The discourse of loss in song dynasty private and imperial book collecting.

ABSTRACT

This article investigates the discourse of loss during the period between the pillaging of the Song Dynasty imperial libraries and the dispersal of private collections in north China in the late 1120s and the rebuilding of the Imperial Library and private collections through the 1140s. It contrasts the different strategies taken by the court and private collectors in managing loss, in developing acquisitions, and in remembering war and peace through collecting.

INTRODUCTION

In 960 Zhao Kuangyin (r. 960-976) established the Song Dynasty (9601270). Zhao was a general who had served under one of the many dynasties in which the Chinese territories had been divided after the fall of the Tang Dynasty (618-907). Twenty years later, in 979, Taizu, as he was officially called, fulfilled his ambition to reunify the Chinese territories. Taizu and his successor, Emperor Taizong (r. 976-97), succeeded in laying the foundations for a lasting empire, even though the Song Dynasty faced major challenges to its rule from the peoples living to its north throughout its three-hundred-year reign. In the early twelfth century the Jurchens, who first lived in the southern part of the area that became known later as Manchuria, began to pose a major threat to Song security. One of their leaders, Aguda, established a Chinese-style dynasty called the Jin Dynasty (1115-1234) in 1115. Soon afterwards, in 1126, the Jin armies invaded Song territories in the north. In 1127 they crossed the Yellow River and captured the Song Dynasty capital of Kaifeng. Jin soldiers imprisoned the retired Emperor Huizong (r. 1101-25) and the reigning Emperor Qinzong (r. 1126-27). Following the capture of Kaifeng, hundreds of thousands of remaining court servants, officials, soldiers, and commoners retreated south. Zhao Gou, Huizong's ninth son, ascended the throne later in 1127 and became known Emperor Gaozong (r. 1127-62). After moving back and forth between several cities along the Yangzi Rivet, Gaozong and his court finally settled down in Hangzhon in 1138. By then the imperial libraries and private collections had suffered major losses. Throughout the 1130s and the 1140s court librarians and private collectors set out to restore the losses sustained during the invasions and the ensuing turmoil.

After the Jin armies captured the Song Dynasty capital of Kaifeng in 1126, they gained access to Chinese books and maps and the printing blocks with which some of these materials had been manufactured. During the peace negotiations that followed, Jin envoys transmitted their emperor's desire for books published in the Song territories. For the first time since it had reunified the Chinese territories in 960, the Song court handed over books and documents that it had carefully kept from foreign eyes because, from its perspective, they contained confidential information about the Song state. In 1127 Jin soldiers carried off not only the two emperors but also a large but unknown number of books, maps, paintings, and printing blocks from the preeminent institutions of court cultural production, the Imperial Library and the Directorate of Education. They transported this cache of artifacts as well as additional materials captured from smaller court libraries on carts to the Jin capital of Shangjing (Manchuria, now Heilongjiang Province) (Z. Wang, 1165/1983, 11.19b-20a, 11.24b, 12.17a; Ren, 2001, pp. 712, 836; Winkelman, 1976, pp. 10-12). According to one very rough official estimate made shortly after the Jin invasions of 1127, about 40 to 50 percent of the Chinese books in existence in 1126 were lost in the turmoil and dislocation that attended the Song court's forced move from northern China to the south (Ma, 14th century/1986, preface, 1.32).

Losses of this kind not only affected the imperial collections; similar accounts circulated about the disappearance of 50 to 100 percent of the holdings of private collectors, although the perpetrators were not always Jin soldiers. The most famous of these accounts is the odyssey of Zhao Mingcheng (1081-1129) and Li Qingzhao (1084-1155), a collector couple who carted their collection by river and over land for about five years, discarding and losing things along the way, until only a handful of volumes remained (Owen, 1986, pp. 80-98). Other examples include the collections of Ye Mengde's (1077-1148) family, who lost about half of their collection of over 30,000 juan; the collection of Wang Zhu (997-1057) and Wang Qinchen (1034-1101), father and son, whose collection was appropriated by a general who had promised to protect it for them; and the collection of Li Chang (1027-1090), whose catalog, extant yet incomplete, still preserves the memory of some of what was lost (Fan, 2001, pp. 89, 97, 102).

This article investigates the discourse of loss during the period between the Jurchen conquest of Kaifeng and the capture of the reigning emperor in the late 1120s, and the establishment and consolidation of the southern Song court in the new capital of Hangzhou through the 1140s. Within the discourse of loss both court officials and private collectors developed and defended strategies to recover what was lost. Discussions of loss moved beyond the nostalgic mental recollection of what was gone; the memory of loss was evoked time and again at court and among local elites as part of a strategy for restoring and expanding the imperial and private collections. This article contrasts the different approaches taken by the court and private collectors in managing loss and in remembering war and peace through collecting.

RETRIEVING AND SHAPING HISTORICAL MEMORY: THE RESTORATION OF THE IMPERIAL COLLECTIONS

Wang Zao (1079-1154) participated in the retrospective compilation of official documents dating to the last two decades before the fall of Kaifeng. This experience familiarized him with official accounts of the events of 1126 and 1127. He testified that nothing remained after the Jin soldiers emptied out the Imperial Library and the Directorate of Education: "Since the history of writing, and after An Lushan's destruction of Chang'an [in the mid-eighth century], [the destruction of] the capital has never been so severe as today. All that had been amassed in the storehouses over the last 200 years has suddenly been swept away" (Z. Wang, 1165/1983, 12.17a.). Wang's testimony voiced the court's anxiety that when the Jin armies captured the Song archival record they carried with them the Song Dynasty's achievements over the past 160 years and thus its political legitimacy.

