Among the key assumptions informing this meeting was an awareness that:

• facilitating the preservation of and access to humanistic information resources across a range of systems and media was of national importance;

• understanding of the human record would be significantly enhanced by developing cross-disciplinary and cross-media compatibility of systems to enable researchers and the general public to search a wide selection of humanities material in disparate locations, and to do so easily and creatively;

• identifying mutually satisfying mechanisms that would enable humanists to work more effectively, productively, and creatively with industry and academic technologists and generate new applications would be of enormous value to humanists and challenge computer scientists into new ways of thinking;

• understanding the intrinsic qualities of the material to be studied to enable appropriate electronic conversion and to foster generation of new material that may be possible only via electronic media would greatly enrich our cultural legacy.

Until very recently, computing and communications science was not seen as a partner to the arts and humanities (hereafter "computer science" and "humanities" for brevity). This perception was due in part to the exceptionally varied nature of humanistic study, an often loose coalition of disciplines that can include philosophy and some schools of sociology, anthropology and music, fine arts and the history of technology, and even some aspects of environmental studies (see Appendix B). Cultural institutions such as museums and libraries are also included in the humanities purview. This realm of intellectual pursuit was deemed antithetical to precise, mathematical computing applications.

Yet all of these varied areas of study focus on the preservation, transmission, and interpretation of the human record, and all of these humanistic fields are emerging as implementers of computer-based technologies in scholarship and teaching applications. Despite an initial lower penetration of information technology, funding constraints, and inherently conservative methods, the humanities, like other groups in society, are making greater use of computing and communications for efficiency, long-distance interaction, and new forms of activity made possible by technology.

Early applications of computer science in the humanities involved statistical analysis of texts (digital concordances and determination of authorship), and computing of statistics relating to historical studies (census results, shifting wartime populations). More recent technological advances have facilitated progress in symbolic and textual storage and analysis, and newer technology enhances the storage and use of audio, image, and video material. Projects such as American Memory, ARTFL (American and French Research on the Treasury of the French Language), Perseus, the Making of America Project, and the Museum Educational and Site Licensing Project testify to increasingly sophisticated applications in the humanities encompassing text, image, sound, and multimedia combinations.¹ They also demonstrate how technology can strengthen links between the creation of new knowledge—research—and its dissemination—teaching and public presentations.

Computer science, in turn, has often focused on scientific and business applications, involving processing of numerical data and simply formatted text. Trends in computing and communications illuminate new opportunities associated with electronic "content": visualization of all kinds of information, digitization of music and other audio material, electronic publishing with increasingly elaborate formats, and more complex paradigms for finding, selecting, displaying, and making use of greater quantities and varieties of information.

These applications align the interests of computer scientists and humanists, and they spurred the convening of this group of practitioners. Recent progress in information technologies in the humanities and computer science also encourages more programmatic collaboration. Going into the March (1997) meeting, it was understood that effective partnerships among computing and humanities professionals call for an understanding of the opportunities in both sets of disciplines: what are shared intellectual interests; which problems in one domain provide challenges for the other; where can collaborations prove particularly fruitful. That understanding is uneven at best today. It also points to a range of important process questions: how do computing and humanities professionals do their work; what are their respective expectations for project size, cost, structure, and outcome; what barriers or constraints tend to shape projects; what incentives and disincentives can affect collaborations?

The March meeting took these questions as its point of departure. (See Appendix C for agenda and participants.) The ad hoc group framed these issues by characterizing the needs of humanities scholars, delineating the systems now available and the economic and institutional context in which they work, and exploring the processes by which future progress can take place.

A digest of those discussions follows; the roundtable members are quoted as often as possible to convey the tenor and interactive nature of the meeting. Among the most important themes of the day's discussions is collaboration, which depends on language, methodology, and institutional context. Another is structure—the technical means (e.g., standards) for humanists to integrate and use computing systems in their research and applications. The third broad theme is resources, including money, time, and talent. These themes interact to the extent that shared intellectual excitement can bring humanists and computer scientists together and inspire new tools for humanities activities—but resource issues may determine ultimate outcomes.