Building collections for all time: the issue of significance.

With only a small percentage of libraries building collections for permanent retention the issue of what these collections should, or should not, comprise has been debated only intermittently in the professional literature. With the hugely increased focus on delivery of information to clients in recent years, this is not surprising. With the move from physical ownership of resources to remote access priorities have been elsewhere, with the result that debate over selection of material of enduring value for long-term preservation and retention has received limited attention. However, the issue has recently come to the fore with the explosion in the nature and amount of digital material being acquired by libraries, raising concerns over its long-term preservation and continued accessibility.

In addition, projects such as the UNESCO Memory of the World Programme, started in 1992, recognise that the documentary heritage, reflecting the 'memory of the world', is a fragile thing, requiring a sustained campaign to ensure the longer-term survival of valuable archive holdings and library collections from around the world. (1) This program is very much concerned with identifying significant materials which countries nominate for inclusion in the Memory of the World Register, in the hope that this recognition of their significance as a cultural artefact of world importance will help ensure their ongoing preservation.

With a growing concern focused on digital preservation, and through major international projects such as the Memory of the World, the question of acquiring, selecting and preserving our documentary heritage, in whatever form, has reawakened the debate into how libraries, as one of the institutions charged with the responsibility for preserving the cultural memory of their society, select material for permanent retention. Traditionally, this has meant identifying physical items (books, papers, maps etc) and conserving the original item in some way in order to ensure the object's long-term survival. At the same time, access issues have been addressed via digitisation, microfilming or other forms of copying. This relatively straightforward preservation process becomes more complex for digital objects, given the virtual impossibility of preserving the 'original' item for the long term, and the need to migrate the data or otherwise copy the content due to the technological imperatives of changing software and hardware. Issues of authenticity, the transience of the original item, context, formats, storage media, standards and so on all impact upon the preservation equation, as well as the sheer volume of material being produced. As Harvey notes, (2) with digital materials, as well as their quantity there is likely to be a wider range of quality issues, with many digital objects free from any quality control process traditionally imposed on print materials by publishers. With concerns like these, it is not surprising that the question of priorities for digital preservation, and thus the need to readily assess levels of significance in material, has become a major contemporary issue for librarians and others.

Thus Lehmann, (3) in a wide-ranging discussion of cultural tradition and the new digital objects, is concerned that 'long term availability is a major criterion for the creation of cultural memory' but that the features of the emergent digital media mitigates against this. Eastwood, (4) in a discussion relating to archival records, makes the point that the long-term preservation of digital objects will be a challenging task, even for relatively simple items, 'placing a premium on the capability of identifying those materials of enduring value', and thus worthy of the effort required for long-term preservation. He goes on to note that for digital objects, an additional complication arises in the determination of 'continuing value' which must include an assessment of the authenticity of the item and thus knowledge of its creation and chain of custody since that time. (5) For unpublished records in particular, the creation of audit trails or similar processes will be essential in providing information to assist in the assessment or appraisal stage of determining a digital item's value. Eastwood concludes that for any of the memory institutions, building and preserving collections of electronic records will require effective selection or appraisal processes to 'a greater extent than does preservation of traditional materials.' (6)

This need for increased attention to the identification of digital objects for long-term preservation is supported by Harvey who argues that the traditional criteria and paradigms used by libraries to identify material for retention may need to be revisited (7) and that guidance may be found in appraisal theory and practice as developed by the record-keeping discipline. (8) In particular, the decision over the significance of an item, which for materials in traditional forms may be undertaken at any time in their existence, needs to be made much earlier for digital records. This early decision making is essential to ensure that they are acquired and stored in a manner which protects their integrity and limits, as far as possible, concerns over the use of proprietary software and other file- or hardware-specific devices that may make longer term access difficult or impossible. This need to make decisions on…