A review of existing practice |
...an eLib supporting study |
To date the majority of searchable electronic catalogues have relied on the heavy structuring of data, and have commonly also used forms of relational databases. In terms of software, market leaders for some years were CAIRS, MODES, STATUS CLIO and ACCESS. STATUS CLIO is the software of choice for the Scottish Record Office's SCAN project. Dominant among new entrants to the market place is CALM 2000 Plus which is being developed in conjunction with a number of County Record Offices and is deliberately being formulated to reflect recent developments with regard to ISAD(G) and Z39.50. All of this software has been used to support cataloguing at all levels of descriptions. SGML (Standard Generalised Markup Language), although of considerable interest to archivists as a tool to support the calendaring of individual documents or other detailed editorial work, was not generally regarded as useful until the development of Encoded Archival Description (EAD) and HTML.
Because of its ease of implementation, rather than its sophistication, a significant number of Follett fund holders are using HTML to deliver information onto the Web, including catalogues commonly prepared as Word documents. For collection level descriptions, in particular, this is being seen as an acceptable way of presenting data pending the development of an easy solution of greater sophistication. Searching is, however, recognised as crude, especially where there are runs of date brackets in biographical and administrative history elements, and scope and content notes inevitably lend themselves to false drops for combined date and subject searches.
Part of the brief of the JISC NFF Archive Subcommittee was to research and develop strategies likely to facilitate networked access to archival catalogues. In scoping the task the Subcommittee took as its working premise that user access would be facilitated by provision of a single or limited number of gateways to a national network. This would provide cross sector access to archival catalogues held in the whole range of university, national, local and private repositories. Further, that it should be possible to search all levels of descriptions down to item or piece used in catalogues, and that any resultant structure should be capable of incorporating catalogues prepared using a variety of software. From the outset ISAD(G) was recognised as the dominant descriptive standard and Fretwell-Downing Informatics, a UK library and information systems supplier, was commissioned to develop a draft Z39.50 profile based on ISAD(G). The resultant draft profile reflected work already undertaken on the widely respected CIMI Z39.50 profile for museums.
The National Networking Demonstrator Project was subsequently mounted to provide a working test of multi-level cross searching of archival catalogues using Z39.50 including the development of a uniform front end to permit seamless public access to catalogues prepared using a variety of software. The project has been the subject of wide preparatory open discussion and follow up demonstrations have been attended by archivists from all sectors. The Demonstrator itself includes catalogues prepared in universities, the Public Record Office and local government, variously using CAIRS, MODES, CALM 2000, word processing, and EAD and HTML. Participation in the project required the provision of fonds level description, and a willingness to mark up catalogues in ISAD(G). Participants were offered a range of options in submitting data: those with their own servers could elect to link them into the Project, making their data available directly through system adaptation, or indirectly via export to a reference target.
The robustness of ISAD(G) as a descriptive standard has undoubtedly been confirmed by the project. The utility of Z39.50 as a means of supporting multi-level cross sector access for structured archival catalogues is also now widely recognised. The first phase of implementation of a cross sector national archival network planned to begin in autumn 1998 will build upon lessons learnt in the course of the Demonstrator and is expected to include the use of Z39.50. The final reports of the contractors Fretwell-Downing Informatics, and the Archival Consultant on particular lessons learnt, and appropriate next steps, including moves towards the formal registration of the profile, are awaited with interest. What is already clear, however, is that the Demonstrator has thrown back on the archival community a number of issues about how the various levels of description available are used, and also concerning access points. Some of these are already being reflected in the recommendations of the Society of Archivists concerning mandatory and new elements for ISAD(G). Others, like the handling of small collections, will need to be the subject of pragmatic decision at the time of a full-scale implementation of a national network.
Another approach to searching a range of archival catalogues is the American Heritage Project that, as a demonstrator, will bring together catalogues relating to nineteenth and early twentieth century American heritage with a view to studying issues relating to content and encoding. Based at the University of California at Berkeley, the project includes material from three other American universities. Collection level data is provided in the US MARC format and detailed catalogues in EAD. Presented as a union database, rather than the gateway approach of the Demonstrator, the intention is to support navigation from the union catalogue to individual detailed catalogues. Results are expected in 1998.
Patricia Methven, King's College London