BOOK COLLECTING IN JIANGXI DURING THE SONG DYNASTY
Joseph P. McDermott
I. Introduction
As the study of the Chinese book expands beyond the concerns of production, technology, and bibliography, the entire subject becomes more central to the study of Chinese cultural and social history in general. Instead of asking merely how a book was produced, how its text may have changed, and what its text meant, we can also explore how it was crucial to the social reproduction of certain kinds of knowledge as well as of the holders of such knowledge. With this turn of scholarly interest to the dynamics of knowledge transmission, issues of book distribution, circulation, and consumption have naturally come to the fore. The uses that the Chinese found for books other than for reading then become vital topics of historical research. Seemingly arcane topics of libraries, book collectors, and book collecting become crucial to any discussion of how bodies of knowledge are transmitted by social groups and institutions and how an individual's deep commitment to book learning may expand the range of his commitments beyond kinship to encompass the education of fellow learners and learning in general. In short, Chinese book history, interpreted broadly, promises to tell us much about the making and makers of China's written traditions.
A basic part of the history of these makers is the history of book collectors.1 As conservers and transmitters of this written knowledge, they have played a vital and respected role in Chinese cultural and
1 The act of book collecting of course does not automatically presuppose the notion
of a book collector as a distinct category of person, but certainly by the Northern Song the Chinese had for this kind of person a phrase, cangshu zhi jia DUcH^Hc (Chao Yuezhi, Songshan Jingyu shengji, 6.11a), that was linked as much to a family as to an individual. Although the collectors discussed here often relied on family resources to be a collector, the focus on book collectors as individuals is largely justified, if only because most book collections survived in a family for only one generation and rarely for more than two or three (ibid.).
64
joseph p. mcdermott
social life over the past millennium. This essay on the practice of book
collecting in the Gan f| River Basin—that is, Jiangnanxi fXj^H and the western stretches of Jiangnandong fZQ^fjfl circuits, or roughly the area of present-day Jiangxi Province—during the Song Dynasty (960-1279) will explore three questions about these conservers and transmitters of China's written culture: Who were these book collectors? How did they collect their books? And, what were they able to do with their books? The aim is to explore some of the implications of book collecting, especially at a time like the Song when individuals or their families began in increasing numbers to form private collections away from the Buddhist temple and eventually away from the court as well. Whereas past study of the institutional history of Song Confucian learning has understandably focused on schools and the civil service examinations, the focus here will fall on certain individuals and the objects—that is, books—they collected that were basic to the operation of these Song Confucian institutions. We hope thereby to explain how Confucian scholars became central to learning and learned circles in the Yangzi Valley. Firstly, these collectors of books can be seen, individually and as a group, to have acquired collections larger, more comprehensive, and more widely dispersed than those of any rival, especially Buddhist temples and monasteries. Secondly, during the Song, especially the Southern Song, these scholars located and made use of new sources of imprints outside of court and eventually all government institutions. Thirdly, the collectors devised a variety of uses for their books, not just for the examinations or for the life of a literatus. Some went so far as to turn their collections into private academies that became local centers of learning, sometimes to satisfy the needs of general readers and sometimes to foster patron-client ties with local students in need of books and lectures.
In each of these three sections the emphasis will be on the diversity of the book collectors' practices that put the book at the center of their cultural and at times social life. The study of such diversity has not been easy. Song and later sources mainly disclose the number of book chapters or scrolls (juan in a Jiangxi collection. Few of them explicitly describe the development of a collection. Still fewer give an account of a collector's aim in forming these holdings. Some of these collectors, especially in the Northern Song (960-1126), kept their books outside of Jiangxi, often in the capital, and none of them has left a catalogue that survives today. In fact, the list of Song Jiangxi's book collectors found in Table 2.1 below, though culled widely from
book collecting in jiangxi
65
Song sources and modern surveys,2 cannot be described as complete. A collector's inclusion in this table is due as much to historical accident as to his bibliophilia. His friendship with a writer willing to write about his collection, his fame for reasons other than book collecting, a son's wish to promote his father's achievements, and the chance survival of writings about a collection, all these circumstances—at least as much as a collection's cultural significance—explain the presence of the names in this list. Although the most important collectors have been included, the statistics presented here on Jiangxi's collectors and collections should be read not as precise data but as general indicators of changes in Jiangxi's book world between the tenth and thirteenth centuries.
II. Scholar-Officials as Book Collectors
During the Song Dynasty the practice of book collecting in Jiangxi underwent a transformation that was both quantitative and qualitative. In the preceding two and a half millennia of recorded Chinese history fewer than five natives of this area of south China had formed or acquired a collection of books that had won them renown as book collectors.3 During the three centuries of Song rule, however, the number of Jiangxi book collectors multiplied fourteen-fold, to seventy. This figure is the highest for Jiangxi in any dynasty and, according to the most comprehensive survey of private book collectors, the highest for any circuit (i.e., proto-province) in the Song Empire.4
This Song Dynasty surge in the number of Jiangxi's book collectors betokens a major change in Chinese cultural and institutional history, a shift that histories of Chinese book collecting have tended to overlook due to their focus on court collections and, especially, individual book collectors. Past scholarly emphasis on individual book collectors has in effect turned the history of book collecting into a series of biographies, most notably of Chinese scholars (shi it), who are cast in the role of the main transmitters of literate culture in China outside
2 These surveys include Fan Fengshu, Zhongguo sijia cangshu shi; Li Yu'an and Chen Chuanyi, Zhongguo cangshwia cidian; Fu Xuancong and Xie Zhuohua, Zhongguo cangshu tongshi; Ren Jiyu HMM., comp., Zhongguo cangshulou; Pan Meiyue, Songdai cangshujia kao.
3 Fan Fengshu, Zhongguo sijia, 40-45, 53-57.
4 Ibid., 62-82, 138-45, 168-86, and 271-320.
66
joseph p. mcdermott
of the court for all of Chinese history. Yet, for a substantial portion of the second half of the first millennium ce the situation was quite different. During the Sui (589-617) and Tang (618-906) Dynasties the removal of southern court collections to Chang'an left the important
book collections in much of central and south China in the hands of neither private individuals nor their families. Instead, in contrast to the practice there from the eleventh century onwards, "in the Tang, before the spread of printing, the libraries in Buddhist monasteries probably constituted the only sizable collections of books outside the capital."5 The largest and most distinguished Tang library in central and south China (with the exception of one late T'ang private collection) was housed in the Donglin JHW Monastery in the northern Jiangxi mountain range of Lushan Hi[_L|. Its generally accessible collection of 10,000-odd juan, half of them the Buddhist Canon and the other half additional Buddhist and non-Buddhist texts, attracted numerous non-Buddhist visitors;6 more than thirty future high officials, including three prime ministers and fifteen celebrated poets, studied on its slopes during the latter half of the Tang.7 In fact, this library's carefully managed holdings of "the fine words of Confucians and Buddhists"8 enabled the Donglin Monastery to serve as the cultural and intellectual center of south China in the late Tang. In the Northern Song it became a center for literati Buddhism, while numerous other Jiangxi temples and monasteries acquired their own collection of the Buddhist Canon.9 But during the Song the overall significance of the Donglin and other monastic collections in Jiangxi declined. The increasingly popular Chan and Pure Land schools of Buddhism downplayed the importance of texts and textual learning, many larger and more comprehensive libraries were independently established in the Yangzi Valley by private individuals and families, and the Donglin collection was ransacked by a crazed emperor's persecution of Buddhism, unruly soldiers, and invading cavalry.10
5 Zürcher, "Buddhism and Education," 28.
6 Fu Xuancong and Xie Zhuohua, Zhongguo cangshu tongshi, vol. 1, 264; Ren Jiyu, Zhongguo cangshu lou, vol. 1, 675; Bai Juyi, Bai Juyi ji, 43.940.
7 Yan Gengwang, "Tangren xiye shanlin siyuan zhi fengshang," especially 387-95.
8 Quart Tangwen xinbian, 721.8268.
9 Halperin, Out of the Cloister, 87-89.
10 Considerations of space prevent me from providing more detail here on Tang Buddhist monastic collections, but I intend to present my findings on them in the near future.
book collecting in jiangxi
67
Thus, it is important to recognize that the presence of so many individual collectors in Table 2.1 represents a sea change in the character and concerns of the book world of Jiangxi and south China (outside of the court area in the Southern Dynasties).11 The change was not just from a few to many collectors and collections but also from one kind of collector and collection to another kind. A largely temple-based world of principally Buddhist textual learning, centered at one main site during the Sui and Tang, evolved into the far more secular and dispersed network of Confucian learning based on private individual and family holdings during the Song. The eventual de-centralization of knowledge transmission and conservation conferred great power on the owners of these collections, that is, the scholar-officials (shidaifu ~il y^^)> whose examination degrees and official careers often depended on their knowledge of certain Confucian books. While recent research has questioned the conventional wisdom that during the Song Neo-Confucianism, Neo-Confucian scholar-officials, and private academies replaced Buddhist thought, institutions, and monks as the center of intellectual culture,12 the clear shift to a preponderance of scholar-officials in Song Jiangxi's book-collecting circles helps to suggest why Confucian scholars have consistently regarded the growth of Song book culture as a crucial chapter in their rise to cultural and social pre-eminence in late imperial Chinese history. While the Song may no longer be seen as the era when the imprint replaced the manuscript as the principal form of book,13 in Jiangxi it was certainly the era when the scholar-official collector became the dominant social type of book collector outside of the court and capital.
If it then comes as no surprise to learn that Jiangxi's private book collectors were more than mere book collectors, it is best to dismiss one tempting interpretation right at the start: book collection figures do not easily translate into broader indicators of a family's wealth. Rich Chinese, to state the obvious, tended to put most of their wealth into forms of property other than books. As Yang Wanli IHUrS observed in the late twelfth century, "South of the lake (i.e., Jinghunan IRtfSJI^f Circuit) not a few great households are rich in property, but few are
11 Tian, Beacon Fire and Shooting Star, 77-95, shows the great size of Liang imperial and court collections.
12 Foulk, "Myth, Ritual, and Monastic Practice," and Halperin, Out of the Cloister.
13 McDermott, Social History, 43-81.
68
joseph p. mcdermott
rich in books."14 Thus, we cannot view the figures of these book holdings as reliable indicators of which Jiangxi families were relatively rich and had surplus capital or even of what portion of its wealth a bookish family tended to put into its books.
