3rd National Roundtable on Libraries & Asia
Issues Papers
National Bibliographic Access To Asian Materials
Introduction
This paper discusses bibliographic access to the Distributed National Collection of Asian materials, and proposes recommendations which would enable us to answer the question 'What Asian materials are held in the collections of Australian libraries, and where are they?' The National Bibliographic Database is advocated as the primary tool for bibliographic access to Asian materials in Australian libraries, and the paper reviews the coverage of the NBD, the relationship between the ABN and CJK databases, and ways in which the NBD may be enhanced. The recommendations are intended to enhance the coverage and effectiveness of the NBD for Asian materials.
Coverage of the National Bibliographic Database
The National Bibliographic Database (NBD) is currently a single database accessible via the Australian Bibliographic Network. However, the National CJK Project will involve the building of a National CJK Database, which will be the CJK arm of the NBD. There will be gateways between the ABN and CJK systems to provide access to both databases. The CJK database will be the first extension of the NBD towards a distributed national bibliographic database.
There have been suggestions from time to time that the NBD comprises all databases and catalogues in Australian libraries, ranging from valuable card catalogues to online catalogues accessible via the Internet. However, such a definition is still more notional than practical at this stage. Certainly, it is possible to search most of the catalogues of Australian university libraries one by one on the Internet, but we do not yet have the ability to perform a global search and find data which is linked to national interlibrary loan and document supply services and billing arrangements. For the time being, the most practical way to achieve bibliographic access to individual titles in the Distributed National Collection in a one stop look up is the NBD on ABN, and soon, the NBD on ABN and CJK.
Current Contributions to the NBD
The results of the DNC Asian Library Materials Survey question 8 'What percentage of your library's Asian library materials has been reported to the NBD?' show a splendid performance by Australian libraries. Contribution of data for current materials to the NBD is very high among libraries which contribute regularly to the NBD, and the National Library is aware that all but one of the other libraries are currently planning to contribute data in the future.
However, it should be said that the DNC Survey masks the issue of timeliness of these contributions. Libraries which catalogue their current acquisitions on ABN, and eventually the CJK system, achieve a much more timely contribution than libraries which contribute cataloguing records in batch files. It would be reasonable to look to the NDIS to fully rectify this situation by offering libraries the facility to upline load their data for rapid updating of the NBD. However, until the implementation of the NDIS, the best currency will be achieved by online cataloguing on the ABN and CJK systems, with batch reporting offering less currency. Timeliness is critical for some aspects of DNC management which require national consultation between libraries, such as the cancellation of serial subscriptions. Timely contributions allow libraries to take quick decisions on the acquisition of expensive items, or to obtain rapid physical access to the collections of other libraries, or to coordinate preservation and retention policy between Australian libraries.
Recommendation 1:
Australian libraries with significant Asian collections should continue to contribute holdings data for current acquisitions promptly to the National Bibliographic Database, including the National CJK System when it is implemented. Where online cataloguing is not feasible, libraries should contribute current update files twice a year until the implementation of upline loading in the NDIS.
Retrospective Contributions to the NBD
The DNC Survey shows that coverage of retrospective holdings on the NBD is stronger than anticipated, with the older libraries having significantly less coverage than the newer libraries. This is typical of mainstream patterns of data contributions to the NBD, and reflects the fact that older libraries have a significant proportion of their data in non-electronic formats, with the data reflecting a wide range of standards used during their history. On the other hand, almost all the libraries with a low level of retrospective contribution to the NBD have full bibliographic access to their Asian materials in their local catalogues, so there is a good foundation for special projects to convert this data for contribution to the NBD. The retrospective coverage of these collections on the NBD will need to be boosted by special funding if libraries as a whole are to enjoy national bibliographic access to them. This is particularly true for libraries whose Asian materials are catalogued in non-electronic formats.
Priorities for Enhancing the NBD
Question 9 of the DNC Survey asked libraries to describe the level of cataloguing for their Asian library materials, and question 11 asked libraries 'What materials would you nominate as the highest priority for loading to the NBD, either on ABN or the CJK system when it is implemented?'. The responses to both these questions can be used as a source of ideas for identifying the most important Asian data files for loading to the NBD. Unfortunately question 11 was ambiguous, for which we apologise, and produced a mixture of responses covering both local and national perspectives. Most libraries seemed to respond with a view on their own library's priorities for loading to the NBD, and if any pattern can be detected, it seems that Japanese language, Chinese language, and serial publications are the highest priority for most libraries. The importance of serials is reflected in the resolution of the 1993 Roundtable calling on the National Library to produce a list of Asian newspaper titles from ABN and the Newspapers in Australian libraries files. While such a list has been compiled, it is a very poor substitute for libraries contributing their Asian newspaper holdings to the NBD. No special searches or listings derived from outdated sources will substitute for timely and accurate reporting to the NBD. However, the National Library will distribute the list of locations from the outdated Newspapers in Australian libraries file, which might not otherwise be available to libraries, with the exhortation that it be used in conjunction with online searching of the NBD.