After the retired Emperor Huizong and the reigning Emperor Qinzong were taken as captives to the Jin capital of Shangjing, Zhao Gou ascended the throne in 1127. His reign lasted for more than thirty years, but Gaozong faced questions about the legitimacy of his succession to the Song throne for the duration of his rule. The new Song emperor actively sought to gather the historical records of his predecessors. In effect this was his way of rescuing Song history from its northern captors. The recovery of the Song archival record was a precondition for the Song court's continuation of its own history. It powerfully underscored the Song court's refusal to give in to the logic of the dynastic cycle according to which the succeeding dynasty wrote the history of its predecessor based on the historical records it appropriated from the latter. The recovery of the archival records further underscored the court's and high officialdom's collective ambition to rebuild Song authority on the basis of dynastic precedent.

As soon as the Imperial Library was revived, first in Shaoxing Prefecture in 1131, private donations started coming in. During 1131 alone donations were reported from the families of high officials, one local official, one general, and one examination graduate without an official position (Xu, 1809/2003, chong ru, 4.20-21). In 1131 the new Imperial Library took in partial and complete copies of the archival compilations that had been produced up to 1126, including the veritable records (abstracts of the materials collected in the combined daily calendars of an emperor's reign); the draft histories (collections of the archival records of several reigns, which served as the drafts for the dynastic history edited under each dynasty's successor); the collected statutes (classified compilations of state documents), which court offices had been regularly compiling since the eleventh century; and similar but abridged compilations of "precious instructions" of the first six reigns of the Song Dynasty, which individual court officials edited (De Weerdt, 2006). The families of high officials further contributed a wide variety of compendia on ritual. They deemed these compendia of court, bureaucratic, and family ceremonies and conventions equally symbolic of the court's intent to restore order on the basis of precedent and the ritual canon.

Court officials articulated the centrality of the archival and ritual texts to the continuity of the dynasty repeatedly in the following years. In 1132 the Imperial Library collections were moved to the new capital of Hangzhou. In 1132 and 1133 vice-directors of the Imperial Library and other court officials prompted the emperor to issue decrees requesting contributions from specific collections and collectors. In all but one case, the requests specifically demanded copies of the draft histories, the veritable records, and the collected statutes to fill in remaining gaps in the copies already collected (Xu, 1809/2003, chong ru, 4.22-23). Other genres and titles requested at this time included texts written by any of the Song emperors, the correspondence of the former court official Han Qi (1008-1075), and the records of Councilor Cai Jing (1046-1126). These texts were needed to write the history of past reigns, and their presence in the Imperial Library and its associated compilation bureaus signified not only the new regime's ability to recover lost books but also its control over the memory and the history of the Song Dynasty through the present and into the future.

The directors of the Imperial Library and other court officials concerned with rebuilding its holdings initially targeted private donors, mostly the families of prominent officials and monastic libraries. They relied on reports submitted by individual officials to identify prospects. Decrees of 1132 targeted Buddhist monasteries, which had in the past served as depositories for the works of prominent officials and writers (Xu, 1809/2003, chong ru, 4.22). In their first response to the Imperial Library's needs, officials reported on private collections in their vicinity and, within these collections, designated for court acquisition only those items that fell within the category of archival and historical collections, imperial writings, or the letters of prominent former court officials.

The method of acquisition differed. The court urged monasteries to part with the items in question. It charged the local prefect with the transfer. In decrees targeted exclusively at the collections of families or individuals, the court suggested that the owners submit the works for copying. The originals were to be returned to the owners upon the completion of their duplication. The emperor granted rewards to those who submitted titles. In the early 1130s there was no standard scale of rewards. The requests sent out in 1132 and 1133 contained references to "handsome rewards," but none specified what kinds of rewards came with what kinds of submissions. Instead, petitioning officials determined rewards on a case by case basis. Some donors received monastic certificates, ranging between five and ten. Each certificate granted the holder one tax exemption. These certificates could be used by the family itself or could be sold to others. Donors who were officials typically received a promotion in rank.

The rewards went back to precedents set in collection efforts under the second Song emperor, Taizong. In 984, Emperor Taizong decreed that those among the high officials who contributed titles listed in the eighth-century catalog of the Tang court but missing in the growing Song imperial collections should be rewarded for their contributions. For those who contributed missing volumes numbering more than 300 juan an official rank for a descendant was to be arranged depending on qualifications; those who submitted fewer juan would be remunerated based on the quantity of the materials submitted (Xu, 1809/2003, chong ru, 4.15-17). The court accepted both donations and loans and rewarded them equally. Although some court officials in the early 1130s were aware that Song emperors had rewarded gifts of books and other artifacts to the Imperial Library in the past, this institutional memory could not be fully revived during the first years of the library's reopening. In the absence of relevant archival records, ad hoc decisions immediately became precedents. One decree issued on November 29, 1131, referred back to a decree issued three months before in which an official received a promotion for his donation of one edition of the collected statutes (Xu, 1809/2003, chong ru, 4.21). It was not until 1146 that the court finally answered the demand, first voiced in 1136, for the revival of a graded scale of rewards. Emperor Gaozong then invoked Emperor Taizong's scale as a model, but it was modified in accordance with the changes that had taken place in Chinese book culture during the intervening 150 years.

