... Ever since the Five Dynasties, one no printing press substantially enlarged the literate population.7 During this longer asks about family background when selecting officials, and one same period, China would also witness a number of major developments no longer asks about family prestige when arranging marriages" mff i ;g in thought and religion, with the emergence of Nee-Confucianism-a rToJ:., 't~f4~k, ~~t¾f, 'ti::..ilV.i.l, ,;lm#f~~k, ~i::..!~:111!1, ,5lmi{' fundamentally new approach to understanding Classical texts-and the ttf .. . . mJi.,ti,x*, J[ll±Z-P05~i!t-, i\Hl!IZ-P05fl~IJ/j_10 In ocher words, popularization of new religious cults and practices.8 So fundamental whereas one's pedigree was critical in the Tang, by rhe Song, people were these changes chat many Chinese of lacer times imagined that the no longer felt that it mattered. Over the past several decades, historical Song regime itself had ushered in an entirely new era. Writing several scholarship has made enormous strides in elaborating upon the nature of hundred years later, the historian Chen Bangzhan ~f .,:~l}t (d. 1623) both the Tang and the Song el ires. David Johnson, Patricia Ebrey, Mao asked rhetorically, "The state system of today, the customs of the popu Hanguang, Sun Guodong, and others have characterized che relatively lace, the administration of the bureaucracy, the dogma of scholars, did circumscribed medieval aristocracy char defined its status on rhe basis of any of these not first emerge during rhe Song?" 4'-l)~i::..-$1], t\fb1i::..1~, blood.11 Robert Hartwell, Robert Hymes, Peter Bol, and Beverly Bossler ~-z-w*.W:-f-? 't~zr,Hr, 1fl;~zrfi'~, 9 have all described rhe new, more diffuse collective of elite families who first emerged in the early Song, families who justified rheir dominance of society and politics on the basis of ralenr and edu.carion.12 l11ey repre is credited with first describing this great transformation. 1l1e Naito thesis became sented rhe core element of what has sometimes been called che "Chinese inAucnrial among Japanese scholars of China in the mid-twentieth cenrnry, and among meritocracy" and would constitute one of rhe most strikino-distinctions t> American scholacs during the 1960s and 1970s (although Naitii's claim that the Song between Chinese and Western societies over rhe course of rhe subse- marked the beginning of the "early modern period" is no longer taken seriously). for quent millennium. an ovcr\'icw of the rhcsis and a more rcccnr critique, sec Miyakawa, "Outline" and Lau Nap-yin, "He wci 'Tang Song biange.'" More recently, there has been a Aurry of But beyond this general account of the sharp distinction between the publications on the thesis in Mainland China. See, for example, Li Huarui, "Ershi shiji rwo elites, disagreement remains as ro how this transformation actually zhong ri 'Tang Song biangc' guan"; volume 11 (2005) of the journal Ta11gya11 jiu, espe cially the introduction by Zhang Guangda; and Li Huarui, ed., "Tang Song biange"/1111. 5. Shiba, "Urbanization"; Shiba, Commerce and Society; Twitchett, "T'ang Marker I 0. Zheng Qiao, To11gzhi, 25:439. Shen Gua i;l,:H, (1031-95) made a similar obser System"; Twiccherr, "Merchanr, Trade, and Government"; Skinner, "lntroduccion: vation. He concluded an essay describing rhc medieval cusrom of ranking clans by Urban Development in Imperial China"; Shiba, "Sodai no mshika o kangaeru." observing rhar "chis cusrom gradually fizzled out at the end of the Tang" Jt.1/;;-.f.Jg* -}j 6. Hartwell, "Demographic, Polirical, and Social Transformarions," pp. 365-94. ;t;f ;!'t,1,1.• . Sec Shen, Mengxi bita11, pp. 772-73- . 7. Elvin, Pattern; Cherniack, "Book Culrnre and Texrnal Transmission." 11. Ebrey, Aristocratic Families; Johnson, Medieval Chinese Oligarchy; Johnson, 8. Bol, "This C11lr11re"; Hansen, Cha11gi11g Cods. "Lase Years"; Mao Hanguang, "Tangdai rongzhi jieccng shchui biandong"; Sun 9. Chen Bangzhan, Songshi jishi benmo, pp. 1191-92. Chen argues that rhe Tang Guodong, "Tang Song zhi ji shehui mendi." Song cransirion constituted the third of rluee grear hisrorical transformations, the first 12. Harrwcll, "Demographic, Political, and Social Transformations," pp. 405-25; two being the initial establishment of civilizarion in antiquity followed by the cstablish Hymes, Statesmen 1111d Gentlemen; Bol, "This C11!ture," pp. 32-75; Bossler, Powerfiil menr of a unified empire at the end of rhe third century BCE. Relations.