It is far more profitable to examine the educational and political achievements of these seventy book collectors. Some 80% of them can be seen to have had close ties to the government, holding an examination degree (65%) or just an official appointment or title (15%). As one might expect from the recent scholarly literature, this link was particularly strong in the Northern Song, when over 80% of all our Jiangxi collectors attained both a metropolitan degree (jinshi jJtlt) and at least one government post. Of the seven remaining Northern Song collectors in this list, that is, its seven non-degree holders, four received either a court appointment (Chen Jingyuan, Fan Duan) or a local posting (Wang Yi zE#H, Liu Yanbi), and the other three (Li Fen, Duan Zhong, and Wu Liangsi) were from families with traditions of book collecting and Confucian learning (Wu Liangsi's father, for instance, had attained both a metropolitan degree and an official career before passing his collection on to his son). This formal government link played a crucial role in the development and growth of many of these Northern Song collections, since they were by and large formed after their owners' appointment to office. However much these men had relied upon their own families' wealth to fund the education needed to win their examination degrees, it was only after their official appointments that most of their own collections grew significantly.
Moreover, this shift to the dominance of scholar-officials in book-collecting circles may have come earlier in Jiangxi than elsewhere in south China. Whereas private book collectors outside of the court appear to have gained their significant role in lower Yangzi delta collecting circles only in the twelfth century, in Jiangxi the change came earlier, in the eleventh century, when two-fifths of its recorded Song collectors were born or old enough to have at least started to collect books (See Table 2.1 below). In fact, nearly three-fifths of the table's collectors were either born or active in the Northern Song, a further sign of how decisive the first half of Song rule was in the history of Jiangxi's book collectors.
Yang Wanli, Chengzhai ji (2), 76.8b-10a.
book collecting in jiangxi
69
The eleventh century also saw a dramatic expansion in the size of these scholar-official collections in Jiangxi. Common sense requires that we treat general figures with caution, taking '10,000 juan to mean simply 'a large number of books' and 'several 10,000 juan to mean 'a very large number of books.' Even so, the overall rise in the size of private individuals' libraries over the Song is undeniable. In the tenth century, Jiangxi's few notable individual collectors gained local fame for having just 'several thousand juan!15 Even as late as the end of the eleventh century a Jiangxi scholar-official might be judged "erudite" simply for having "read several thousand juan of books, all of which he mastered."16 Yet, most recorded book collectors in Jiangxi from the eleventh century onward had 10,000 or more juan of books,17 and one collection reportedly contained as many as 40,000-odd juan.ls In the late twelfth century the peak figure rose to 50,000 juan in the collection of Zhao Shanying (1118-77), a scion of the imperial clan who was newly resident in Jiangxi. This figure, the highest recorded for any Jiangxi book collection in the Song Dynasty, ranks also as the second largest of all privately held Southern Song collections.19 It is also five to ten times higher than the top figure for any individual or family collection in Jiangxi before the Song.20
These Song collections, furthermore, were far more widely dispersed within Jiangxi than during the Sui and Tang, even if less so than at first appearance. The northern tip of Jiangxi, favored by its access to both the Yangzi River and Poyang Lake, had been the location of the Donglin Monastery and other Buddhist library collections at Lushan. It was also the site of most of the first generation or two of these Jiangxi Song collections. But in the eleventh century significant collections were formed by natives of more central prefectures like Yunzhou $tjt[, Jizhou cf'Jtl, Fuzhou $SJJ11, and, on the eastern side of Poyang Lake, Raozhou During the Southern Song collectors
15 E.g., Zheng Yuansu W&TCM, a book-collecting resident of Lushan for forty years in the Five Dynasties era, collected up to just 1,000-odd juan of old books (Lu Shan zhi (1933), 9.39a).
16 Liu Huan et al, San Liu jia ji, 40b, quoting Huang Tingjian.
17 See Table 2.1.
18 Zhang Bangji, Mozhuang manlu, 5.142.
19 McDermott, Social History, 51.
20 The largest collection I have found for a pre-Song individual in Jiangxi is a 10,000
juan collection that belonged to a native Sichuanese who lived on Lushan in the late Tang (Lushan zhi, 9.31b).
70
joseph p. mcdermott
from these same central prefectures still predominated. Jizhou alone provided over a third of all of Jiangxi's recorded book collectors, and along with Raozhou accounted for nearly half of the Table's seventy book collectors. Its Luling IS HI County had a concentration of book collectors matched by no other Jiangxi county during the Song. This preponderance naturally enough created a sharp imbalance in the spatial distribution of these book collections within Jiangxi. Of modern Jiangxi's fourteen Song prefectures and commanderies, twelve have at least one book collector in the table; but only five had more than five collectors over the course of the Song.
Finally, during the Song the concerns of these collections broadened far beyond classical and literary subjects. In the Northern Song the bookish interests of the Liu family of Gao'an developed from erudition in the Confucian classics and its commentaries21 to a very inclusive knowledge of history, ranging from calendars and administrative positions to geography, great families, and popular stories, for the entire length of dynastic and pre-dynastic China.22 While some late Northern Song collectors had an interest in books concerned with divination and Daoism to philosophical writings and fictional stories,23 this wider range of interests was expressed more often in the twelfth and thirteenth centuries. Collectors then admitted to an interest in military books, medical drugs, and divination (Wang Cao);24 astronomy and geography (Wen Yi);25 and, secret information in books about divination and crafts (Zhang Daxun).26 Overall, then, the book world of Jiangxi in the Song saw the emergence of a group of scholar-official, and would-be scholar-official, collectors, whose libraries over the course of the dynasty grew impressively in size and number to reflect a wider range of intellectual concerns. To study how these men formed and maintained these libraries will reveal how they found a place for their books and their families in the midst of great changes to book-collecting practices during the Song.
21 Liu Huan et al, San Liu jia ji, 24a.
22 Ibid., 56b, 63b; Li Yu'an and Chen Chuanyi, Zhongguo cangshujia cidian, 74-75; Pan Meiyue, Songdai cangshujia kao, 115-17.
23 Ren Jiyu, Zhongguo cangshu lou, vol. 1, 776; and Fan Fengshu, Zhongguo sijia cangshu, 90.
24 Cheng Minzheng, Xin'an wenxian zhi 94 _t, 2351-59.
25 Li Yu'an and Chen Chuanyi, Zhongguo cangshujia cidian, 97-98.
26 Wei Liaoweng, Heshan xiansheng quanji, 86.13b-16b.
book collecting in jiangxi
71
III. Song Ways to Build a Book Collection
How then did these men acquire their books? On the face of it, it would seem likely that they found them in Jiangxi. The geographic
and economic conditions that are usually presented to explain why some areas of Song China produced books existed in Jiangxi throughout the Song. Like the lower Yangzi delta, Chengdu j^KP, and Jian-yang W$k in the northern Fujian Circuit, Jiangxi had in abundance all the material resources required for book publication. Its mountains were covered with trees useful for making woodblocks, paper, and ink, and its diverse network of navigable rivers would have facilitated the shipment of these products to the lowland cities of the Gan River Basin and elsewhere. Indeed, Jiangxi's earliest known imprint dates from 847-49,27 two full centuries before that of Jianyang,28 just at the time when its registered population increased six-fold from less than 2,000,000 in 742 to more than 12,000,000 in 122329 and when agricultural improvements turned it into the richest or second richest rice-growing region in the Song Empire.30 Its increased demand for books, so evident in the relatively large number and size of its major book collections, would suggest that Jiangxi had its own important site for book production and distribution.
Oddly, it did not turn out this way. Of the 1,500-odd extant Song imprint titles, just 150 (or 170, according to another survey) were printed in Song Jiangxi.31 Hence, no Jiangxi location appears in contemporary lists of important Song publishing centers. Also, whereas northern Fujian had numerous commercial publishing establishments and the lower Yangzi delta's cities had a mixture of private and government publishers, Jiangxi's book printing, at least for the literati readership that would have included our book collectors, appears to
27 Seo Tatsuhiko, "Publishing Industry in Chang'an's Eastern Market."
28 Chia, Printing for Profit, 77.
29 McDermott and Shiba, "Economic Change in China, 960-1279."
30 In 987, the combined annual quota in Jiangnanxi and Jiangnandong Circuits for the basic land tax, the Twice a Year tax, was 2,200,000 bushels of rice, that is, more than a third of all the rice sent from south China to Kaifeng. In 1159 this quota was recorded at just 10,000 fewer bushels, thus accounting for nearly half of the rice then sent to the court in the abbreviated empire of the Southern Song (Shiba, Sodai shbgyb shi kenkyii, 155-56).
31 Poon, "Books and Printing," 468, 470-71; Du Xinfu and Qi Shenqi, comp., Jiangxi lidai keshu, 1-21, which lists 182 titles, of which 25, all county government publications, are no longer extant.
72
joseph p. mcdermott
have been done overwhelmingly by government offices: 86% of the titles of Jiangxi's extant imprints were printed by government offices
and another 5% by private or semi-private academies (shuyuan 1=|$t:). By contrast, its private publishing houses, commercial as well as noncommercial, played only a minor role in its publishing activities, printing just fourteen of its 157 (or 170) surviving Song imprints during the twelfth and thirteenth centuries.32
Moreover, even these Jiangxi government offices published few titles, quite likely because Jiangxi—unlike modern Henan Province with the Northern Song capital of Kaifeng flfW and modern Zhejiang Province with the Southern Song capital of Hangzhou l/l'JII in the lower Yangzi delta—never became the home of a Song imperial capital. Nor, but for the last few years of the Southern Tang Dynasty (937-76), did it serve like Chengdu for the Kingdom of Shu ID (906-65) as the home of the capital of any of China's numerous regional kingdoms in the Five Dynasties period (907-60). Its dynastic links thus being relatively weak from the tenth through the thirteenth century, Jiangxi's Song government printing houses likewise had a production level below that of their counterparts in these capital cities of the Song.33 In sum, despite the availability of resources, Jiangxi's growing number of major libraries must have been stacked with books acquired elsewhere in the Song Empire.