Emperor Gaozong reportedly ordered that the new scale be printed for broad dissemination. It differentiated between officials and scholars (shiren). Officials who made large donations were given promotions; scholars who made large donations were either permanently or temporarily absolved from taking the prefectural civil service examinations. This gave them direct access to the triennial examinations at the metropolitan level. The addition of the category of the scholar-collector underscored the court's recognition of the importance of the rapidly expanding class of examination candidates in the circulation of cultural goods. Studies of eleventh- and twelfth-century printing and book culture strongly suggest that the hundreds of thousands of students preparing for the civil service examinations played a major role in the expansion of commercial publishing during these two centuries (Poon, 1979; Chia, 2002; De Weerdt, in press). (1)

The new reward scale fit into an effort to broaden the search for missing titles. The court complemented its orders to specific collectors with general calls for submissions starting in 1133. It repeatedly issued general calls throughout the 1130s and 1140s. With the new calls came new techniques devised to heighten the effectiveness of the collection campaign. The general calls covered the whole empire, targeted public collections and private collectors of a variety of social backgrounds, and engaged central and local government offices in the collection campaign. The Imperial Library and the Historiography Institute took the lead in the search effort. Officials attached to these central collection and compilation agencies went on occasional scouting missions (Xu, 1809/2003, chong ru, 4.25). Typically, however, they delegated this task to regional and local administrators such as the fiscal intendants, prefects, and county magistrates, instructing them to survey the public collections of prefectural and county schools, as well as the private collections (sijia) of high officials, retired and local officials, local scholars, and "the people" (minjian) or "the commoners" (shu). The list of donations preserved in the current edition of The Collected Statutes of the Song Dynasty (Song huiyao jigao) suggests that from the 1120s onwards the latter categories became more prominent as the court began to receive donations in increasing numbers from non-office holding examination graduates (jinshi) and scholars (Xu, 1809/2003, chong ru, 4.20-29). (2)

As far as the categories of materials are concerned, the general calls for donations differed from requests sent to private collections and collectors. While the latter targeted specific types of materials and often specific tides, the vast majority of which fell into the category of archival and historical compilations, the court's bibliographers showed a much broader interest in their appeals to all collectors. Materials in all four of the standard main bibliographic classes (classical texts and commentary, history, philosophical texts, and literary collections) came within their purview.

Despite the widening breadth of the Imperial Library's acquisition policies, its search was also selective. The bibliographers compared extant copies of bibliographies and catalogs of the imperial collections, such as The Bibliographical Treatise of the Tang Dynasty (Tang yiwen zhi) and The Comprehensive Catalog of Venerating Literature (Chongwen zongmu), to the Imperial Library's current holdings and posted lists of missing titles at the Petition Drum Bureau (Dengwen guyuan) and the Petition Depository Bureau (Dengwen jianyuan). Both of these offices were direct channels of communication between the court, local officials, and the common people. Local officials and commoners were allowed to submit first and second appeals to these offices when dissatisfied with the regular administrative authorities. The published lists simplified the library's work. Until 1143, when this list appears to have been first published, the library had asked local officials to submit lists of titles from which the bibliographers could then make their choices. With centralized lists of missing titles, the court potentially avoided the review of duplicate titles.

Local officials also welcomed such lists as a convenient tool in their collection efforts. One local official prodded the court in 1143 to make the lists more readily available in print (Xu, 1809/2003, chong ru, 4.26). He suggested that the Imperial Library have the catalogs printed in new editions that clearly marked which titles were missing. He explained that it was difficult for those in outlying provinces to copy the lists by hand and that lack of knowledge of the missing titles was likely to impede ongoing local collection efforts. Emperor Gaozong approved this request and issued a decree ordering the distribution of a printed list of missing titles to all prefectures in the same year. The Imperial Library published The Continuation of the Catalog of the Imperial Library Indicating Missing Books in the Four Repositories (Bishusheng xu biandao sihu que shumu) in line with these efforts in 1145 (Van der Loon, 1984, pp. 12-14).

The printing and dissemination of the list of missing titles in 1143 was part of a consistent effort to render the court's collection campaign more effective. Based on the list of donations in The Collected Statutes of the Song Dynasty, a steady flow of donations came in between 1131 and 1135. With the exception of the year 1134, the Imperial Library collected several thousand juan annually during the first five years of the campaign. The numbers of registered donations, however, showed a downward trend, and this decline was a sign of the waning interest among donors. The number of donations never again reached the level attained in 1131. Only one donation, consisting of one title, is on record for the period between 1136 and 1142.

In 1143-44 the collection was moved into a new building. In preparation for the move, it became apparent that the Imperial Library was still lacking a substantial number of titles that were, according to its librarians, available in private collections. Private collectors were thought, however, to be unwilling to contribute their treasures in the absence of clear and substantial rewards. An upsurge in the collection campaign followed to address these problems.