Taking a different approach to identifying the priorities for enhancing the NBD, Flinders University interpreted question 11 with a national perspective, and nominated three known 'gaps' in NBD coverage: ANU holdings for Asian materials, NLA holdings for Indonesian materials, and data about South Asian materials nationwide.
It would be useful to use the responses to these survey questions to develop a consensus on what are the most important data files for improving bibliographic access to Asian materials nationally. The resulting list could be used as a basis for seeking funding for special retrospective projects which promote national access to the most valuable library collections. Clearly the ambiguous nature of the questionnaire does not allow a full list of important projects to be constructed from the responses, but the following list is offered merely as a starting point for discussion:
- Serials, including newspapers- contribution to NBD
- Japanese language - contribution to NBD
- Chinese language - contribution to NBD
- South Indian materials - contribution to NBD
- Retrospective conversion of valuable card catalogues to NBD, e.g. University of Sydney
- ANU data files - contribution to NBD
- NLA's uncatalogued Indonesian materials - contribution to the NBD
If such a list could be used as the basis for ARC funding proposals, the benefit would be twofold: firstly, the application would carry the weight of national consensus from the Roundtable, and secondly the competition between different proposals would be reduced.
Recommendation 2:
That the Roundtable attempt to achieve a consensus on what are the most significant files of cataloguing data which should be loaded to the NBD, with a view to seeking ARC funding for significant costs associated with data preparation, data upgrading (eg with vernacular script) or format conversion.
On the question of ARC funding proposals, a critical aspect often omitted from funding proposals for the creation or conversion of bibliographic data, is the contribution of that data to the National Bibliographic Database. It is suggested that any proposal which involves the cataloguing of library materials, or the conversion of card catalogues to machine readable form, should include contribution to the NBD as an integral part of the proposal, so that the project may be completed in a way which promotes national access to these materials for cataloguing, reference and interlibrary document supply. For example, cataloguing projects should create data which conforms to the minimum standard for contribution to the NBD, or should include conversion of data to meet that standard. They should also make provision for costs associated with loading the data to the NBD, where those costs exceed the National Library's usual levels of support for file loads.
Recommendation 3:
That any proposals for ARC funding of projects for the cataloguing or retrospective conversion of data for Asian materials should always include provision for contribution of data to the NBD as an integral part of the project.
Relationship between the ABN Database and the CJK Database
The broad terms of the relationship between the ABN and CJK databases have been established. The National Library, in consultation with the National CJK Project Steering Committee and the Network Committee, has accepted financial responsibility for the NBD aspects of the CJK system on the understanding that the CJK database is conceptually the CJK arm of the National Bibliographic Database. The National Document and Information Service (NDIS), which is currently under development, will not provide the full range of support for CJK scripts in its initial implementation. This policy was adopted for technical and financial reasons, and the National CJK Project is envisaged as a low cost 'stop gap' solution to the support of CJK scripts.
The existence of separate databases using different systems has implications for bibliographic access to CJK materials at the national level, and the Roundtable meeting provides an opportunity to clarify the relationship between the two databases and to achieve consensus on the practical management of this relationship. Some factual background information about the coverage of the two databases will set the scene.
Coverage of the ABN and CJK Databases
At the time of implementation of the CJK system, all CJK data (bibliographic records and holdings statements) on ABN will be copied to the CJK database. The CJK data will not be deleted from ABN, so on 'day one' the CJK content of the two databases will be identical, except that the CJK records from Library of Congress, OCLC and RLIN will be re-loaded in their original form with the CJK characters. (Under current arrangements, the CJK characters have been stripped from the records before loading to ABN).
From the implementation of the CJK system, the two databases will start to diverge in the following ways:
- The National Library plans to cease loading the Library of Congress CJK tapes to ABN, and will load them to the CJK database instead. This policy will be implemented as soon as practicable after the implementation of the CJK system. A limited period of double updating would be feasible, at least until most Australian libraries have had a reasonable opportunity to start accessing the CJK system for their cataloguing of CJK data.