How then did Jiangxi's private book collectors overcome these obstacles and disadvantages to become the largest regional group of private book collectors in the empire? Elsewhere I have noted the central role of manuscript production in the reproduction and collection of literati books in the lower Yangzi delta right up to the sixteenth century and beyond.34 Such activity was certainly important in Jiangxi as well; witness the predominance of manuscripts in the 9,000-plus juan of Li Chang ^'fif35 (1019-68) and a temple's payment for the
32 Ibid., 19-21. Note that seven of these private imprints were published by Zhou Bida M'J&k, a Luling native, at the end of his life and another one after his death by his family's school. As Zhou spent a busy official career almost entirely in Hangzhou,
there is no guarantee that the seven titles listed as his family's publications were actually carved and printed in Jiangxi. If so, then the surviving private imprints from Song Dynasty Jiangxi might number as few as four.
33 See n. 31. Moreover, Poon, "Books and Printing," 113, mentions that "at least two hundred titles were published directly by the emperors' various agencies in K'ai-feng and Hang-chou."
34 McDermott, Social History, 83-94.
35 Su Shi, Jingjin Dongpo wenji shilue (1), xia, 859-61.
book collecting in jiangxi
73
hand copying of the entire Buddhist Canon.36 If our information on Jiangxi manuscript culture is nonetheless too scanty for the writing of its history during the Song, it is enough to indicate the shortage there of all types of books in general and imprints in particular.
In theory, such a shortage would have led Jiangxi collectors to pursue two options to initiate or expand their collections throughout the Song. That is, these collectors might have persuaded book owners (including merchants and peddlers) elsewhere to bring their books and learning to Jiangxi, or they themselves or their agents might have set off to acquire their books outside of Jiangxi and eventually bring them back home.37 The first occurred more frequently than our sources indicate, such as when Jiangxi collectors received special presents of books from the Song government (e.g., in 994 it donated 1,000 juan of books to the family of Hu Zhongyao ^H^H,38 whose members fled home with their collection from Kaifeng at the downfall of the Northern Song in 1127, and opened up a school with a teacher summoned from outside the region).
For the great majority of our collectors, however, the second of these options, whereby a book collector left Jiangxi and avidly hunted down books during his stay elsewhere, seems to have been the more common practice (recorded use of external agents is negligible). This practice began in Jiangxi no later than the mid-tenth century, when members of three influential families of Jiangzhou in northern Jiangxi—the Chen W., the Hong and the Hu —held a series of scholarly posts at sizeable libraries (eventually 20,000-plus juan) at the Southern Tang court in Nanjing l^fiH.39 They then proceeded during the first two generations of Song rule to form the first notable individual or family collections in Jiangxi.
IV. Northern Song Practice
In the Northern Song the most explicit instances of outside book collecting by Jiangxi collectors involved the ordering and purchase of books in the capital. For example, having failed to gain the
36 E.g., the decision in 1073 by one Jiangxi temple head to pay for the hand copying of the Buddhist Canon, rather than purchase a printed copy of it (Huili si zhi 1, "Yiwen," lb).
37 Xu Xuan, Xu gong wenji, 28.5a.
38 Li Yu'an and Chen Chuanyi, Zhongguo cangshujia cidian, 56.
39 Xu Changbian (1), 16.352.
74
joseph p. mcdermott
appointments he desired, the metropolitan degree holder Wen Cao ~3CW in 1038-39 visited Kaifeng where he presented a letter to the palace, requesting permission to purchase with his own family funds copies of books in the Guozi jian ll-pls: (Directorate of Education). His request granted, he brought these non-religious texts back to his home in Shicheng ^S'iffi, County.40 More commonly, examination candidates would take advantage of their visit to Kaifeng to acquire books, often at the Xiangguo ^iS Temple.41 Northern Song monks in search of Buddhist imprints also made visits to Kaifeng and Hangzhou to purchase books. Sent there by their Jiangxi temple, they ordered the printing of an entire copy of the Buddhist Canon from woodblocks held in the capital42 or had woodblocks carved in Hangzhou in 1085 for the quality printing of a specific sutra title.43
Few eleventh-century collectors were as rich as Wen and these monasteries. Thus, some serious book collectors from Northern Song Jiangxi, after attaining a metropolitan degree, moved away from their home circuit and set up residence elsewhere. In this new home, they proceeded to acquire books, often by means other than by purchase.44 One such degree holder moved to Changzhou JJt| in the lower Yangzi delta, where in the early twelfth century he made copies of books in the collection of a close friend.45 But most Jiangxi degree-holders anxious to become book collectors hoped for appointments to one of the many central government offices concerned with the collection, publication, or preservation of books (e.g., the posts of director, collator, corrector, or assistant in an institution such as the Bishu sheng
40 Li Gou, Li Gou ji, 23.253; Shicheng xianzhi (1781), 6.22a, 8.44b-45a.
41 McDermott, Social History, 63, 98, 108. Hence, I suspect the ability of Yang Xiaoben to acquire his collection. Having failed the metropolitan examination, he returned to Ganzhou if JJt|, where he spent twenty years reading in seclusion books he spent his entire fortune on purchasing. Fortune eventually turned his way when he was summoned to the court and promoted to a post in the Palace Library in the early twelfth century (Fan Fengshu, Zhongguo sijia cangshu shi, 70; SS (2), 458.13448).
42 Huang Tingjian, Song Huang Tingjian quanji, vol. 1, "Zhengji," 17.445. Also Cao Ganghua, Songdai fojiao shiji yanjiu, 26-27, tells of the formal procedures required to obtain a copy of the Buddhist Canon in Kaifeng during the Northern Song.
43 Abe Choichi, Zdtei Chugoku zenshu shi no kenkyu, 233-34.
44 The cultural attractions of living in Kaifeng were not, of course, restricted to book collecting. The capital, in addition to being supplied with the goods and luxuries of the empire, had the best education facilities for imparting the skills most needed to pass the civil service examinations. Hence, many officials brought their sons to the capital to be educated (Chaffee, Thorny Gates, 63-64).
45 Cheng Minzheng, Xin'an wenxian zhi 94 _t, 2351-59.
book collecting in jiangxi
75
^$ (Palace Library). These posts, which they sometimes held for many
years, provided relatively easy access to the books in the court and palace collections and also enabled younger men to make contacts crucial for later success in official circles. Men holding such appointments were commonly promoted to editorial positions at court, such as the editorship of the annals of a previous reign in the dynasty and the compilation of a book catalog like Chongwen zongmu tk^C!§,|=I for the Palace Library.46 Such access was available to officials, but not to monks or even former monks.47
To mention just two well-known Northern Song examples of such career progression will, I hope, suffice. The famous historian Ouyang Xiu, who is said to have accumulated 10,000 juan, gained easy access to court books and records, when he was appointed to compile some dynastic histories and to make an extensive catalog of the holdings of some court libraries.48 His own favorite among the next generation of scholar-officials, the book collector Zeng Gong WVi from Jian-chang jHH Commandery, followed up his attainment of a metropolitan degree in 1057 with twelve years of editorial appointments in institutions like the Institute of History and the Jixian yuan ^HiHn: (Academy of Scholarly Worthies). After a series of provincial postings in south and central China during the era of Wang Anshi's ZE^c ^ reforms in the 1070s, he was summoned back to the capital in the early 1080s to hold the even more prestigious editorial position of the compiler in the Institute of History of Longping ji (Collected writings on the period of great peace), a dynastic history for the reigns of the first five Song Emperors.49
At times the link between these appointments and the building of their holders' book collections is made explicit, as in the report that Chen Jingyuan PjflJtjC returned to Jiangxi from his appointment in the Hanlin Academy ftf^P^: with his baggage containing 100 bushels (dan ^E[) of books, all of them classified as works on the classics and history.50 More often, the link becomes evident only after repeated
46 Li Yu'an and Chen Chuanyi, Zhongguo cangshujia cidian, 68.
47 Chaffee, Thorny Gates, 54, on the 1044 ban on defrocked monks taking the official examinations.
48 James T.C. Liu, Ou-yang Hsiu, 32, 37, 102-3.
49 Zeng Gong, Zeng Gong quanji, "Zeng Nanfeng nianpu," 5-12.
50 Li Yu'an and Chen Chuanyi, Zhongguo cangshujia cidian, 71; Ren Jiyu, Zhongguo cangshu lou, 776.
76
joseph p. mcdermott
generations of a family gained scholarly appointments at court and proceeded to build up an important collection of books. Song degree-holders enjoyed relatively easy access to book collections in the Three Institutes (San guan mflf) and other government institutions in Kaifeng. Despite fires and management abuses, these collections contained books far more numerous and of better quality than virtually all private collections outside of the capital for most of the Northern Song. Residence in the capital and especially appointment to a position within these libraries offered aspiring book collectors golden opportunities to read and/or acquire copies of rare and not so rare titles in both government and private collections. Even if an official somewhat exceptionally abided by the government's ban on borrowing books from these libraries, he was still allowed to read them and to have copies made for his personal possession.51
The Liu H'j of Gao'an [ri;£c County
Most book collections of high officials and Confucian scholars in the
Northern Song were broken up upon the death of their owners,52 but about a third of our thirty-one Jiangxi book-collecting families from the Northern Song were able to avoid this and to establish hereditary collections over two or three generations. The most informative example of such a Northern Song family that built up a collection in the capital and still retained close ties with Jiangxi—they returned with their books to, and were buried in Jiangxi53—is the Liu family of Gao'an County in Yunzhou 3~JJJt| Prefecture. Three successive generations in turn saw to the creation of the collection (mainly the Confucian classics), its expansion (mainly through the acquisition of Han and later dynastic history texts), and then its completion (mainly by the inclusion of pre-Han texts and a full catalog). Although the third generation of the Liu family had difficulty building on the successes of the first two generations, their decades of successful book collecting show the opportunities available in the Song capital to ambitious bibliophiles from the provinces.
