1999
President's Report
The Foundation’s annual report serves, first of all, a stewardship function. All of the individual grants made in 1999 are listed at the end of the report, and a summary tabulation is provided by category. The Trustees appropriated over $161 million in 1999, and nearly two-thirds of this amount (64 percent) was in support of a wide variety of projects in Higher Education and Scholarship. The remaining grants were concentrated in five other categories: Conservation and the Environment (11 percent); the Performing Arts (10 percent); Population Studies (8 percent); Museums and Art Conservation (4 percent); and Public Affairs, including the Refugee Program (4 percent). The Foundation’s longstanding commitment to the humanities is reflected in this pattern of grantmaking, as is the belief that a reasonable degree of continuity in support is important if lasting results are to be achieved.
We have not followed a "stop-and-start" approach. Each year is also witness to new emphases and new initiatives, and to the phasing out of activities that either have served their purposes or no longer have as strong a claim on the Foundation’s resources as they once did. A second function, then, of the annual report is to inform potential grantees and others of the ways in which various programs are evolving and of new ideas being explored by the staff and Trustees. It is not practical, however, to cover every program on an annual basis; in order to provide some focus, we concentrate each year on a small number of topics that seem especially timely.
This year, the president’s report has a single primary focus: how the Foundation is addressing the impact of information technology (and especially digitization) on scholarship, scholarly communication, and libraries. Because this is such an important area, and is changing so rapidly, we thought there would be merit in providing enough detail to allow grantees and other interested parties to understand the Foundation’s thinking as it has evolved. In addition, this year’s report contains a special essay on the Foundation’s new initiative in Refugee Studies and Forced Migration written by Carolyn Makinson, the staff member who is also responsible for the Foundation’s population program. (Note 1)
Focusing on these particular areas provides only a partial picture, however, of the range of the Foundation’s activities. The extent of continuing support for established programs—such as the Mellon Fellowships in the Humanities, the Sawyer Seminars, and a wide variety of post-doctoral fellowship programs—is evident in the list of grants at the end of this report. There are also important new developments in areas of higher education in which the Foundation has long been active, and next year’s annual report will focus on initiatives at both liberal arts colleges and universities, including the continuing evolution and deepening of the Foundation’s efforts to encourage promising minority students to pursue academic careers (most notably through the Mellon Minority Undergraduate Fellowship Program).
Staffing
This account of programmatic directions needs to be prefaced by a discussion of major staffing changes that have occurred at the Foundation, since programs and the people responsible for leading them cannot be considered apart from one another. This tenet—that basic judgments about program directions and staffing must go hand in hand—is basic to the Foundation’s style of grantmaking, which depends heavily on the presence of highly accomplished staff members who are able to work closely and effectively with key institutions in the fields of particular interest to the Foundation and its Trustees.
The Foundation has been fortunate in having attracted in earlier years individuals who continue to contribute outstanding leadership in fields of central interest to the Foundation: Harriet Zuckerman and Pat McPherson (higher education); William Robertson (conservation and ecology); Carolyn Makinson (population, refugee studies and forced migration); Catherine Wichterman (performing arts); Angelica Rudenstine (museums and art conservation); and Tom Nygren (South Africa, technology).
In 1999, three other senior colleagues, who have contributed enormously to the work of the Foundation, moved to important new positions outside the Foundation. Stephanie Bell-Rose was elected president of the newly created Goldman Sachs Foundation; Richard Ekman was appointed Vice President for Programs at Atlantic Philanthropic Service Company, Inc.; and Jacqueline Looney returned to Duke University to serve as Associate Dean of the Graduate School.We also anticipate that two senior advisors will retire this June: Henry Drewry, who created the Foundation’s Mellon Minority Undergraduate Fellowship (MMUF) Program and currently works with the Historically Black Colleges and Universities(HBCUs) and Alvin Kernan, who directed the Foundation’s portable fellowship program in the humanities and has served with distinction as senior advisor in the humanities.