The renewed interest in developing the collection in the early 1140s was shaped by changes in the political climate. After a decade of reconstruction in the 1130s, court policy toward the Jin Empire gradually shifted toward a pro-peace stance. The nomination of Qin Gui (1090-1155) to the position of councilor in 1138 marked the beginning of the ascendancy of the pro-peace camp. Qin Gui had been captured by Jin soldiers along with 3,000 members of the imperial family and entourage in 1126. Originally an advocate of resistance against the Jurchen advances in the north, he began to espouse a pro-peace stance during his captivity. He escaped from his Jurchen captors and returned to the Southern Song court in 1130. At the court of Gaozong, he advocated peace with the Jurchens in the north at whatever cost. Emperor Gaozong, whose position on the throne depended on the continued captivity of the former emperor Qinzong, became gradually convinced that a pro-peace policy would guarantee the consolidation of his rule over the southern territories. Between 1138 (when the peace conditions were announced) and 1155 (the year of Qin Gui's death) peace with the Jurchens was upheld as the "court line" (Yu, 2003, 1:373). Voices transgressing this line were the object of political suppression.

In 1140, as part of a series of attempts on now Grand Councilor Qin Gui's part to fashion the library's collection and compilation activities according to the political needs of the pro-peace faction of which he was the leading proponent, he centralized the operations of the court's collection and compilation agencies. He abolished the Historiography Institute and housed its operations under the Imperial Library (Huang, 1993, p. 52; Hartman, 1998). At the beginning of 1142 he arranged for his brother Qin Zi (?-1146) to be appointed to the vice-directorship of the library. Six months later Qin Xi (?-1161), Qin Gui's adopted son, became assistant in the library. Within another three months Qin xi was appointed to the vice-directorship, and in 1144 he became director of the Imperial Library. Through this reorganization and the subsequent restaffing Qin Gui established control over the historical memory of the dynasty. The library's directors and assistants now decided what would be collected and what would go into the compilation of the dynasty's archival and historical record.

Capitalizing on the political authority of his father, Qin xi pushed through a series of campaign reform measures. First, he established the fixed scale of rewards that Emperor Gaozong approved in 1146 and had it printed and distributed to lower officialdom. He called for the submission of missing book titles as well as pieces of calligraphy and painting, which were also stored in the library. He insisted that the 1146 reward scale should only be used for donors whose books and artifacts had been reviewed and deemed rare at the library. Officials were only to compensate owners for those titles that were listed as missing or otherwise considered unique. This was a precautionary measure aimed at eliminating the deceit of those collectors and officials who donated common materials or stuffed book cases with heterogeneous materials to reap monetary rewards.

Second, Qin Xi exerted pressure on local officialdom to take the Imperial Library's collection campaign seriously. In a report submitted in 1145 he charged that the gaps in the imperial collection were not due to the lack of leadership on the part of the central government but rather to the lackluster implementation of central directives in the prefectures and the counties. He noted that the collection of missing and rare materials had not been a high priority in local administrations and argued that this could be changed by providing clear incentives to local administrators. He demanded that the Imperial Library be given the authority to reward productive administrators and to penalize those who continued to ignore its directives. The emperor granted him such powers; Qin Xi's proposal that prefects be asked to submit regular reports on their collection efforts was accepted (Xu, 1809/2003, chong ru, 4.27-28).

Qin Xi's efforts seem to have borne immediate effect. The collecting campaign went through a second upsurge between 1145 and 1148. Two collections of several hundred juan arrived in the first year, followed by a large collection of 2,990 juan in 1147; all of these titles were listed as missing in the court's catalogs. The donors were rewarded in accordance with the new scale, except for some minor modifications (Xu, 1809/2003, chong ru, 4.25-29).

The librarianship of Qin xi led, according to Mo Shuguang (jinshi, 1163), who directed the Imperial Library in the mid-1180s, to the restoration of the imperial collections (Xu, 1809/2003, chong ru, 4.31; Chen, 1178/1998, p. 174). Qin Xi's policies, however, illustrated not only the contributions to but also the detractions from the memory of war in book collecting. The pursuit of peace ironically resulted in the persecution of those who cherished the memory of war as policy. Qin Xi's demand for greater authority for the Imperial Library in its supervision of the collection efforts of local administrators was not simply an answer to a collection campaign gone dormant; it fit into a concerted effort to scrutinize private collections and weed out those materials that were deemed incompatible with the new regime's pro-peace stance. The defense of the pro-peace policy required in Qin Gui's view the support of a history of the preceding decades that made the support for peace with the Jin Empire the preferred diplomatic policy and him the untainted protector of the Song Dynasty's best interests. The reinvigoration of the collection campaign coincided with the imposition of a ban on privately compiled histories of contemporary events. Qin Gui proposed such a ban in 1144 and received the emperor's support for a larger project to ensure the preparation of a uniform account of the history of the restoration of the Song Dynasty (Hartman, 1998).

Several cases have come down to us that illustrate the additional loss of books that resulted from the ban on private histories and the collection campaign of the early 1140s. Even though the collection efforts thrived due to the memories of war as an agent of loss recounted at the beginning, they were also a response to literate elites who kept alive the memory of war as a policy to be continued. According to contemporary accounts the families of those who had been advocating war with Jin and the recapture of the northern territories burned entire collections or, more frequently, letters, memorials, and historical accounts that could be deemed subversive in the eyes of local administrators or librarians at the Imperial Library committed to the pro-peace policy. The materials were in several cases burned as a matter of precaution, typically when the owners learned about the persecution of a like-minded soul.