51 McDermott, Social History, 128-30.
52 Chao Yuezhi, Songshan Jingyusheng ji, 16.11b.
53 Liu Huan et al, San Liujiaji, 71a, which tells of the virtual disappearance of their graves in Jiangxi by the late twelfth century.
book collecting in jiangxi
77
The first successful scholar-official in the Liu family of Gao'an, Liu
Mei §i!j$fe, thanks in part to his reputation for great classical learning and integrity, won an appointment as a Companion to the Heir Apparent (Taizi zhongyun jk/^^ft) and to the Palace Library. But in mid-life he returned to Jiangxi with a small collection of books in ca. 1050. His son, Liu Shu Hll^S, having just acquired a metropolitan degree at the age of eighteen, replaced him in Kaifeng, where he spent most of the rest of his life in scholarly appointments at the court and enjoyed many opportunities to build up his own book collection. His desire to move to Kaifeng had been fueled by his repeated frustration in obtaining some famous books to read. He early on read the Confucian classics, even gaining an exceptional command of Li ji W. HH (Book of rites) and Chunqiu (Spring and autumn annals).54 But such basic historical texts as Hou Han shu tJ£?Jtll (History of the Later Han) and [Jiu] Tang shu [HflUfUs ([Old] dynastic history of the Tang) he had to borrow from others (presumably his father had no copy).55 According to one late Northern Song scholar, he "detested the fact that families of scholars in the south (nanfang shiren l^7l it A) did not collect books, and so he paid special attention to collecting."56 Liu Shu's hopes about northern collectors were soon dashed. As his friend Weng Shan m^M. observed, Kaifeng's private book collections were at first closed to him:
When Liu came to the Eastern Capital, his appearance was like that of a crane in the forest. I had heard of his reputation, and now I saw it. Trust is difficult in the dusty world to attain. The various Confucian scholars in great numbers stored up their treasures, and men who made advances to them were turned away. He made clear the events of a thousand years, and it was I alone who repeatedly discussed them with him.57
Some court appointments in the 1050s and 1060s, especially promotion to the Vice-Directorship of the Palace Library, seem to have improved his position, as he soon had access to a far wider variety of texts, everything from historical annals and chronicles to the texts of numerous schools of thought and popular stories (beiguan xiaoshuo
Sang Qiao, Lushan jishi, 6.18a, quoting from SS.
Liu Huan et al, San Liu jia ji, 56a.
Chao Yuezhi, Songshan Jingyusheng ji, 6.10b.
Liu Huan et al, San Liu jia ji, 29b.
Ibid., 56b.
78
joseph p. mcdermott
Improvements in his learning soon followed in two important ways. Predictably, he gained a breadth and depth of knowledge that won him some fame, even though his writings seem not to have circulated widely during his lifetime.59 Whereas some contemporary scholars studied solely Han shu illl (History of the Han) and others acquired a reputation for their erudition just by knowing both Shiji ^.HS (Record of history) and Hou Han shu, his command of Chinese history ranged from high antiquity up to the Five Dynasties.60 Sima Guang ^IMiyt, when compiling his magisterial Zizhi tongjian Hftn jrUm (Comprehensive mirror for aid in governance), specially invited him to join the project and then relied heavily on his grasp of the historical chronology of dynastic China.61 Their mutual admiration—Liu regarded Sima Guang as his closest friend and Sima treated him with great respect—was often commented on by their contemporaries.62
Secondly, his own collection expanded with acquisitions from nongovernment collections as well. In the Zhiping tn^F era (1064-67) he, along with others, was able to donate to the court a copy of Hou Wei shu t^iHlt (History of the Later Wei Dynasty) that was far more complete and accurate than any version then in the Palace Library.63 Nothing is said of how this rare book entered his hands, but clearly it, or at least part of it, must have come from a non-government—that is, private—collection, and his acquisition of it would have required wealth, close links to bibliophile circles, or special permission from its owner for the transcription of a copy. Thus, regardless of how he acquired this book, it is clear that his prolonged tenure in scholarly institutions at the court—and the attendant privileges he could share with selected others—opened doors once denied him as a young man from the provinces. These opportunities he then used to build up a book collection that eventually won the accolade of being described as "rich."64 Disgruntlement with political machinations over Wang Anshi's reforms drove him to bring his collection into retirement in
59 Xingzi xianzhi (1871), 14.84a.
60 Liu Huan et al., San Liu jia ji, 63a, 74b.
61 Ibid., 41b, 57a, and 63b.
62 Ibid., 59b.
63 Chao Gongwu, Junzhai dushu zhi jiaozheng, 5.184. See Kosen Hiroshi and Kawai Kozo, eds., Zuisho keiseki shi shoko, 265. Note, however, that Ren Jiyu, Zhongguo cangshu lou, 775, identifies the submitted text as Hou Han shu.
64 Liu Huan et al, San Liu jia ji, 76a.
book collecting in jiangxi
79
Jiangxi, where in his few remaining years his deep commitment to scholarship survived an eye affliction and palsy in his right hand.65
Liu Shu's son Liu Xizhong H'Jltf^ made a far shorter commitment to this blend of court scholarship and politics, and this probably contributed to the subsequent decline of the family's role as important book collectors. Like his father and grandfather, Liu Xizhong gravitated to the court and its scholarly institutions, but his dissatisfaction with its politics came relatively soon. Withdrawing in the early twelfth century from Kaifeng to the life of a recluse in northern Jiangxi, he brought along both his father's and grandfather's collections of books: "From the two previous generations the only thing handed down had been diagrams and books (tushu Hit)."66 Liu Xizhong's contribution to this hereditary property was his addition of pre-Han texts, such as Yijing WM. {Book of changes).67 In the eyes of one contemporary he had "a very rich collection of books,"68 one put together by scholars for scholars and "surely not formed to fill bamboo boxes and boast of silken slipcases and in the manner of a stupid merchant adorn the house with gold and pearls."69 Thus, in his retirement shortly before the end of the Northern Song he put this collection to scholarly and filial use: he finished a historical compilation initiated by his father, made another of his own about the pre-Qin era, and compiled a catalog of the family's books (which unfortunately is not extant). In sum, all three of these men spent their lives immersed in books, reading, writing, and not least collecting them, largely through government institutions, before returning with these acquisitions to Jiangxi.
Initially, it might seem odd that the book collections of these three members of the Liu family, separately as well as collectively, are said to number only a few thousand or ten thousand juan. Given the length of time these men spent at court, the range of their Kaifeng connections, and their scholarly achievements, these figures may strike us as far too low. Yet, it would be a mistake to assess the command that these scholars (and many other educated Chinese collectors) had of the literary culture of China solely on the figures for their book collections.
65 Li Yuan and Chen Chuanyi, Zhongguo cangshujia cidian, 74-75; and Liu Huan et al, San Liujiaji, 4b, 41b.
66 Liu Huan et al, San Liu jia ji, 76a.
67 Pan Meiyue, Songdai cangshujia kao, 116.
68 Liu Huan et al, San Liu jia ji, 75b.
69 Pan Meiyue, Songdai cangshujia kao, 116.
80
joseph p. mcdermott
The explanation is not simply the vagueness of these numbers. The
learning of these members of the Liu family, and in a wider sense their working collection, consisted of more than the volumes arranged on their shelves. It included, at least as much, the books stored in their memories, or to use the Chinese term, "the 10,000 juan in their stomachs."70 Their remarkable ability to recall texts, as recounted by one twelfth-century observer, would have easily enabled their literary and scholarly resources to go far beyond the confines of their own collections: "Liu Shu daily remembered 10,000 words, and did not forget them to the end of his life. Liu Xizhong could remember 5,000 or 6,000 characters [every day], and what his own son [daily] remembered was 3,000 characters."71 However exaggerated, this account rightly underlines how important access to a major court collection was to Northern Song scholars wishing to improve not just their family's book collection but also their own 'working collection' of learning. Consequently, after Liu Xizhong left the court, the next generation of the Gao'an Liu family had no degree holder, no court appointments, no close contact with the court's libraries, and thus made no additions to the family's collection. Indeed, its books and book catalog as held and compiled by Liu Xizhong ended up in the Nankang Pff HI Commandery's storehouse; but they soon disappeared from it, perhaps pillaged by the Jurchen invaders in Jiangxi at the start of the Southern Song.72 In the opening decades of the Southern Song the family's fortunes further declined, as by no later than 1179 its members had stopped making sacrifices to any of these ancestors' graves.73 By the early thirteenth century the family was no longer heard of and had dropped out of the historical record for good.74
The Liu ofXinyu if Nftj County
Compare then these three Northern Song generations' pursuit of books and official appointments with similar but longer efforts on the part of another Jiangxi family with the surname Liu, this time from Xinyu County in Linjiang       Commandery. At first glance, the resemblance
Liu Huan et al., San Liu jia ji, 27b. Ibid., 74b.