In my view, the Foundation has been extremely successful in identifying and attracting able new colleagues. Michele Warman, an accomplished lawyer, has come from Davis Polk & Wardwell to accept the position of General Counsel and Secretary, and Gretchen Wagner, who worked with Ms. Warman at Davis Polk, has come as Assistant General Counsel; Donald Waters, who most recently was Director of the Digital Library Federation, has joined the Foundation as Program Officer for Scholarly Communications; Lydia English has come from Brown University, where she was Associate Dean of the College, to succeed Ms. Looney as Director of the MMUF program and as a Program Officer in Higher Education; Danielle Carr, a mathematician who was on the faculty at Bryn Mawr, is now working in the liberal arts college area and will succeed Henry Drewry as the staff member with direct responsibility for the Foundation’s relationships with the HBCUs; Joseph Meisel, who recently completed his PhD in British history at Columbia University while working in the university’s budget office, has come to the Foundation to work principally with Harriet Zuckerman on issues facing research universities and centers of advanced study. Finally, Bernard Bailyn of Harvard and Paul Hunter of the University of Chicago have accepted invitations to advise the Foundation on future directions in the humanities. For as small an organization as this one, this is a lot of change!
We will of course miss those who have done such excellent work here; we wish them well, and we are proud that those who are not retiring have taken on such important new responsibilities. At the same time, we are delighted by the ideas and considerable energy that our new colleagues are already contributing and by the wide variety of experiences and skills that they have brought to the Foundation.
Information Technology, Scholarly Communication, and Libraries: Introduction and Organization of the Discussion
The relation between staffing and programmatic directions is nowhere clearer than in the area of information technology and its applications in higher education, the arts, and other fields that depend on access to databases and other scholarly resources. The Foundation has assembled individuals with a range of special skills and experiences that will permit it to play a useful role, we believe, in helping to conceptualize new directions at a time when technological advances are changing dramatically how scholarly materials are published, presented, stored, and used by scholars worldwide.
In a previous discussion of this general subject, last year’s annual report was unequivocal in emphasizing that the Foundation intends to maintain its traditional emphasis on the importance of substance and content, rather than on technology for its own sake.
At the same time, the staff and the Trustees believe that the Foundation can play a special role in mediating between rapidly developing technological possibilities and those applications most likely to benefit education and scholarly research. Work with grantees in 1999 has reinforced this conviction.
The discussion that follows is intended to highlight themes that connect a number of grants made during 1999 (and earlier). These include:
• The evolution and future prospects for JSTOR, an electronic database containing the backfiles of leading scholarly journals, that was the Foundation’s first major project in the applicationof information technology to scholarly communication;
• Related projects sponsored by the Foundation, includingthose devoted to the digitization of monographs and digital archiving;
• New opportunities that are emerging in the imaging of art and related scholarly materials that are otherwise inaccessible or widely dispersed, and the possible creation of a new "ARTSTOR" entity (a parallel to JSTOR) that would organize and distribute electronic archives of art images, manuscripts, and relevant scholarship;
• Issues of intellectual property rights that affect many of these projects, and especially the imaging of art;
• The status of college and university libraries in a changing environment shaped increasingly by electronic technologies.
At the December Board meeting, our colleague, Harriet Zuckerman, directed attention to the interconnections among what she called "a growing array of Foundation activities aimed at bringing content and coherence to the broad field of electronic scholarly communication, including the changing role of the library."
"JSTOR’s contribution," she noted, "is to provide textual content , to ensure its preservation, and to facilitate easy access to scholarly literature that might otherwise never be located or used." Grants in support of the proposed Dunhuang Archive and the digitization of MoMA’s design collection (discussed later in this report) will provide another kind of electronic content, in these instances in the form of images linked to text. Other grants (including the one to the Institute for Advanced Technology in the Humanities at the University of Virginia, also discussed later) aim to improve the means of delivering content by supporting the design of sophisticated databases and other kinds of software. She then referred to a substantial series of grants focused on libraries, which are intended to help students and faculty become proficient users of electronic content. Finally, grants have been made for the purpose of assessing the legal and economic implications of electronic publishing and archiving.