The most famous case was that of Li Guang (1077-1155). Prior to the capture of Emperor Qinzong in 1126, Li Guang opposed negotiating territory with the Jurchens. In the 1130s he became an advocate for strengthening defenses along the Yangzi River, which had become the de facto border between the Song andJin Empires. He agreed to serve as assistant councilor under Qin Gui in 1139. When the latter began to remove from office all generals who had served in the war against Jin, Li became a harsh critic of Qin Gui's pro-peace diplomacy and argued that the Jin forces could not be trusted to abide by peace regulations. He was dismissed from office one year later. In 1147 Li Guang's family burned his library of at least 10,000 juan (M. Wang, 1195/2000, 7.174; Tuo, 1345/1977, 473.13760; Fan, 2001, p. 104). (3) The collection of Wang Zhi (?-1145?) suffered a similar fate. He was dismissed from court around the same time as Li Guang. Qin Xi's call to local administrators first resulted in the alleged confiscation of over half of the collection estimated at 25,000 juan. In 1147 anxious relatives burned the other half of the collection (Hartman, 1998, pp. 93-94, 99-102; Fan, 2001, pp.106-8). (4)

The combination of active collecting and censorship instilled fear and paranoia among those related to or associated with officials ousted by Qin Gui. Soon after the death of Councilor Zhao Ding (1084-1147), who had been exiled to Hainan Island in 1147, his son burned the entire family collection of books and weapons (Huang, 1993, p. 60). Not all those fearing the impact of the Qin family's cultural policies resorted to such extremes. Xue Jixuan (1134-1173) wrote that his father hastily removed those pieces in his collection that contained prowar sentiments and criticisms of the propeace policy and hid them in a separate cabinet. The effect of the campaign was reduced in this case, but Xue still blamed it for the dismal state in which he later discovered his father's writings (Huang, 1993, pp. 60-61).

LOSS AS OPPORTUNITY: PRIVATE COLLECTORS AND PRINTING IN THE TWELFTH CENTURY

When remembering the losses that resulted from warfare and the forced migration in the 1120s and 1130s, private collectors shared a common historical analogy. Several of them referred back to the destruction of private collections under the First Emperor of Qin in 213 BCE (Petersen, 1995). They were united in their estimation of the scale of the losses that resulted from this deliberate destruction, but they connected it to the subsequent restoration of the classical textual legacy. They attributed this restoration to the collective efforts of the community of Han Dynasty scholars. In Ye Mengde's words,

   I reflected on the fact that in the beginning of the Han
   Dynasty [ca. 200 BCE] it had not been that long since the
   time of Confucius, and, yet, after the chaos brought on by
   the Qin, fifty-one chapters of The Book of Documents had been
   lost, six chapters from The Book of Songs, and the sections
   "The Officials of Winter" in The Rites of Zhou had been
   entirely lost. If this was the case for the classics, it must
   have been much worse in all other categories! Fortunately, the
   rest had been preserved in the collections stuck in the walls
   of private homes. What has been preserved until today is the
   result of scholars upholding [these texts'] transmission. (Ye,
   12th c./1983, 4.1b) (5)

Beyond their personal losses (well over 10,000 juan in Ye Mengde's case), private collectors, following the Song migration south and the associated loss of texts, demonstrated grave concern over the impact of these losses on the cultural memory of the literate elite. Their response to the losses occasioned by the Jin conquest of the northern territories accordingly fell into two broad categories. First, they engaged in the same kinds of activities, albeit on a smaller scale, as the librarians at the Imperial Library. Several private collectors who had lost substantial parts of their collection in the 1120s and 1130s were inspired to undertake heroic acts of book collecting. Both Ye Mengde and Jing Du were known to have built their collections through the acquisition of handwritten copies (Fang, 1999, pp. 292-304; Ren, 2001, 1:805-6). The epitaphs written for collectors typically laud the care they took in collating their newly copied editions. These practices were a continuation of shared acquisition methods. Collectors who had been faced with the occasional burning or looting of individual collections resorted to the generosity of colleagues in reconstituting lost titles. More generally, copying by hand and collation were practices associated with the image of the true scholar-cure-collector (McDermott, 2005, p. 65; Y. Wang, 2005).

Second, private collectors differed from the court librarians and their superiors in their enthusiasm for private and government printing as a way to ensure the future preservation of cultural memory. Accounts of loss and the lack of scholarly sources in the decades following the Jin conquest of the north were frequently a pretext for the mobilization of private and government moneys for the printing of a wide variety of texts. Ye Mengde related his own efforts in this area in an inscription for "The Pavilion for the Assemblage of Books" (Choushu ge), a library he established for the common use of scholars in Jiankang in 1140:

   In the beginning of the Shaoxing period, I became prefect
   [of Jiankang]. After the major military upheavals, we encamped
   the soldiers in linked camps. The walls of the city were thickly
   overgrown with thorns and brushwood; the scholars' respect for
   ritual had not yet returned. When I searched for The Changes, I
   could not find it. Therefore, I greatly feared that the sacrifices
   would be abandoned. In order to establish a school and invite
   students I freed up 600,000 strings of cash from the military
   budget. With this money, I paid teachers and had the six
   classics printed....