Hong Mai, Rongzhai suibi, 398; Lu You, Laoxue an biji, 9.114. Liu Huan et al, San Liu jia ji, 51a, 52b. Ye Changchi, Cangshu jishi shi, 25.
book collecting in jiangxi
81
is striking. If anything, the Xinyu Liu family were the more successful in linking their collecting to their official careers: eleven metropolitan degrees in the first century of Song rule and another seven over the next century and a half up to 121775 (four others won official appointment through the hereditary yin iH privilege).76 Their founder Liu Shi H'J;c£ (948-997) set the pattern for this success when studying Confucian books to become the top-ranked graduate in his examination cohort in the final years of the Southern Tang Dynasty:
In his youth he was fond of learning and did not work at production. At the age of eighteen or nineteen he took leave of his family and lived at Lushan, where he borrowed books to read. He studied Zuo IjL (Zuo commentary), Gongyang (Gongyang commentary), Guliang (Guliang commentary), and Chunqiu ^$fX (Spring and autumn annals), and on the side took in other classics. For a total of five or six years he did not return home, and in his career as a Confucian scholar he became increasingly proficient.77
Smoothly navigating a transfer of his dynastic loyalty from the vanquished Southern Tang to the triumphant Song, Liu Shi was soon appointed Vice-Director to the Palace Library and eventually headed one of the three agencies in the State Finance Commission as a comptroller. His own collection of books, retained by his wife at his death, was used to educate his five sons, facilitating their success in the examinations and officialdom and eventually winning literary immortality thanks to the Neo-Confucian philosopher Zhu Xi's (1130-1200) encomium on this book-laden inheritance as "an ink estate" (mozhuang Hit).78 His second son became an Erudite of the National University, another a Director of the Palace Library and a third a scholar at the Academy of Scholarly Worthies. In the third generation two members' mastery of the Confucian classics won them further scholarly appointments at court and the respect of more highly placed scholar officials: Liu Fen (1023-89) "won the esteem of eminent officials such as Han Ji and Wu Qing" and for a time "the mental submission" of Wang Anshi,79 and Liu Chang gained the appreciation of Ouyang Xiu for his extensive knowledge of everything from Buddhism, Daoism,
Xinyu xianzhi (1673), 10.1a-9b. Ibid., 10.49a, and 12.21b. Liu Chang, Gongshi ji 51.607.
Zhu Xi, Hui'an xiansheng Zhu Wengong wenji, 77.18b. Ibid., 12.17b-18a.
82
joseph p. mcdermott
divination, and medicine to geography, past and present biographies, local gazetteers, astronomy, Chunqiu, and pre-Qin bronze inscriptions.80 Liu Chang's second son Liu Fengshi (1041-1113) also gained fame for his scholarship, being proficient like his father at Han shu and in gaining appointments to scholarly court institutions like the Institute of History and the Guoshi yuan (Historiography Academy).81
Yet it was not just the duration of their success in officialdom throughout the whole of the Northern Song that collectively distinguished these Liu of Xinyu from the three Liu of Gao'an. In addition, from the end of the tenth century the Liu of Xinyu removed themselves from Jiangxi to the capital. In leaving behind the rural fastness of their Jiangxi home village seventy // east of the county seat,82 they shifted their officially registered residence to the outskirts of Kaifeng.83 Henceforth, not only would they and their wives (non-Jiangxi natives) be buried here, but also they would keep their book collection here right up to the end of the Northern Song.84 In addition, despite the success of some members of the family in continuing with scholarly work at the court, other members of the family with metropolitan degrees branched out to pursue more varied and practical appointments in officialdom throughout the eleventh century.85 These members seem to have added little if anything to the family's collection of books, but doubtless did much to broaden the scope of its connections. A further difference probably followed on from this: the collection of this Liu family, despite the praise showered on it by later Neo-Confucians, amounted to just "several thousand juan." Their decades of access to Kaifeng's libraries may well have fed their memory-stomachs more than their own collection's shelves. But, even apart from the question of the size of their collection, the family never matched the Liu
80 SS (2), 319.10386-87; Huang Zongxi, Song-Yuan xuean (1), 4.69.
81 SS (2), 319.10388-90; Xinyu xianzhi (1673), 12.21b.
82 Ibid., 13.4a-b.
83 Zhu Xi, Hui'an xiansheng Zhu Wengong wenji, 90.4a.
84 Yang Wanli, Chengzhai ji (2), 127.2b-3a; Liu Bin, Pengcheng ji, 38.509. By contrast, Liu Huan and Liu Shu were buried initially on Wulao Precipice JH in Xingzi County (Liu Huan et al., San Liujiaji, 66a), and Liu Shu reburied by his son at Mount Longquan flJl in Dehua t§»^§ County, Jiangzhou (ibid., 59a). Also, Liu Chang, Gong shi ji, 36: 435, tells of his father's elder brother leaving Kaifeng and setting up—not in Jiangxi but in Changshu County in Suzhou fsfc'jrl Prefecture in the lower Yangzi delta—a library building to house his collection of several thousand juan of books.
85 E.g., Liu Fengshi WWW (SS (2), 319.10386-87).
book collecting in jiangxi
83
of Gao'an in establishing a reputation for scholarly commitment and achievement.
This account of two Northern Song patterns of book acquisition probably downplays the contribution of book purchasing to the growth of these scholar-officials' collections. Although these collectors' purchase of books from other scholar-officials is only infrequently recorded, scholars did play a significant role in the distribution and sale of literati books in the capital even in the eleventh century.86 Furthermore, purchase would have become more common for many collectors once they left these central government institutions for appointments elsewhere. It would have been most common for those collectors, including the sons of official collectors, whose failure in the examinations would prevent them from repeating either of the two Northern Song patterns of book acquisition at the court that we have discerned.
V. Southern Song Practice
The fall of north China to the Jurchen in 1127 brought an end to the court-based arrangements for Jiangxi book collectors. The next century and a half of Southern Song rule saw the emergence of a more socially diverse group of collectors as well as collecting methods that gave greater prominence to purchasing. For the first time one can find in Table 2.1 men who themselves had acquired neither a degree nor a government appointment and whose fathers had not done so either. Born as commoners, they died commoners, and though they did not share, as far as we can see, any identical occupation or source of wealth other than perhaps landownership, their mere presence in this list indicates a broadening of the social base for the ranks of Southern Song book collectors compared to the Northern Song. Not having direct access to court collections and imprint gifts from the government or officials, these collectors would have understandably relied more on the purchase of books than their predecessors in the Northern Song.
Yet, note that this type of book collectors represents a distinct minority in Table 2.1 and does not signify a decline in the desirability of an examination degree or official position for at least three-quarters of these book collectors. No fewer than twenty-three (i.e., 70%) of the
McDermott, Social History, 108.
84
joseph p. mcdermott
table's forty collectors active in the Southern Song acquired an official degree, position, or status: fifteen had a metropolitan degree; two had
the status equivalent to a Ming Dynasty provincial degree; two were students at the National University hoping for a degree; and four had no degree but held at least one official appointment (two of whom as members of the imperial clan gained an appointment to the Hanlin Academy).
Eleven other collectors, despite their not having gained a degree or an official position, sought or had ties to officialdom. For instance, Peng Weixiao l^lti^, armed with recommendations from his influential Jiangxi friends Zhou Bida JeOi&^C and Yang Wanli, traveled to Hangzhou, where his knowledge of current affairs and his skill at prose-poems (fu reportedly won him the offer of employment from no less an authority than Emperor Ningzong ^P^c. (He ended up refusing the offer, presumably because it failed to meet his expectations).87 Four others were brought up in families with a reputation for Confucian learning and with a recent appointment to officialdom, and a further four, including Xu Bochen f^ffi^, who received a prestige title in his old age,88 collected books for their sons' education for the civil service examinations (information is lacking for only Zhang Hong 5H§£ and Luo Jingfu JHHafc^c). Altogether, then, nearly 80% of the Southern Song collectors as individuals had some sort of link, however complicated, stubborn, or loose, to the central government; the figure for this link rises even higher once we consider family ties.
The lure, of course, was the continuing status of officialdom, especially as some of its privileges were extended in the Southern Song to those who passed merely the local and not the metropolitan exams.89 One privilege for officials and degree-holders remained facilitated access to books in government collections or published by the government either at the court or, significantly, in the provinces. Some mid- and late Southern Song book-collecting officials like Zhou Bida remind one of the Liu of Xinyu in the Northern Song, as they continued to follow the pattern of spending most of their careers in the capital, Hangzhou, and keeping most of their books there. Beginning with sinecures in the Palace Library and other government book
87 Taihe xianzhi (1753), 17.4b-5a.
88 Fengcheng xianzhi (1664), 9.37b.
89 Bol, "Sung Examination System and the Shih," and Takahashi Yoshiro, "Sodai no shijin mibun ni tsuite."
book collecting in jiangxi
85
collections, they worked their way up to high appointments in other ministries. Zhou himself had several library buildings erected for his impressive collection of rare books and paintings, all assembled on the way to becoming a powerful figure around the court for three decades. Born to northern parents who migrated to Luling County, Jizhou, he claimed to have modeled himself on three Northern Song book-collecting officials from Luling.90 He spent so few years of his adult life in Jiangxi, however, that one can understand his absence from most modern lists of Jiangxi book collectors.
Other court-based collectors with closer ties to their native Jiangxi, such as Ma Tingluan Mj^M and Chen Zongli PJItk"!!, followed an alternative Song pattern already seen in the careers of the Liu family of Gao'an. They spent a long time in the capital before their despair over politics at court and over military defeats in western China drove them home to a retirement spent reading books in Jiangxi. Both of these men spent their court careers largely in scholarly appointments to the Palace Library, the National University, and the Compilation Office for the Veritable Records and the Dynastic History.91 They would also have benefited from the presence in the capital of numerous government offices and commercial establishments that printed books for their members and friends.
Yet neither of these court-based patterns of book collecting was dominant in Jiangxi's book-collecting circles during the Southern Song and especially the first half of the Southern Song. The loss of Kaifeng's court collections to Jurchen fire and capture initially made the court's newly formed libraries in Hangzhou far less attractive to collectors than the scholarly collections and institutions of the Northern Song had been. Since the Hangzhou court's Palace Library holdings for officials only regained their peak eleventh-century size in the late 1170s (the very time when Emperor Xiaozong still thought it neces-
sary to send a man to Sichuan in western China to hunt down largely Northern Song government imprints for the Palace Library),92 it is probably not surprising that Jiangxi's degree-holding book collectors by and large spent less time at the court during the first half-century of Southern Sung rule than they had in the Northern Song. Of the forty
SS (2), 391.11965-72.