- The members of the consortium for the National CJK Project will start adding cataloguing and holdings data to the CJK database instead of ABN. The major growth in CJK cataloguing and holdings data will be on the CJK system rather than ABN from that point onwards, making the CJK database the preferred first point of call for CJK cataloguing and holdings data.
- The general picture is that after the implementation of the CJK system, the two databases will diverge steadily, with some unique CJK material on each database, some overlap between the two databases, but with the CJK database being the richest source of CJK data and the logical first point of call for all DNC operations relating to CJK materials.
Updating of the ABN and CJK Databases
The National Library has adopted a policy of 'no mutual updating' of the two databases. The National CJK Project Steering Committee has considered the possible implications of not attempting to cross-update between the ABN and CJK databases after the initial ABN load to the CJK system. It was agreed that after a year to 18 months a follow-up transfer of CJK data from ABN to CJK would be considered, but the Steering Committee endorsed a strategy of 'no regular cross-updating between CJK and ABN databases'. It recommended that the National Library should follow a strategy of minimal overlap between the two databases, and after the initial load the ABN and CJK databases would not be synchronised (23 September 1993).
Libraries will be free to add records and holdings to either system, but there will be no attempt to automatically capture this data and copy it to the other database. This is consistent with the National CJK Project as a 'low cost solution' to the problem of accessing CJK materials using CJK scripts. In this environment, it would be useful to encourage libraries to direct all their CJK data contributions to the CJK database.
Recommendation 4:
That the Roundtable encourage all universities with significant collections of Chinese, Japanese and Korean materials to contribute their bibliographic and holdings data to the National CJK Database on an ongoing basis.
It would also be useful to encourage libraries which have not contributed their retrospective data to ABN, to contribute this data to the CJK database.
Recommendation 5:
That the Roundtable calls on libraries with existing machine readable files of bibliographic records for Chinese, Japanese and Korean materials, which have not yet been contributed to the NBD on ABN, to contribute these records directly to the CJK database in the format specified by the National CJK Technical Committee.
This raises the question of the CJK data on the ABN database, particularly in relation to holdings statements. Much of the CJK data on the ABN database will gradually age. Bibliographic data will not be maintained and enhanced in the shared cataloguing environment. For example, libraries may add subject access or CJK script to the records on the CJK database, and they may modify their holdings statements to indicate that they have cancelled a serial subscription or discarded particular a work. These changes will not be made to the ABN database unless individual libraries choose to do so. We can anticipate a gradual decline in the usefulness of the ABN data for copy cataloguing and interlibrary document supply.
A medium term objective for the CJK database would be to become a practical 'one stop shop' for national bibliographic access to CJK materials in Australian libraries, as well as a useful source of copy cataloguing data. Once Australian libraries are routinely using the CJK system, and the CJK database is effectively fulfilling its NBD role, the outdated and duplicate CJK data could be removed from ABN. However, this would certainly not be undertaken before Australian libraries have made a successful transition to the CJK system, and from a technical viewpoint it is unlikely to be a straightforward task before ABN's migration to the NDIS. While we are in transition between the two systems, it will be important to overcome any confusion by education of system users on the best strategies for locating CJK data.
Improving Bibliographic Access to Asian Materials
The DNC Survey question 12 invited libraries to suggest the most important needs for improving bibliographic access to Asian materials in Australian libraries. Responses to this question could be broadly grouped as follows:
- More data available for copy cataloguing
- Retrospective conversion of printed catalogue records
- Retrospective upgrading of existing records to include vernacular scripts
- Standards issues to be resolved - especially romanisation of Chinese (Pinyin), but also character sets (Unicode)
- System improvements - extension of support to non CJK scripts, improved online display of Vietnamese
- Staff language skills
- Bibliographic guides to materials and sources for Asian studies and languages in Western Australia [See Attachment 1 to this paper, a document submitted by the University of Western Australia to illustrate their answer to Question 12 in the DNC Survey]
Authority Control
It is interesting to note that the question of authority control, which was the subject of a resolution at the 1993 Roundtable on Libraries and Asia, was not mentioned in the DNC Survey responses. The obvious inference is that there are more pressing issues concerned with the creation and sharing of basic bibliographic and holdings data. Authority control is a part of the cataloguing operation, and high quality cataloguing will always include resolution of synonyms and conflicting forms of authors' names. All MARC records received from the Library of Congress will have been subjected to authority control, providing a head start for the NBD, and it is up to Australian libraries to extend this authority control activity in their own cataloguing. However, this paper does not propose any resolutions with regard to authority control, in order to focus more clearly on the most pressing issue of creating basic catalogue records and reporting them to the NBD. It is suggested that the detailed implementation of cataloguing standards is best tackled by the ABN Standards Committee and the CJK Technical Committee.