"Separately and together," Ms. Zuckerman concluded, "this array of grants should help faculty members, students, libraries, and scholarly publishers navigate—and shape—the still inchoate and rapidly changing world of electronic access to information."
JSTOR
When the Foundation made its first JSTOR-related grants in 1994, the intention was to explore the feasibility of creating and making available a new kind of electronic archive of journal literature.
The history of JSTOR—now a free-standing, not-for-profit organization with its own Board—is fascinating and full of lessons. (This history deserves to be recorded, and a colleague at the Foundation with experience at JSTOR hopes to begin work on this assignment in the summer of 2000.) Kevin Guthrie, the president of JSTOR, will be presenting his own account of its recent results and current plans, and my objective here is not to report for JSTOR but to note important points of intersection between JSTOR and the broader interests of the Foundation (Note 2)
JSTOR Usage and Content: Priorities and Funding Principles
College and university presidents and librarians have spoken with essentially one voice in endorsing what JSTOR has accomplished to date. Echoing Samuel Gompers’ famous response when he was asked what the American labor movement wanted("More!"), they have urged the Foundation to help JSTOR include more and more content. This is not surprising. By the end of 1999, JSTOR had more than met its Phase I goal, which was to make available to participating libraries the full backfiles of at least 100 core scholarly journals. In fact, the database contained 117 journals in 15 academic disciplines (over 750,000 articles and more than 4.6 million pages of content). In 1999, over 4 million searches were performed on the database and over 1.3 million articles were printed from it. Usage continues to grow at an astonishing rate (tripling in the most recent year), and one challenge for JSTOR is to build server capacity fast enough to maintain its high standard of performance. At the time of this writing, in mid-February of 2000, 651 participating libraries, including 121 outside the United States, had demonstrated their confidence in the utility of JSTOR by paying the fees necessary to gain access.
JSTOR has succeeded in faithfully replicating the "look and feel" of the original articles, it is easy to use, and it contains a high fraction of the most prestigious and widely cited journals in each discipline that it covers. From the beginning, the plan was for JSTOR to become a self-sustaining entity—beneficiaries of JSTOR’s services, initially college and university libraries, were expected to cover its costs on an on-going basis. This plan is working. JSTOR is nearing the point when the revenue it receives from annual access fees paid by participating libraries will cover the costs of maintaining the existing Phase I database. In other words, if JSTOR decided not to add new content or take on more projects, it would soon generate sufficient revenues to cover the costs of fulfilling its original set of promises to the scholarly community.
Operating in a maintenance mode is not, however, what the academic community is imploring JSTOR to do. Indeed, in response to the insistent demand for "more," JSTOR’s Board has decided to embark on an ambitious enhancement of the original list of titles and is now planning to double (more or less) the content of the arts and sciences core collection. In addition, it plans to add a series of stand-alone collections in fields such as general science, ecology and botany, business and finance, and language and literature.
A key issue for everyone interested in this project, and one which the Foundation’s Trustees have discussed at length, is how additions to the JSTOR database should be funded. In the main, JSTOR will continue to follow the basic principles that have guided it to date. Thus, when libraries agree to join JSTOR or to take new content, they will be asked to pay both a one-time "Archive Capital Fee" (formerly called the "Database Development Fee") and an "Annual Access Fee." (Note 3) The one-time fee is intended to cover part of the initial costs of digitization as well as to provide resources to ensure that data and software can be migrated to new platforms as technology evolves (thereby addressing, for this body of literature, the "electronic archiving issue"). The annual fee will cover the recurring costs associated with: (a) providing the server capacity and technical support needed to make sure that users have convenient and reliable access to the archive 24 hours a day and 365 days a year (this "library" never closes); (b) training for new users; (c)adding a substantial amount of new content each year as the "moving wall" that divides current issues from the backfiles moves forward (which means adding another year’s worth of articles published by the growing number of journals); (Note 4) and (d) maintaining JSTOR’s infrastructure.
