THE UNIVERSITY OF CALIFORNIA DIGITAL LIBRARY:
A FRAMEWORK FOR PLANNING AND STRATEGIC INITIATIVES

APPENDIX D
Content: Collection, Access, and Management

Copies and Intellectual Property

In the digital environment, access to content will often be secured by licensing the right to use and archive the intellectual property contained in a work rather than, as in the paper environment, purchasing a copy of a work. Content should be mounted in a variety of locations, including local servers, central campus servers, central UC servers, and remote servers. Content owned by UC should be mounted on University servers with clearly established conditions and terms on which it may be used within and beyond the University. Management of intellectual property will complement the acquisition of physical copies as the primary mechanism for operationalizing a libraryís mission. For the UCDL, this transition should mean that ownership will focus on unique objects ranging from manuscripts and art objects housed in UC special collections and museums to new scientific knowledge created in UC laboratories. Ownership should carry all the obligations now embedded in ownership of paper collections: preservation across technological platforms, organization for access, cataloging and provision of finding aids, support for users. Ownership will also require new kinds of agreements between the UCDL and faculty creators of intellectual property. For content representing intellectual property held elsewhere, the preferred mode of access will likely be licensing in perpetuity, including either a guarantee that the owner will archive the licensed material or the right to archive it locally. Hybrid arrangements will mean that collection quality is no longer meaningfully measured by traditional research library indicators: the UCDL should be measured on the satisfaction of faculty and student needs rather than the "size" of its "collection."

As in the paper environment, organizing access to content should be a significant part of collection management, with the probability that the organizing structure may eventually drive "collecting" priorities by creating categories of information in which more or less completeness is sought. It will be necessary to develop pathways to content created and held elsewhere that are consistent with the organization of UC material but that do not require significant UC intervention. By creating an orderly information system and guaranteeing archival and service support, the UCDL will construct a collection which is clearly demarcated from the undifferentiated content of the Internet. The metadata that describes content to which the UCDL provides access should become a prominent element in its collection for which skilled staff or reliable vendors should be employed. One of the essential characteristics of the DLís core collection should be its provision of finding aids that are accurate and contain enough information to assist users in making reasonable choices about what items to pursue. Without such indicators, the time cost of seeking information in an expanding universe of electronic materials will be excessive, and researchers will be faced with the choice of restricting themselves to content with which they are familiar or engaging armies of graduate students to do their searches for them.

Developing improved search tools that enable individual users to discover and negotiate their access to externally held content should become an important aspect of "collection development," as should licensing access to remote content in frequent demand at UC. Determination of when reasonable access requires a UC license or the provision of multiple copies on more than one server should be driven by the volume and nature of demand and the advantages to be gained by incorporating specific content into the collection that UC manages. This includes assurance that the material will be archived and that access paths and storage media will be maintained and refreshed. In principle, access licenses should provide for users from every campus, with technical issues related to bandwidth and cost issues related to the level of use governing the choice between one and many servers. It should be affirmed that the UCDL will collect primarily for the research, instruction, and public service needs of the University and that, although public access will be facilitated to the extent economically feasible, it may be the responsibility of others to create organizing viewers and links to those parts of the UCDL that are most valuable to other users, such as K-12 educators, industry researchers, health practitioners, and public policy makers.

Digitization

In establishing a balance between paper and digital "collecting," efficient access and cost should be determining factors. Costs must be calculated to include acquisition, cataloging, storage, circulation, and user support services. Despite a growing preference for acquiring new works in digital form, paper collecting should continue until on-demand printing is convenient, cheap, and reliable since paper containers of information (books and journals) are more portable than digital ones. Provision should be made for meeting the challenge of creating a seamless interface for digital and paper content; these formats require different maintenance and retrieval functions and present different cataloging requirements. Currently, it is also important to consider the distinction between digitized and digital material, especially textual, although this distinction is likely to become less important in the future. The former are works originally held in paper or other tangible format and translated to electronic media but with varying levels of functionality, e.g. bitmapped images of journal pages placed on a server that can be viewed and downloaded but are not fully searchable or manipulable vs. searchable texts. Digitization may be carried out for ease of delivery or preservation as well as improved functionality. Digital material, on the other hand, originates in digital formats and makes possible the inclusion of many kinds of data besides text and images; it may be executable as well as manipulable and searchable. On functional grounds, higher priority should be placed on incorporating digital content into the UCDL than on digitizing paper content in limited-function formats--despite the appeal of using a single format for all collections. Why this is so may be understood by considering that the reader of a digital article may be able to retrieve the raw data underlying it, rerun the author's analysis, and even conduct new analyses; a digital article may include or be linked to software as well as data. Digitization should be employed when it offers an efficient form of access, particularly to unique material, and when it is the most appropriate way to deliver content to academic programs. The Mellon Foundation's experimental JSTOR project to digitize and license the entire corpus of ten leading journals, may provide some insights into the feasibility and value of large-scale digitization for existing paper material.

Archiving

It is important to state up front that no one yet understands how to archive digital materials. It is critical for UC to participate fully in the national debates currently underway in this area. Nonetheless, certain principles can be articulated which appear reasonable at this time.

Priority in archiving should be given to materials that are unique to UC, and standards should be equivalent to those already prevalent in managing special collections. Additional unique materials should enter the UCDL through cooperation with other institutions: some of these will be small and lack the capacity to mount their own archiving operations, while others may be large but will contain special collections that complement those at UC. Archiving decisions for non-unique materials should be made with reference to archiving activities elsewhere in such a way that academic libraries as a collective body assure the survival of significant works created in digital formats, with access assured to users at any location. The UCDL should also retain the right to archive any material it has purchased. The ability to print single copies on acid-free paper, where digital materials are printable, should be part of UC contracts and licenses. The UCDL system should be designed for distributed storage of content at sites both within and outside UC. Digital content should be stored in duplicate at multiple sites to secure against system failure or destruction of data in single-site disasters like earthquakes, fires, and floods.

Partnerships

Developing a collection by licensing access to information will necessarily involve the UCDL in an expanding network of partnerships. Partnerships should link UC to other institutions that are developing their own digital libraries, to national university and professional associations, to international creators and holders of information, to government agencies, and to local entities. All should have content to share and needs for the content that UC controls as well as contributions to make in developing standards and formats. In particular, the UCDL should rely on the national library community, which will continue to nurture systems, standards, and techniques as well as looking to UC for contributions. Close contact with alumni and other corporate affiliations are also likely lead to cooperative agreements to share the information resources held by many corporations in California.

Dissemination

Providing digital access to unique materials controlled by UC will sometimes blur the distinction between library and publisher functions and should offer opportunities for the UCDL to play a proactive role in making information available to University and public users. On a small scale, libraries have been performing some of these functions as they manage course reserves that combine traditional library material with original works created by course instructors or students. For at least some sets of data, the UCDL should facilitate peer review and information organization functions similar to those now associated with journals.

Virtually all content created at UC should be considered as potentially a part of the UCDL, regardless of the organizational nature of the unit in which it is created and held. With this possibility, it will be necessary to develop demarcations between content for which the UCDL is responsible and that to which it merely provides links. Faculty should necessarily play a continuous and strong role in developing criteria of quality and significance, with consultation taking place at all levels from research and instructional units to campus-wide and University-wide committees empowered to set priorities between competing demands for limited resources.

Last updated: 31 January 1997