SS (2), 414.12436 and 421.12594-95.
McDermott, Social History, 55 and 217 n. 63.
86
joseph p. mcdermott
book collectors in the table who were active in the Southern Song,93 just eight made a career at court in scholarly institutions, half in the twelfth century and half in the thirteenth century. These men who continued a preferred pattern for Northern Song collectors thus constituted a minority of the Jiangxi book collectors of their time. None of their families rivaled the Liu of Xinyu or of Gao'an in the persistence of their success in the examinations, appointments to scholarly institutions at the court, and access to their books.
Instead, most of Jiangxi's Southern Song collectors with a degree or government appointments spent their official careers largely in the provinces. It is at this point that the bi-polar distinction usually made to contrast the social and political halves of the Song Dynasty— the capital centralism of the Northern Song political elite versus the localism of the Southern Song political elite—loses its cogency. This model, retaining too much of the bureaucratic structure of the Chinese political order, overlooks the ties that people in different Southern Song provinces formed with one another outside of government mediation, especially when searching for particular items. Whereas in the Northern Song a string of provincial appointments would have cut a collector off from the main offices of government publications in the capital, the seven-fold expansion of provincial publishing sites from thirty in the Northern Song to over 200 in the Southern Song94 suggests a greater availability of imprints away from the court than before (the actual rate of expansion is even greater, since some Northern Song publishing sites were in north China and thus not part of the Southern Song Empire). Many of the new Southern Song publishing sites were attached to government offices, whose officials enjoyed special access to the use or purchase of such publications. The information on this means of book acquisition remains mainly suggestive (for example, Luo Wujing ^M^a served as the secretary of Jianning $H ^ in northern Fujian on his way to accumulating 10,000 juan and a
93 I have chosen to use "active in the Southern Song," rather than birth in the Southern Song, as my frame of reference, since it allows me to deal with all book collectors active as adults after 1126. As a rule then, this category includes all men who were twenty years of age or older in 1127 or whose lives were largely lived and shaped by their experiences in the Southern Song. The number of book collectors active in the Southern Song is thus forty.
94 Poon, "Books and Printing," 1, 11, 468-74.
book collecting in jiangxi
87
reputation for being a bibliophile,95 while Zhang Daxun's ijjI^HHI collection of several 10,000 juan surely benefited from his appointment to the book-publishing center of Chengdu Prefecture in ca. 1200).96 The importance of non-court official connections for provincial collectors is underlined by the relative sizes of collections owned by officials and commoners in Southern Song Jiangxi. Scholar-officials, with their privileged access not just to the Palace Library in Hangzhou but also to books published in provincial government offices, had collections far larger than did those Jiangxi commoners who stayed in Jiangxi and thereby forfeited such privileged gifts or even purchases (no Southern Song commoner's library is said to have contained more than 10,000 jwtm).
Under these circumstances, for both commoner and official collectors the outright purchase of books from government and commercial establishments was an obvious alternative means of acquiring them. Hangzhou, with publications from its numerous central government offices and some twenty-odd private commercial bookstores, certainly catered to the thousands of officials serving or visiting there at any one time.97 Yet, even as purchasers serious Jiangxi book collectors no longer had to get their books in the capital. Witness the practices of the Liu of Xinyu, whose "books stored up [in Kaifeng] had been scattered and not preserved" with the fall of north China to the Jurchen in 1127. Some of the family survived the flight southward to Jizhou Prefecture in Jiangxi, where, bereft of the libraries, appointments, and emoluments that had made their life in Kaifeng comfortable and secure, they were obliged to face a radical decline in their living standards. Some became so poor that they could afford to eat only vegetables for a third of their meals.98 Under these conditions the family's repeated failure in the examinations threatened to end the interest it had had in book collecting over the past one and a half centuries, until one fifth-generation member took special steps to revive the tradition: "Liu Chu H'J^ was the only one who was concerned about what had been
95 Jianning xianzhi (1919), 8.6b; and Hu Quan, Hu Danan xiansheng wenji, 31.1b-2a.
96 Wei Liaoweng, Heshan daquan wenji, 86.15b; Fan Fengshu, Zhongguo sijia cangshu shi, 74.
97 McDermott, Social History, 98.
98 Yang Wanli, Chengzhai ji (2), 127.3a.
88
joseph p. mcdermott
lost and scattered from what former generations had stored up."99 For twenty years this man, who never acquired a degree but eventually obtained a provincial office and its precious salary, saved and scrimped for a single purpose: "He economized on his clothes and food, devoted all his energy to assembling [books], and only in 1152 did he restore the previous several thousand juan."100 In 1182 his sons were said to be protecting the collection and making some additions. They gave Zhu Xi a copy of Han shu that Liu Shi had studied in his youth at Lushan,101 but by this time it was being reported that "only barely did [the collection] avoid a collapse."102 Come the mid-thirteenth century, the Liu family's interest in books (and success in the examinations) had completely faded. Described now as "lacking wealth" (wu fu U) and with only twenty or thirty persons (ren A; males?) remaining, they had become primarily concerned with maintaining granaries and providing themselves with famine relief rather than with libraries and books.103 In contrast to their Northern Song practice, they had kept their restored collection not in the capital but in Jiangxi throughout the Southern Song. Yet, as their string of examination successes came to an end in the early thirteenth century, they became "local." Their book collection "collapsed," and soon afterwards they too dropped out of the historical record.
Liu Chu's purchase of these books outside of Jiangxi in the mid-twelfth century is one further indication that this region still had no significant publishing site. Admittedly, some Jiangxi book owners would seem to have made their purchases in Jiangxi. Wang Yi 3i)M is said to have returned to Luling after the disturbances of the early Southern Song and purchased 10,000 juan with much of his family's wealth,104 and in the second quarter of the thirteenth century another Luling native, Wen Yi, would also seem to have purchased his books in this area, as he rarely traveled and was reportedly so fond of reading books that in addition to personally copying over 100 volumes (ce M) he is said to have pawned his clothes to buy books he had not yet
99 Zhu Xi, Hui'an xiansheng Zhu Wengong wenji, 77.19a; Luo Yuan, Ezhou xiaoji 4.37.
100 Zhu Xi, Hui'an xiansheng Zhu Wengong wenji, 77.19a.
101 Ibid., 81.25b-26a.
102 Ibid., 19a, 20a for the date.
103 Liu Chenweng, Liu Chenweng ji, 3.59-60.
104 Yang Wanli, Chengzhaiji (2), 128.8a.
book collecting in jiangxi
89
read.105 These two men and eleven others in our list are identified as
"men of Luling," and conceivably they bought and sold books among themselves.
But Jiangxi merits not even a dismissal as a book distribution site in an account written in the 1190s about the buying practices of book collectors from Changsha MfJs. When these collectors sought books outside of their native circuit of Jinghunan 7f!J$jip^, they travelled instead to northern Fujian, Hangzhou, or Sichuan.106 One is forced to conclude that Jiangxi's book publication level remained low, as Jiangxi collectors usually continued to turn to outside sources—publishers, stores, or collectors—to purchase their books, regardless of whether they were from commercial or government outlets.
Thus, throughout the Southern Song, even more often than in the Northern Song, we read of Jiangxi men seeking to buy Confucian as well as Buddhist imprints outside of Jiangxi, especially for large orders. One external site of purchase they visited was Nanjing, where in the 1180s Zhu Xi and Lu Jiuyuan H~J~L$tfl dispatched men to find books for their recently re-opened Bailudong shuyuan SJSfl^ISI^ (White Deer Hollow Academy).107 But Northern Fujian's publishing houses, famous for their relatively cheap editions, seem to have been favored by Jiangxi's Southern Song collectors, editors, and book patrons, and not only by Liu Chu in his ca. 1150 purchase there of "500 volumes."108 In the mid-1170s a monk at a village temple in Anfu ^rlM County, Jizhou, collected money from local people to buy sutras for a temple that previously had no sutras at all; he then walked 2,000 // to Fuzhou IS'JII in central Fujian to purchase the 5,048 juan of a complete Buddhist Canon.109 Two decades later, in 1196 Lu Jiuyuan's disciples sent an agent to northern Fujian to purchase books for their private academy, "since the private academy had few books (shuji U?^)."110 And,
i"5 Wen Tianxiang, Wen Tianxiang quanji, 11.416-20.
we Yang Wanli, Chengzhai ji (2), 75.3b-4a. Their hunts appear to be for high quality and relatively expensive books—that is, rare books, books with fine paper, and
books with big characters—that would interest a keen collector.
107 Mao Deqi, Bailu shuyuan zhi, 2.1068.
108 Luo Yuan, Ezhou xiaoji, 4.37.
109 Yang Wanli, Chengzhai ji (2), 72.7a-b, "Shuyuan."
110 Huang Zongxi, Song-Yuan xuean (1), vol. 3, 20.11; Fu Xuancong and Xie Zhuo-hua, Zhongguo cangshu tongshi, 400. On this trip Lu Jiuyuan's disciple visited Zhu Xi and to his surprise heard some world-weary advice on book collecting from the contemporary most famous for his exhortations to Chinese to read books: "When during his visit Zhu Xi asked him why he had come, Peng Xingzong ^fS^c said that
90
joseph p. mcdermott
in 1261, while serving in coastal Fujian, the son of the Jiangxi-born official Xu Yuanjie f^JC^u paid for the carving in Xinghua PKfc Com-mandery of woodblocks for his father's writings.111 The persistence of the Jiangxi collectors' practice of going or sending others to order the purchase and/or production of books outside of Jiangxi speaks volumes about Jiangxi's continued lack of book publishers and merchants, its structural problems of book distribution, and its shortage of books, especially new imprints.