Pinyin Romanisation of Chinese
Among the suggestions for improving bibliographic access to Asian materials, is the standard for Pinyin romanisation of Chinese. This issue has aroused more intense and broadly based interest in the Australian library community than any other standards issue in recent years, which probably reflects the intractable nature of the problem. The ABN Standards Committee commissioned a study of this subject by Susan MacDougall in 1990, and consulted university libraries (which have significant collections of Chinese materials) on a proposed standard for Pinyin romanisation. The Standards Committee consequently adopted a standard of single syllable word division as the lowest common denominator among the existing practices at the time. (Copies of report of the MacDougall study are still available from Julia Trainor at the National Library).
The Pinyin standard has proven neutral and accessible for cataloguers, but not popular. In particular it is unpopular among public libraries, whose users are unaccustomed to using romanisation at all. The Pinyin standard has been a less pressing issue for university libraries, where academic staff are more accustomed to coping with romanisation. The advent of the National CJK System has also reduced the urgency for standardisation of Pinyin romanisation among university libraries.
The issue of Pinyin romanisation was raised at the Annual Users Meeting of ABN and Ozline users in Hobart, July 1994. Technilib proposed the adoption of a standard for Pinyin which uses polysyllabic word division, based on a dictionary. The meeting referred the issue to the ABN Standards Committee and the CJK Technical Committee for resolution. The ABN Standards Committee established a Pinyin Committee comprising two members of Standards Committee (Robert Ho and Julia Trainor), two members of the CJK Technical Committee (Susan Prentice and Andrew Gosling) and two members representing libraries which use Pinyin (Robert Apedaile and Malika Xayasy). The Committee was set up in November 1994, when a large bundle of background reading was distributed to each member, along with the proposal by Technilib. The Pinyin Committee meets by email, but to date there has been no further discussion of the Technilib proposal, and no alternative proposal has been submitted. The Pinyin Committee can be expected to make more progress in preparation for the next Standards Committee meeting on 29-30 March, but any individuals interested in contributing to the development of a standard for Pinyin based on polysyllabic word division are invited to join the Pinyin Committee's discussion. If progress has been slow, it is because we have moved beyond the mere statement of the problem, and are focussing on the development of an acceptable solution.
Bibliographic Guides
The suggestion by the University of Western Australia for the preparation of bibliographic guides to materials and resources can only be applauded. There will always be a role for such guides and directories, particularly when the compiler has subject knowledge. Of course, there will be some overlap between the coverage of such directories and the content of the National Bibliographic Database. It is hoped that the compilation of guides and directories does not become a substitute for reporting holdings to the NBD, and it is also hoped that the NBD is able to provide valuable data for inclusion in such works. The coverage of the NBD and the relationship between the NBD and specific bibliographic projects is already the subject of recommendations earlier in this paper.
Conclusion
Although the future for access to information is full of promises, there are two fundamental elements to this access: data and systems. While computer systems will come and go, the data we create will live forever, providing access to materials in formats ranging from microfiche to electronic full text documents. The process of creating bibliographic data is very expensive, but the intellectual value of bibliographic data is difficult to assess in simply monetary terms. However, the value of bibliographic data for the management of the Distributed National Collection will only be realised if the data is nationally accessible, and the best way libraries can do this is to support the National Bibliographic Database on both the ABN and CJK systems.
Summary of Recommendations
- Australian libraries with significant Asian collections should continue to contribute holdings data for current acquisitions promptly to the National Bibliographic Database, including the National CJK System when it is implemented. Where online cataloguing is not feasible, libraries should contribute current update files twice a year until the implementation of upline loading in the NDIS.
- That the Roundtable attempt to achieve a consensus on what are the most significant files of cataloguing data which should be loaded to the NBD, with a view to seeking ARC funding for significant costs associated with data preparation, data upgrading (e.g. with vernacular script) or format conversion.
- That any proposals for ARC funding of projects for the cataloguing or retrospective conversion of data for Asian materials should always include provision for contribution of data to the NBD as an integral part of the project.
- That the Roundtable encourage all universities with significant collections of Chinese, Japanese and Korean materials to contribute their bibliographic and holdings data to the National CJK Database on an ongoing basis.
- That the Roundtable calls on libraries with existing machine readable files of bibliographic records for Chinese, Japanese and Korean materials, which have not yet been contributed to the NBD on ABN, to contribute these records directly to the CJK database in the format specified by the National CJK Technical Committee.