      Nowadays everywhere we are focused on what was lost. With every
   title that is engraved on woodblocks, the number of good things
   increases. It behooves us to shave and spread this book collection
   in order to be prepared for all eventualities. It so happens that
   we have 200,000 strings in reserve in the prefectural coffers. I
   do not dare to spend it on anything else, but use it to buy
   classics and histories. Overall we have been able to acquire a good
   number of juan. (Ye, 12th c./1983, 4.1a-b)

Ye Mengde undertook his school- and library-building activities while serving as an acting local prefect in Jiankang. His zeal for these projects and the large amounts of cash he allocated for them, however, were inspired by the difficulties he suffered as a scholar and collector to gain access to books. His emphasis on the lack of even the most basic texts such as The Changes provided justification for his printing projects. He envisioned his ventures in printing as contributions to a larger enterprise aimed at the preservation and dissemination of scholarly texts. Even though he voiced strong feelings about the decline in philological skills attendant upon the increased use of print technology, he shared the observation made by other contemporaries that print facilitated access to the basic sources of scholarly discourse. By extension, the wider distribution of print editions also increased the survival rate of texts. It did so even more effectively than the earlier Han scholars' copying of texts that had been recovered "between the walls of private homes."

The connection made in Ye Mengde's inscription between the interest in things lost and their recovery in print is also evident in the post-conquest activities of other private collectors. Jing Du, fiscal intendant for Sichuan in the early 1140s, allegedly spent half of his salary on the acquisition of books. Sichuan, a major player in commercial printing during the Song Dynasty, had suffered less destruction than the areas in the north and further east along the Yangzi River. Jing Du, therefore, was able to acquire substantial numbers of books. At the same time, he turned his attention to the dissemination of some of the materials he was acquiring. After he had collected a complete set of the seven dynastic histories from various sources, he had the histories engraved on woodblocks and distributed throughout Sichuan and beyond (Chao, 1151-1240s/1983, 2A.6b-7a).

As documented in Lucille Chia's study of the commercial printers of Jianyang in northern Fujian, the twelfth century witnessed an unprecedented growth in printing, with the number of printing centers in Song territories increasing from 30 before the conquest of the north to around 200 thereafter (Chia, 2002, p. 66). The activities of Ye Mengde and Jing Du suggest that the development of commercial and private printing during the Southern Song period may have been boosted by the memory of lost books as well as by the increasing numbers of students preparing for the civil service examinations.

Moreover, private and commercial publishers also continued to circulate those texts that were weeded out during Qin Gui's collecting campaigns. The best example in this regard is the first private history to be banned under Qin Gui, A Record of Rumors by Sima Wengong (Sima Wengong jiwen; later also known as A Record of Rumors from Su River, or Sushui jiwen). It continued to circulate independently and in excerpts throughout the twelfth century. This book was in all likelihood based on notes about court events taken by Sima Guang (1019-86) in the late eleventh century (Sima, 11th c./1989, preface). He took the notes in preparation for a history of the Song Dynasty, which was intended as a sequel to his more famous survey of Chinese history up to 960. This notebook was not printed until around 1145 in the commercial printing center of Jianyang. It was then part of the general interest in the recovery of materials that had allegedly become rare as a result of the war with Jin. Prior to its first documented commercial circulation, in 1136, Councilor Zhao Ding (1084-1147) asked for and received imperial permission to assign the noted historian Fan Chong (fl. 1090s-1130s) the task of recovering and arranging these notes. Zhao Ding noted that copies of Sima Guang's notebook of court events had become sparse since the outbreak of war, but he expressed hope that it could be recovered from the private collections of scholar-officials.

One decade later this private notebook became suspect because of its connection to Zhao Ding, who had been ousted by Qin Gui in 1140 and had become a symbol of the prowar faction. Sima Guang's great-grandson, Sima Ji, feared that the commercial publication could implicate his family as Qin Gui called for a ban on private histories. He submitted a memorial denying any link between this work and Sima Guang and asked that the book be banned and the printing blocks burned. His request was granted and his effort rewarded with a promotion, even though this cost him the ridicule of contemporary scholars, many of whom dismissed any doubt concerning its attribution. The emperor ordered the prefect of Jianyang to burn the woodblocks, but private catalogs, histories, biographical collections, and commercial encyclopedias from the twelfth and thirteenth centuries referred to it and quoted from it, suggesting that the scholars had access to it and remained committed to the transmission of cultural memory broadly conceived.

The enthusiasm among literate elites for the reprinting of rare titles was so extensive that in 1186, some forty years after Qin Gui's revamping of the imperial collecting campaign, the director of the Imperial Library alerted the emperor that his collections needed to catch up with the many rare titles that had been reproduced in print locally in the aftermath of the Jin conquest:

   In the beginning of the Shaoxing reign [early 1130s] a decree was
   issued to borrow and collate books. By 1143 a decree was issued to
   search for missing titles. And in 1146 a scale was set up to reward
   those who donated books. Thereafter the [imperial] book collection
   became complete. However, by now, another forty years have passed.
   During the long period of peace, people all over the empire have
   increasingly come to hold books in high regard. The fiscal
   intendants and prefects searched and obtained all the rare items in
   the collections of the scholarly elite. They often engraved them on
   woodblocks and produced government editions. I ask that a decree be
   issued to all intendants and prefects asking them to submit a
   catalog of the books produced in their jurisdiction to the Imperial
   Library. This Library should compare the catalogs to The Catalog of
   the Imperial Libraries during the Period of Restoration. If there
   are items that are not held by the Library, an order should be sent
   down to the relevant jurisdiction to obtain it. We hope that this
   will expand the collection of the Library (Xu, 1809/2003, chong ru,
   4:31).