In sum, throughout both the Northern and Southern Song the book collectors of Jiangxi relied on a variety of government and private sources outside of Jiangxi to load their shelves. But during the Southern Song the court's institutional book collections became less central to their activities, as did also the capital's costly government and commercial imprints. New sources were found elsewhere, mainly, it seems, in northern Fujian, thus allowing the formation of a greater number of relatively large collections in Jiangxi than during the Northern Song. While collections of just a few thousand juan were still considered noteworthy, the establishment in twelfth- and thirteenth-century Jiangxi of a 30,000 juan, a 50,000 juan, and four "several 10,000 juan" collections (regardless of what number we ascribe to this vague term) bespeaks a growth in book accumulation there despite the declining centrality of government collections in the capital. This growth in scale, it needs to be added, was by and large limited to a handful of collectors and, equally significantly, to just three of Jiangxi's wealthier prefectures. In all likelihood, then, the Jiangxi collectors' varied and bumpy path to the accumulation of books provides a more representative example of how relatively significant collections were formed and preserved in most parts of Song China than does the corresponding history of book collections formed in publishing centers like northern Fujian, the lower Yangzi delta, and the Chengdu region.
he came here to buy books, since the private academy had few books." Zhu Xi replied, "If you want books, you still do not need many juan. I in the past loved them like this. Afterwards, I thought that upon assembling they will invariably disperse. What need is there to be servile to things (yi yu wu fi£^R^|)?"
111 Zhu Shangshu, Songren bieji xulu, 1317. Further Southern Song examples of Jiangxi authors' writings being published outside of Jiangxi can be found in ibid., 810 (Huizhou fift'jfl), 891 (Chizhou ttW'H), 936 (Sichuan), 1107 (Jinghunan), 1245-47 (Sichuan and Fujian), and 1344 (Zhejiang).
book collecting in jiangxi
91
VI. Uses of Book Collections
What then were these book collections intended and used for? Their owners rarely provide helpful answers to this question. They never put
on paper a vision or seldom even proposals as to what their library could or should be like. A comparison with practices in Jiangxi's rival region, the lower Yangzi delta, however, provides some insights into the priorities of these Jiangxi collectors. Firstly, for most of the seventy collectors in the table scholarship seems to have been a minor concern. In fact, Jiangxi's noted scholars—such as the Liu family of Gao'an, Ouyang Xiu, Wang Anshi, and Zhou Bida—tended to spend their collecting and reading careers outside of Jiangxi. Secondly, the practice of collating texts, so often observed for Song and later collectors in the lower Yangzi delta, is mentioned far less for Jiangxi collectors.112 Thirdly, very few Jiangxi collectors, in contrast to their counterparts in the lower Yangzi delta in the mid-Southern Song, published books themselves as family publishers (jiake WM\).n3 And, finally, biblio-philia seems to have operated at a less feverish pitch inside Jiangxi. An obsession with owning the first edition, the oldest copy, the rare titles, and the most costly version, or with discerning the difference between manuscript and imprint copies, between old editions and new imprints, and between various imprints' quality of ink and paper— these bibliophile concerns appear not to have fired the passions of the great majority of Jiangxi's book collectors. The exceptions to this observation, examples of men who spent all their wealth on books, who devoted all their time to reading and revising texts, largely come from the Southern Song, and mainly its final years.114 What instead excited the majority, at least on the basis of a few examples, was a book's provenance—especially the official status of a book's previous owners115—and the number of books they could collect.116 Their
112 Three Jiangxi scholars, two of them in our list of collectors, who somewhat
exceptionally acquired a reputation for their collation of texts are Chen Jingyuan (Lushan zhi, 9.63a), Zhang Hong (Li Yu'an and Chen Chuanyi, Zhongguo cangshujia cidian, 89), and Wang Tingzhen zEHSt (Luling xianzhi, 32.15b).
113 The most obvious exception, mentioned already, is Zhou Bida, but it is not certain that the woodblocks of his family school's imprints were actually carved in Jiangxi; see also Gu Zhixing, Zhejiang chuban shi yanjiu, 31-32, 157-60, 194-96, 207, 244.
114 E.g., Wang Bochu (Zhou Bida, Lushan lu, 73.13a-b); Wang Tingzhen (Luling xianzhi, 32.15b); Xu Bochen (Fengcheng xianzhi, 9.37b).
115 Yang Wanli, Chengzhai ji (2), 71.6a-b.
116 Xie Zhaozhe, Wenhai pisha, 7.14a.
92
joseph p. mcdermott
bibliophilia seems singularly unaffected by the scholarly bibliomania
and eccentricity that was at times endemic to the practice of book collecting in the lower Yangzi delta.117
Jiangxi book collectors, instead, proved more focused, and their relatively low-key bibliophilia underlines how a fair number of these collectors sought to use their books less to pursue their individual interests than to build an institution around the presence of a book collection, be that institution a large family and its descendants, a library, or a school. Thus, although some used books, as ever, for moral self-cultivation and, occasionally, to provide the learning required in compositions they wrote to eke out a living,118 the most commonly mentioned purpose and use of their collections was education. Usually, this education was directed to their kinsmen. Not only was Li Fen's commitment to his studies, as well as his own and his family's book collections, described in an account of his family's success in the examinations and official life,119 but also the father of Xiahou Lin K f^c^f, the scion of a twelfth-century commoner family, bought 10,000 juan to entice some highly-placed scholar-officials to come to his house and undertake some teaching.120 In addition, no fewer than ten of the forty collectors active in the Southern Song (as opposed to only two in the Northern Song) were prompted by their family's commitment to education to put up a library building to house their books.121 Usually located in a separate building near their family residences and in a scenic setting alongside a stream, these structures proclaimed the wish of their owners to preserve their books for generations. Since seven of these ten Southern Song library buildings were put up by men without
117 See the interesting denial of the concept of bibliomania by Chao Gongsu, Song-shan ji, 50.4a-b, where he claims that one can have an obsession about things in general but not about books.
118 Zhou Bida, Lushan lu, 73.13b; Fengcheng xianzhi, 9.37b.
119 Kong Pingzhong, Qingjiang san Kongji, 19.16b.
120 Yang Wanli, Chengzhai ji (2), 129.15a.
121 por tne Southern Song: Duan Liaozhong (Zhou Bida, Lushan lu, 35.5b-7a), Wang Yi ZEul (1092-1173) {Luling xianzhi, 33.3b), Zhang Hong (Hong Gua, Pan-zhou wenji, 331.16a-b), Wang Yi IEJI (1157-1202), Luo Jun (Liu Kezhuang, Houcun xiansheng da quanji, 164.8a), Zhao Buyu (Yanshan xianzhi, 8.8a-b), Luo Jingfu (Fan Fengshu, Zhongguo sijia cangshu shi, 79), Wen Yi (Jin Junqing, Jinshi wenji, 11.6b, Xu Luqing (Li Yu'an and Chen Chuanyi, Zhongguo cangshujia cidian, 95), and Chen Zongli (Ren Jiyu, Zhongguo cangshu lou, vol. 1, 789). For the Northern Song: Chen Jingyuan (ibid., 776), and the singular case of Xu Bi (Chen Boquan, Jiangxi chutu muzhi xuanbian, 58).
book collecting in jiangxi
93
a metropolitan degree, they can easily be seen as a sign of a family's claim to the attainment of literati learning.122
The owners of Jiangxi's book collections were anxious also about the perpetuation of their families. Their aim, only sometimes acknowledged by themselves or their biographers, appears often to have been the establishment and continuation of a hereditary collection of books that would instruct multiple generations of their descendants, especially for the civil service examinations.123 This emphasis on using books for their kinsmen's classroom learning suggests that these books were used more by the younger than the older members of their families. If so, the education imparted would have been at an intermediate rather than advanced level. Quite likely then, these collections had few rare books of interest to bibliophiles and scholars, and so the production of relatively few scholarly works on the basis of these collections is more understandable.
In addition, at least eight of Song Jiangxi's seventy collectors made their books available also to non-family members, as they sought to make their collection central to their area's cultural and even social activities. The earliest recorded instances of such sharing were undertaken by the three influential Jiangzhou families we have already mentioned as prospering during the first few generations of the Northern Song: the Chen, the Hong, and the Hu. In the process of establishing and maintaining large co-residential surname groups (yimen HFI) each of these families gained fame for opening schools and acquiring books to instruct their own offspring and students from elsewhere.124 The Chen family's rules of 890 backed up these goals by requiring its school to add regularly to its collection of books and to place all of them under the curatorial care of one of its students. Its 10,000 juan, somewhat exceptionally, could be borrowed, but the student curator was to register their check-out and see to their return without fail.125
A century later Wen Cao housed his book collection in the large school he had built in his native Shicheng County, Qianzhou; he invited scholars from far and near, "taking as his own responsibility
122 The three metropolitan degree holders who built their own libraries were Zhou
Bida, Xu Luqing, and Chen Zongli.
123 E.g., Nancheng xianzhi, 11.112b; Shicheng xianzhi, 6.22a (ibid., 8.44a-45a, reprints a 1747 account of this library which treats it as a moral example for contemporary book collectors).
