As of 22 April 2009 this website is 'frozen' in time — see the current IFLA websites

This old website and all of its content will stay on as archive – http://archive.ifla.org

IFLANET home - International Federation of Library Associations and InstitutionsAnnual ConferenceSearchContacts

62nd IFLA General Conference - Conference Proceedings - August 25-31, 1996


Soren Edgren
The Research Library Group

Princeton, New Jersey, USA


RLIN (Research Libraries Information Network) is the automated information system developed and maintained by RLG (Research Libraries Group) in Mountain View, California, U.S.A. The CJK (Chinese/Japanese/Korean) component of the RLIN database was introduced in 1983 and today contains over 1,000,000 titles of modern books published in East Asian languages. In the late 1980s RLG proposed an init iative to create full bibliographic records in machine-readable form for pre-nineteenth century printed books and manuscripts in the form of an international union catalog of Chinese rare books. For a number of reasons the timing was right and, indeed, it is unlikely that a project of the kind seriously could have been undertaken anytime earlier. For example, by the last decade of this century collections of Chinese rare books had become stabilized and accessible, either in public collections or in known private libraries. New inventories and simple catalogs in book form had been produced or were underway for many of the books. A convergence of sinological and bibliographical research methodologies made possible more sophisticated levels of cataloging and at the same time demanded a greater variety of bibliographical data for scholarship, especially for emerging fields such as East Asian book history. And finally, after several years of use the RLIN CJK system had become refined enough to cope with the complexities of on-line cataloging of Chinese rare books.

Great quantities of Chinese rare books exist in the libraries of China and Japan, where most have been under some form of traditional bibliographic control, and where simple descriptions of many have appeared in various published book catalogs. Although the East Asian libraries of Europe and North America hold lesser quantities of books, they nevertheless contain many rare editions, and these o ften have not been adequately cataloged. All these disparate collections and their idiosyncratic catalogs had never previously been viewed comprehensively as a single resource that could be subjected to uniform cataloging standards, and few attempts had been made to bridge the gap between the realms of traditional Chinese rare book scholarship and modern library computer technology. It was in t his environment that the RLG Chinese Rare Books Project was conceived.

At the time, in 1988, only the National Central Library (NCL) in Taiwan had created automated records for its holdings of Chinese rare books, so the first step was to devise a conversion program to enter them in RLIN in order to review their form and contents, with the hope that they would serve as a "base file" for the proposed project. The jingbu section of Chinese "classics&quo t; (c. 1,400 titles) was successfully loaded into the RLIN database and, although the records could not be used for the kind of copy cataloging intended, the titles have served as a valuable on-line reference source. It is understood that these records have been updated and expanded, and a new priority will be to negotiate a means of converting the entire NCL rare books file for entry in RLIN in a manner consistent with the current standards of the Chinese Rare Books Project.

The next step of the so-called "pilot" phase of the project was to create an International Advisory Committee (IAC) of experts in the field to compile draft cataloging guidelines and standards, which could be tested by specialist rare book catalogers together with professional RLIN CJK catalogers. In February 1989 the first IAC meeting was convened at the Library of Congress and was at tended by various librarians and scholars, including Gu Tinglong, Director of the Shanghai Library, Peter Chang, Deputy Director of the National Palace Museum in Taiwan, and T.H. Tsien, Professor Emeritus of the University of Chicago, among others. At the same time a group of five rare book catalogers was invited from China, four from the Peking University Library and one from the library of the Chinese Academy of Sciences (Beijing), to implement the draft guidelines and enter the first trial records for the project with colleagues from the East Asian libraries of Columbia University and Princeton University. Altogether about 400 records were created on behalf of the two libraries over a period of five months. On the basis of this experience it was possible to solicit participants for the project, find a location for the central editorial office (Princeton University), and submit a cohesive plan for the funding of the project.

After funding for the first two years was secured, the Chinese Rare Books Project officially was launched in September 1991. From the start the project faced staunch challenges, such as the need to create the very guidelines by which the cataloging was to be carried out. Although the pilot phase had produced a draft set of guidelines by which the initial records were cataloged, it was up to the new central editorial office to refine them through the trial and error of cataloging the widest possible range of books. By spring of 1993, after a special IAC meeting held in Princeton, guidelines for the cataloging and creation of machine-readable records for Chinese rare books were submitted to the cataloging standards committee of the American Library Association (ALA), which formed a spec ial task force made up of western rare book specialists as well as East Asian catalogers to review the guidelines. Most of the recommendations of the ALA task force were incorporated into the guidelines and implemented during 1993 and 1994. In April of 1995 another IAC meeting was held in Princeton and the guidelines once more underwent point by point scrutiny. After some additional changes to the Chinese and English versions of the text, "Cataloging guidelines for creating Chinese rare book records in machine-readable form" will be published during the second half of 1996 in a bilingual edition. Although this painstaking review process has necessitated considerable updating and revision of previous on-line records, we feel confident that it has insured the validity of the results.