CONCLUSION

Imperial librarians and private collectors described losses to their respective collections in similar terms. While their methods overlapped to an extent (both used catalogs to find rare items and relied on hand copying to collect large numbers of items), their goals diverged. The imperial librarians' efforts were guided by the imperative to reconstitute the Song Dynasty's collections and its historical memory. By the 1140s the memory and politics of war steered the efforts to recover the imperial collections in opposite directions. On the one hand, the memory of war as loss motivated a very ambitious recovery program. On the other hand, the ongoing politics of war and peace established informal guidelines for acquisitions and inspired a campaign against materials that opposed the court's propeace stance.

The recovery campaign was relatively successful. Within five decades the court was able to restore the Imperial Library's holdings to its previous level of around 50,000 juan. While new acquisitions figured in this number, the court's success also derived from a concerted effort to collaborate with private collectors, local officials, and commercial printers to acquire copies of lost books and archival collections. Librarians used extant and newly compiled catalogs of the Imperial Library as checklists for the recovery campaign and shared them with local officials. The expansion of print culture in the twelfth century aided the librarians' efforts. The tens of thousands of literati preparing for the civil service examinations fueled the expansion of commercial printing in different parts of the Song Empire, which, in turn, led to a substantial increase in the number of private collectors. The court encouraged private collectors to make voluntary contributions or to make their holdings available for copying. The volume of such contributions picked up as the court developed a scale of rewards based on the rarity and quantity of the contributions. Accompanying legislation provided among other things for the establishment of professional copyists attached to the local bureaucracy. Commercial printers who cut new blocks duplicating lost editions issued by the Directorate of Education were required to submit these and other titles to local and court authorities for approval and deposit (Ozaki, 2003).

Despite the recovery campaign's successes, the influence of factionalist debates over war and peace also resulted in further loss. Personal and political grievances led to the confiscation of private collections or their burning by fearful or recalcitrant owners. Legislation prohibiting the publication and possession of materials touching upon the question of war and peace also resulted in confiscations.

Private collectors associated loss with the scholar's responsibility to transmit and preserve cultural memory. The task, prefigured in Han Dynasty scholars' recovery and transmission of the classical legacy, was in their view facilitated by print technology. The private experience of loss justified large-scale printing projects. The twelfth-century enthusiasm for printing among local officials, private collectors, commercial publishers, and scholars in general was thus in part a response to the memory of war as loss. Despite Qin Gui's legislation, editors, printers, private collectors, and scholars continued to invest in a wide variety of materials, including those that celebrated or embodied the memory of war as policy.

REFERENCES

In references to pre-modern Chinese sources, the first date refers to the first date of publication, the second to the date of the edition used. Dates are followed by division title if applicable, juan (fascicle/chapter) number if applicable, and page number followed by a and b for front and reverse sides of the page if applicable.

Throughout the references, the Siku quanshu edition will be referred to as SKQS.

Chao, G. (1151-1240s/1983). Junzhai Dushuzhi. SKQS.

Chen, K. (1178/1998). Nan Songguan'ge lu, xu lu. Beijing: Zhonghua shuju.

Chia, L. (2002). Printing for profit: The commercial publishers of Jianyang, Fujian (11th-17th centuries). Cambridge, MA: Harvard University Asia Center.

De Weerdt, H. (2006). Byways in the imperial Chinese information order: The dissemination and commercial publication of state documents. Harvard Journal of Asiatic Studies, 66(1): 145-188.

De Weerdt, H. (in press). Competition over content: Negotiating standards for the civil service examinations in imperial China (1127-1279). Cambridge, MA: Harvard University Asia Center.

Fan, F. (2001). Zhongguo sijia cangshu shi. Zhengzhou: Daxiang chubanshe.

Fang, J. (1999). Songdai da cangshujia Ye Mengde de cangshu houdong ji qi dui tushnye de gongxian. In Huang Jian'guo and Gao Yuexin (Eds.), Zhongguo gudai cangshulou yanjiu (pp. 292-304). Beijing: Zhonghua shuju.

Hartman, C. (1998). The making of a villain: Ch'in Kuei and Tao-hsueh. Harvard Journal of Asiatic Studies, 58(1), 59-146.

Huang, K. (1993). Qin Gui yu wenzi yu. In Huang Kuanchong (Ed.), Song shi eonglun (pp. 41-73). Taibei: Xinwenfeng.

Kong, Q. (1355/1987). Zhizheng zhi ji. Shanghai: Shanghai guji chubanshe.

Lu, Y. (1190s/2000). Laoxue'an biji. In Hanji quanwen ziliaoku. Taibei: Zhongyang yanjiuyuan jisuan zhongxin.

Ma, D. (late 13th-early 14th c./1986). Wenxian tongkao--jingji kao. Taibei: Xinwenfeng.

