CLIR believes that preservation remains a fundamental mission for libraries and that preservation has a long-term payoff, even though it has grown more difficult to fund and manage. Libraries must balance local needs with the national agenda, weigh the value of content on each new medium against its typically short life span and funding constraints, and monitor the changing research priorities of scholarship. Preservation of nondigital materials is complicated by the difficulty in setting priorities for selection and by the unrealistic, if understandable, view of many that all resources are of equal value and must be preserved. Preservation of digital information is challenging not only for technological reasons but also for legal reasons. Traditionally, libraries have had a mandate to preserve what they own. Few if any libraries are willing to take responsibility for preserving licensed electronic resources that they do not own, even though they recognize that publishers are not archivers.

Increasingly, the preservation of cultural and scholarly resources is becoming the responsibility of all who have a stake in them—creators, publishers, and users—as well as of the traditional custodians in libraries and archives. Preservation awareness must reach beyond the walls of the library and the grounds of the campus.

Preserving Multimedia Resources

In July, CLIR published a report on the state of resources documenting the art of dance. Dance documentation exists in all formats, from print to video to digital, and the challenges of preserving and enabling access to these materials mirror those of all the performing arts. The report, entitled Securing Our Dance Heritage: Issues in the Documentation and Preservation of Dance and produced by the Dance Heritage Coalition, identifies the numerous formats in which information about dance is recorded and the range of individuals and institutions that have some responsibility for providing long-term access to this information. Not surprisingly, many of the individuals charged with this responsibility are not professional librarians or archivists. Therefore, any strategy to preserve these materials must involve many small groups and communities in a coordinated national effort to raise awareness of what is at risk. CLIR lent further support to the Dance Heritage Coalition to convene of series of meetings that developed a national strategy for the preservation of dance documentation.

Recorded Sound in Peril

Like dance documentation, recorded folklore collections are an endangered resource. Such collections consist mostly of field recordings, some of which date from the Depression era or earlier and exist on a variety of fragile and obsolete media. Besides being physically vulnerable, these materials are imperiled by a lack of agreement on standards for description and access. Some have been accessioned into libraries and archives, but many exist in private collections or in small cultural agencies that lack adequate storage facilities. To address the needs of this corpus of potentially national significance, CLIR is working with the American Folklore Society and the American Folklife Center in the Library of Congress to develop a strategy for preservation and access. The first step in this project is a baseline survey of ethnographic materials in institutions and private collections to determine the scope of the documentation and the preservation needs. The work will culminate in December 2000 in a conference that will bring together experts—archivists, librarians, recorded-sound technicians, preservation and media specialists, intellectual-property lawyers, and recording company executives—to explore all aspects of the crisis and reach a consensus on collaborative action.

Digital Preservation

Among the most widely used techniques for managing long-term access to digital files is migration, i.e., the transfer of digitally encoded information from one hardware-software configuration to another. Migration is intended to keep selected digital information accessible by ensuring that hardware or software obsolescence does not strand files in unreadable formats. It is, in essence, a translation program, and, as is the case with all such programs, some measure of information is lost in the movement from one encoding scheme to another. CLIR commissioned the Cornell University Library to do a study of what could happen to digital files over time as a consequence of multiple migrations. The study was conducted over 18 months, and the results were reported in June 2000 in a publication entitled Risk Management of Digital Information: A File Format Investigation. This report provides tools for assessing risks to some standard formats and enabling managers to make informed decisions when implementing migration strategies.