As a practical necessity, a chronological cut-off date is used by the guidelines to define the scope of the project as all "Chinese language printed books and bound manuscripts produced in China before 1796" (i.e. through the Qianlong period). 1796 has replaced the traditional date of 1644, or the end of the Ming dynasty, as a reasonable date for inclusion. Chinese language books publ ished outside of China (e.g., in Korea or in Japan), as well as non-Chinese language books published in China, are not included. The cataloging guidelines developed for the project, however, clearly allow for chronological extension or regional expansion of the scope, should that be desired at a later time.

Finding common ground between the special characteristics of traditional Chinese books and the specific conditions and requirements of on-line cataloging, without harming the integrity of either, always loomed as the chief concern of the compilers of the guidelines. Although the guidelines are intended for libraries following ALA/Library of Congress cataloging rules, and although they are based as fully as possible on AACR2 (Anglo-American Catloguing Rules, 2nd edition, 1988 revision), it is hoped that they will find a much wider audience. The following are but a few issues addressed by the guidelines that go beyond the usual comparisons of western books vs. Chinese books or manual cataloging vs. on-line cataloging. Firstly, it was deemed absolutely necessary that the juanshu (number of juan, or chapter-like internal division of traditional Chinese books) not be divorced from the title. Since Chinese books do not have title-pages in the western sense, the caption (juanduan) is preferred as the chief source of title, statement of responsibility etc., but it is declared that the entire contents of the book in hand may be used as the source of cataloging information. New stand ards of edition discrimination have been introduced, especially the means of distinguishing among unacknowledged, contemporary, variant editions. Indexing has been expanded to provide access points far beyond the traditional author, title, and subject indexes of manual catalogs, offering the capacity to search for all sorts of corporate and personal names and their various functions, even includ ing the names of former book collectors and blockcarvers. The inclusion of blockcarvers' names is an example of the kind of compromise that resulted from the collective judgment of the IAC meetings. From the start it was hoped to be able to record all the names of blockcarvers appearing in each work, but as we began to encounter titles with more than fifty names it became obvious that the labor of listing the names, as well as the six and seven page printouts that resulted, was too unwieldy for the project's objectives, and the decision was made to record representative names only from a single title. Other similar examples could be found to demonstrate how balanced and practical solutions to problems have been sought. Although the painstaking IAC review process has necessitated cons iderable updating and revision of previous on-line records, it is seen as means of insuring the validity of the results.

The value of improved access to early printed books and manuscripts by scholars can hardly be overestimated. The original planning and advisory committees recognized a number of deficiencies that needed to be overcome in order for an on-line cataloging project for Chinese rare books to be successful. The problems included a lack of uniform standards for cataloging, such as no universally accept ed criteria for bibliographic description, no collation data in printed catalogs, and unsophisticated edition discrimination. In compiling the guidelines and developing a work methodology close attention has been paid to these particular problems, and satisfying results have been achieved. A printout of an uncomplicated sample record has been appended and can be referred to as an example of a b asic bibliographic description produced by the project, in this case for a presumed unique edition of a Guangdong province local gazetteer held by the University of California at Berkeley's East Asian Library. A database catalog not only enumerates the titles of books and the names of authors, but it allows for establishing relationships among them, and as it grows the likelihood of fruitful sea rches increases immensely.

By the time the working phase of the project, begun in September 1991, came to an end in April 1996, a total of 7,495 Chinese rare book records had been created by the project and entered in RLIN. The distribution of the titles among the four main divisions of the traditional classification scheme, viz. jingbu (classics), shibu (history), zibu (philosophy), and jibu (belles-lettres), is surprisi ngly balanced, although the jingbu and jibu sections are slightly larger than the others. Altogether eighteen libraries have participated in the Chinese Rare Books Project thus far, thirteen in North America and five in China. The titles within the scope of the project held by the following libraries have been completely cataloged and entered on-line: Columbia University, Freer and Sackler Gall eries of Art, New York Public Library, University of Alberta, University of Hawaii, University of Minnesota, University of Pennsylvania, University of Southern California, and Princeton University. The libraries of the University of British Columbia, University of California at Berkeley, University of Chicago, University of Toronto, Peking University, Chinese Academy of Sciences (Beijing), Liaon ing Provincial Library, Fudan University, and Hubei Provincial Library are still active participants. Libraries expected to join the project when a new working phase is organized within the Department of East Asian Studies of Princeton University by the end of summer 1996 include the Zhejiang Library in Hangzhou, Bodleian Library at Oxford, and Bayerische Staatsbibliothek in Munich. The new pha se of the Chinese Rare Books Project is expected to produce over 2,500 new records each year which will greatly accelerate the growth of Chinese rare book records in RLIN.

The current ratio of North American to Chinese rare book records in the RLIN CJK database is two to one, but that will begin to change as North American holdings become exhausted in the next few years. It is worth noting that, unlike so-called retroactive conversion projects which convey data from one form (uaually catalog cards) to another (automated records), the Chinese Rare Books Project cat alogs for the first time or recatalogs all titles entered in RLIN, often requiring considerable research. In the case of existing catalog descriptions by qualified specialists, the project reviews the data to bring them into conformity with its unified standards, often providing biographical and bibliographical addenda. Although still at an early stage of development, it is safe to say that the future of Chinese rare book resources in RLIN looks bright, indeed.

Soren Edgren, Editorial Director

Chinese Rare Books Project