McDermott, J. (2005). The ascendance of the imprint in China. In Cynthia Brokaw and Kaiwing Chow (Eds.), Printing and book culture in late imperial China (pp. 55-104). Berkeley: University of California Press.

Owen, S. (1986). Remembrances: The experience of the past in classical Chinese literature. Cambridge, MA: Harvard University Press.

Ozaki, Y (2003). Songdai diaoban yinshua de fazhan. Gugong xueshu jikan, 20(4), 167-90.

Petersen, J. (1995). Which books did the first emperor of China burn? Monumenta Serica, 42, 1-52.

Poon, M. (1979). Books and printing in Sung China (960-1279). Unpublished doctoral dissertation, University of Chicago, 1979.

Ren, J. (2001). Zhongguo cangshulou. Shenyang: Liaoning renmin chubanshe.

Shi, S. (1202/1983). (Guiji zhi. SKQS.

Sima, G. (11th c./1989). Sushui jiwen. Beijing: Zhonghua shuju.

Tuo, T. (1345/1977). Song shi. Beijing: Zhonghua shuju.

Van der Loon, P. (1984). Taoist books in the libraries of the Sung period: A critical study and index. London: Ithaca Press.

Wang, M. (1195/2000). Huizhu lu. In Hanji quanwen ziliaoku. Taibei: Zhongyang yanjiuyuan jisuan zhongxin.

Wang, Y. (2005, April). "Ten thousand scrolls": Texts, textuality, and the development of critical discourse on poetry in the late Northern Song. Paper presented at the workshop "The Early Development of Print Culture in China," Fairbank Center for East Asian Research, Harvard University, Cambridge.

Wang, Z. (ca. 1165/1983). Jingkang yaolu. SKQS.

Winkelman, J. (1976). The Imperial Library in Southern Sung China, 1127-1279: A study of the organization and operation of the scholarly agencies of the central government. Transactions of the American Philosophical Society, 64(8).

Xu, S. (ca. 1809/2003). Song huiyao jigao. Taibei: Zhongyang yanjiuyuan jisuan zhongxin.

Ye, M. (12th c./1983). Jiankang ji. SKQS.

You, M. (ca. 1190/1983). Suichu tang shumu. SKQS.

Yu, Y. (2003). Zhu Xi de lishi shijie: Songdai shidafu zhengzhi wenhua de yanjiu. Taibei: Yun chen.

NOTES

(1.) The new standard for the distinction between major and minor donations set in the 1146 decree further illustrates the court's awareness that private collections had grown substantially since the late tenth century. The 300 juan bar, changed to 500 juan in the eleventh century (Xu, 1809/2003, chong ru, 4.17), was raised to 2,000 juan. This number does not reflect the average size of private collections but rather the total number of unique materials that a private collector was able to contribute to the Imperial Library.

(2.) The list includes the names of eight jinshi who made donations between 1131 and 1155; a handful of donors were solely represented by honorary titles given in recompense for their contributions, suggesting that they as well had not held prior office. In its record of donors and donations, the court did not list "commoners" as such. It only used honorary titles in reference to donors without official or examination ranks in the very documents that conferred the titles. For example, Zhuge Xingren, who in 1135 donated 8,546 juan according to one account or 11,515 juan according to another, was referred to by the bureaucratic title of "Judicial Investigator of the High Court of Justice" in the official list of donors compiled at court. According to a local history of Shaoxing compiled in 1202, Zhuge Xingren was a "plain-clothed" commoner (buyi) when he made the contribution (Shi, 1202/1983, 16.29b). The sources further suggest that some of the men listed as jinshi may have been examination candidates rather than examination graduates. Emperor Gaozong granted three of them exemption from the local examinations in exchange for their collections, the reward reserved for non-office holders according to the 1146 scale of rewards. Although such rewards were transferable to male relatives, it is possible that the designation jinshi functioned here, like the honorary titles, to upgrade the status of the donors. There are other cases where scholars arrogated to themselves the title of jinshi. For one example in the business of editing and printing, see Chia (2002, p. 90).

(3.) Both Wang Mingqing and the author of Qin Gui's biography in The Song Dynastic History estimated the collection at around 10,000 juan. According to Kong Keqi, the Li family's collection of books and inscriptions amounted to several tens of thousands of juan (Kong 1355/1987, 2.39).

(4.) According to Lu You, Qin xi requested that the family donate its collection after Wang Zhi's death in exchange for an official appointment. His son, Wang Lianqing (1127-1214) refused the offer and Qin Xi was unable to make the acquisition (Lu 1190s/2000, 2.20).

(5.) For a similar interpretation of the contribution of Han scholars, see You (1190/1983, la-b).

Hilde De Weerdt is assistant professor of history at the University of Tennessee at Knoxville, where she teaches premodern Chinese history. She has a long-standing interest in the organization and dissemination of information. She has written on Chinese encyclopedias, local gazetteers, and archival compilations. Her manuscript on the intellectual history of the Chinese civil service examinations, "Competition over Content: Negotiating Standards for the Civil Service Examinations in Imperial China (960-1279)," is forthcoming from the Harvard University Asia Center. She is currently working on a project on news and identity in late imperial Chinese history. This article is part of a more general research interest in the history of book collections (both court and private) in premodern China.