124 Xu Xuan, Xu gong wenji, 28.5a.
125 Fu Xuancong and Xie Zhuohua, Zhongguo cangshu tongshi, vol. 1, 280.
94
joseph p. mcdermott
the instruction of disciples and the courteous treatment of guests."126 Then, in the 1070s Chen Jingyuan purchased books in Kaifeng with a gift of silver from Emperor Shenzong |$tf for his services at the court. Upon his return to Jiangxi he shared his 100 bushels—or several 10,000 juan—of books with his disciples and students.127 At roughly the same time Li Chang travelled in the opposite direction, from northern Jiangxi to Kaifeng, to take up a post in the Palace Library, and left behind some 9,000 juan for the use of others who came to his former study to read and learn.128 After another century we read of two other Jiangxi-based collectors whose notion of sharing further expanded the scope of the beneficiaries. The scholar Xiao Minwang MKat was an avid book purchaser, who gained a reputation for his willingness to share his acquisitions of Buddhist books with fellow residents of Luling as well as with his students and disciples from all over Jizhou Prefecture.129 Most remarkable, however, was the generosity of
Zhao Buyu jil^F^, a Yanshan in (_I_f County scholar in south-eastern
Jiangxi. Attached to the Hanlin Academy in Hangzhou as an Auxiliary in the Zhi fuwen ge llEfl!3£fftl (Hall for the Diffusion of Literature), he set up a library for Yanshan readers:
Previously, men in the county did not have book collections, and scholars were troubled by requests for books. Now the several ten-thousands of accumulated juan are all divided into four sections of classics, history, literature, and collectanea. Zhao set up one man to be in charge of the key and to handle the collection. Those coming were led into the book pavilion, where he set up several mats and enabled them to look freely [at the books].130
Note that the beneficiaries of this book sharing were by and large not
fellow collectors, such as one can observe at times in the lower Yangzi valley. On the whole, Jiangxi collectors were not primarily interested in forging alliances with their peers, that is, with other scholars and their collections, and thus in strengthening their ties as scholar-officials committed to a specific body of learning. They preferred to use books and lectures to win over prospective students, disciples, and
126 Li Gou, Li Gou ji, 23: 253-54; Shicheng xianzhi, 6.22a, 8.44b-45a.
127 Li E, Songshi jishi, vol. 4, 2130.
128 Su Shi, Jingjin Dongpo wenji shilue (1), xia, 53.859-61. Li Chang's collection would thus seem to have been formed inside Jiangxi, probably from works he copied at Lushan.
129 Yang Wanli, Chengzhai ji (2), 73.1a-2a.
130 Yanshan xianzhi, 8.8a-b.
book collecting in jiangxi
95
fellow readers.131 Individual Jiangxi collectors in the table thus would tend not to use their books to form close friendships or to forge an alternative identity as 'a scholar among scholars' that might weaken or replace their obligations to their family. Instead, they would use books to compete for followers who conceivably could become an extension of their family (hence, it seems, their students' frequent exemption from tuition fees).132
Thus, when a collector sought to institutionalize his books beyond the level of a library building for his family, he commonly chose to open a private school. The fact that during the Song Dynasty the Jiang-nanxi and Jiangnandong Circuits together had the greatest number of recorded private book collections as well as private academies (over 210, four-fifths of which opened in the Southern Song) is probably not accidental,133 as the acquisition of books served as the initial step for at least four Northern Song and five Southern Song book collectors in opening a private academy. Interestingly, half of these collectors lacked a metropolitan degree. It appears as if these men were seeking to overcome this social shortcoming by institutionalizing their books and schools into a large and respectable social base that enabled them to function as powerful local figures, much as if they were scholar-officials. (Some of these schools, of course, could have been set up to cloak their owners' less legitimate activities with the respectability of Confucian learning.) For instance, the mid-Southern Song collector,
131 Obviously, I am dealing here with serious collectors, not just the owners of a very small number of books, who may well have shared their odd copies with friends in both of these regions (e.g., the famous case of Ouyang Xiu's dependence on a neighboring family for books during his youth (Lti Zuqian, Ou gong benmo, 1.1a).
132 In the mid-tenth century the Luling resident Zeng Chongfan II'SkSe probably does not fit this generalization. Though poor, he had inherited copies of the Nine Classics and works of literature and history from the collection of previous generations of his family. When a prefect took them from him and then sought to recommend him for office and pay him for the books, Zeng first rejected him outright, saying, "The classics are the common implements for all under heaven (tianxia gongqi ^ ~F4aIoI). When the world is in disorder, they are stored at home; when the world is
in order, they are stored in the country____I am not a book store (shuli (jit). How
could I assess their value for compensation?" Further blandishments from the official, however, persuaded Zeng to go to the Southern Tang capital in Nanjing to be recommended for a court appointment (Luling xianzhi, 32.5a-b).
133 Walton, Academies and Society, 88-89; and Li Caidong, Jiangxi gudai shuyuan yanjiu, 56-57 and 107-15. Earlier, Chaffee, "Chu Hsi and the White Deer Grotto Academy," 46, provided a provisional figure of 425 academies throughout the Song Empire, 56 in the Northern Song and 261 in the Southern Song (108 are undated). Regardless of the precise numbers, it is clear that a considerably disproportionate share of Song private academies was located in Jiangxi.
96
joseph p. mcdermott
Dong Yi Kilt, the descendant of several generations of Confucian scholars and officials, never obtained an examination degree but opened a school before the age of thirty-two to attract to his side some famous teachers and promising students.134 In other words, these school founders, like the library builders and other book owners sharing their books in Song Jiangxi, were using their collections to establish competitive patron-client ties with examination students and, hopefully, future officials. If a book owner failed the examinations, he conceivably might have used his books and school to form the basis of a broad network of clients capable of providing him, their patron, with otherwise unavailable contacts inside officialdom.
Yet, even when housed in specially designated buildings and transformed into school collections, these libraries seldom lasted more than two generations. They fell victim to the wishes of their founders' heirs, anxious to enjoy their birthright portion of a partible inheritance. Able to circulate through subsequent sales to other collectors, the libraries nonetheless lost whatever potential they might have had on paper to lead to the establishment of long-lasting private institutions dedicated to the preservation and transmission of knowledge. These books thus failed to provide the independent ballast that Chinese institutions have often needed to survive the needs and depredation of heirs and other local parties. In this key sense, Song scholar-official libraries proved less stable institutions than their predecessors in monasteries such as that of Donglin in Tang Dynasty Jiangxi.
VII. Some Questions
The world of learning represented by Jiangxi's book collections in the Song differed noticeably from that of earlier dynasties. These
book collections were far more numerous, their holdings larger, their contents broader, their locations far more dispersed, their sources of acquisition more diverse, and their uses more varied and secular. As a result, not only was the framework of knowledge broadened, deepened, and made more specialized, but also the knowledge contained in these books spread far more widely than ever before in Jiangxi's
134 Zhou Bida, Lushan lu, 75.16a-17a. No mention is made of the examination disappointments that Dong in all likelihood suffered before launching this school.
book collecting in jiangxi
97
history, beyond the confines of the temple and monastery to many
individual and family collectors.
The cultural milieu of ninth-century Jiangxi had changed permanently, and not just for certain kinds of knowledge and their books. The expansion of Jiangxi's book collections during the Song marks an important chapter in the establishment of a certain kind of social and political elite. Tested on knowledge in those collections, this scholar-official elite unofficially adopted the book as one of its symbols in art and ritual (as, for example, the warrior rulers of medieval Europe and Japan adopted the sword as their symbol). Likewise, the private book collection, especially when it was represented by an independently standing library, became a well-recognized symbol of a family accustomed to mastering this knowledge and gaining legitimate political power without recourse to violence. This process of civilizing, so basic to the Confucian scholars' sense of purpose, seems to have become more widely realizable in Song Dynasty Jiangxi than it had been in the days when books had been concentrated in one large monastic establishment on the slope of a mountain famous for its eremites.
The wider implications of these Jiangxi book collections in the Song, I hope, will now be addressed. To build on this essay's findings and gain a deeper understanding of the dynamics of the transmission of this book knowledge and the formation of a new social and political elite, we need to ask more specific questions. How did the rise of these private collections alter the social position and activities of Buddhist monasteries and monks in Jiangxi? What alternative bodies of knowledge and lines of knowledge transmission, if any, did these new collections provide, apart from the Confucian learning transmitted by the disciples of Jiangxi thinkers like Ouyang Xiu, Wang Anshi, and Lu Jiuyuan? Did readers of the books in these collections in, for example, a relatively educated Southern Song prefecture like Jizhou acquire from them a collective framework of knowledge, of questions and answers, that came to constitute a particular "mindset"? What misfortune befell these Jiangxi book collections and collecting families, so that in the early Ming Jiangxi could be described as bereft of notable book collections and collectors?135 And, to what extent did this attested decline of Jiangxi book collections and their owners signal a change in
135 Yang Shiqi, Dongli wenji, 2.14; Hu Yinglin, Shaoshi shanfang bicong, 1.13; Fan Fengshu, Zhongguo sijia cangshu shi, 168-71.
98
joseph p. mcdermott
the social composition and intellectual commitments of this province's scholar-official elite? If these types of issues attract our future study,
the history of the book will have the power to become as central in our study of China's rich and varied past as I claimed earlier. What historian then could dismiss the concerns of book history and book collecting as merely bookish?
Table 2.1. Jiangxi Book Collectors in the Song
Name
Prefecture and county    Size of book holdings (juan)
Chen Jingyuan PjUJStH
(1024-94) Chen Xun gfcH
(js 1015) ChenZongliiL^f
(d.1270) Ding Zhong
(13th century) Dong Yi if fit
(1171-1202) Duan Chong
(1098-1175) Duan Zichong Wtrf-¥?
(aka iHJ)^ Fan Duan ffiMn
(1008-60) He LiangshuffM(
(fl. mid-S. Song) Hong Hao iM^
(1088-1155) Hu Ye iKM
(ca. 1080-?, js 1118) Hu Zhongyao #J#^§
(ca. 940-ca. 1010) Huang Tingjianlf Jill?
(1045-1105) Li Chang
(1027-90) Li Fen ^{ft (1062-82)
Liu Chu MM
(1099-1159) Liu Hang flJ?/L
(995-1060) Liu Jingzhi H'M/^:
Jianchangjun, Nancheng Jiangzhou, Dehua
Jianchangjun, Nanfeng
Longxingfu, Xinjian
Jizhou, Yongfeng
Jizhou, Luling
Jizhou, Luling
Jiangzhou, Dehua
Jizhou, Luling
Raozhou, Poyang
Ganzhou, Ningdu
Longxingfu, Fengxin
Yuanzhou, Fenyi
Nankangjun, Jianchang
Longxingfu, Xinjian Jizhou
Jizhou, Yongxin Jianjiangjun, Xinyu
several 10,000 10,000
several 1,000 up to 10,000 several 10,000
several 10,000
10,000+
10,000
10,000
5,000 or several 10,000
10,000
9,000+
10,000 (family)
several thousand (500 vols.)
has a 2 juan catalogue