Together, these charges should not be so high that they discourage libraries, especially those with limited resources, from participating in JSTOR. Herein lies the challenge: JSTOR must balance its core desire to fulfill its not-for-profit mission with the requirement that it pay its own way. If JSTOR were to follow a strict "market-driven" philosophy in deciding what content to add, it would add only that content for which there is a relatively large set of users willing and able to pay for access. The direct consequence would be a disincentive to add more humanities content or content in relatively esoteric fields such as Near Eastern and Slavic Studies. There simply are not enough departments, scholars, or students in some of the more specialized fields to justify, on market assessments alone, the investments required to develop the content. The conclusion we have reached is that outside funders, principally foundations, should be encouraged to invest in adding new content, while participating libraries and other users should be expected to cover the recurring costs of maintaining an ever expanding database as well as some (but not too large a share) of the initial capital costs. The Foundation’s Trustees have shown their willingness to continue to provide this kind of support. At its December meeting, the Foundation’s Board appropriated $1.75 million to cover part of the upfront costs of adding the full backfiles of another 35 journals in humanities fields such as history, classics, and archaeology. In addition, the Foundation made a second contribution in 1999 to building the JSTOR database through an appropriation of $2 million to the Modern Language Association (MLA). The MLA will assemble a list of key journals in modern languages and literatures and then work both with the publishers and JSTOR to make a separate "languages and literatures" collection available. The MLA and JSTOR will also assess the possibility of developing an indexing interface between the widely-used MLA bibliography and JSTOR.
It is encouraging that other foundations have also seen the importance of adding content to JSTOR based on the academic case for doing so. The Stavros S. Niarchos Foundation made a generous contribution to assist in covering the high costs of digitizing the classics and archaeology journals. Two other grantors, the Howard Hughes Medical Institute and the Josiah Macy Foundation, joined the Mellon Foundation in providing support for production of the general science collection (which will include all the backfiles of Science, the Proceedings of the National Academy of Sciences, and the journals published by the Royal Society in London, which go back to the 17th century).
JSTOR is now in the process of approaching other funders in the hope that they, too, will want to underwrite, at least in part, the upfront costs of adding content in fields such as area studies, the social sciences generally, and selected natural sciences. If the currently planned enhancement of the arts and sciences collection can be completed, nearly 10 million pages of text will be accessible at the desktops of scholars, teachers, and students. Another question is whether all of this content should reside at JSTOR. One could imagine other possibilities. For example, the MLA might have concluded that it should embark on its own program to digitize backfiles entirely apart from JSTOR. A more fragmented model of this kind is inevitable to some degree since JSTOR will never have the resources to digitize all of the scholarly literature that people will want captured in this way. However, there are two advantages to including as much core content as possible in an integrated database such as JSTOR. First, it makes sense that JSTOR leverage the knowledge and experience that it has gained in digitizing content; to the extent possible, the scholarly community should avoid having to re-invent the proverbial wheel. Second, users of scholarly literature naturally prefer to be able to access a single large collection of materials rather than a number of separate collections. Adding new journals and new fields to JSTOR has multiplicative value since one of JSTOR’s great strengths is that it allows users to search across disciplines (joining, for example, titles in literature to history and philosophy content). On purely intellectual grounds, then, there is a strong case for including in JSTOR as much related journal literature as possible. "Going it alone" is not only potentially costly and inefficient, it can also be undesirable from the perspective of faculty and students.
Linkages to Other Databases and to Current Issues
Equally important are ongoing efforts to find workable ways of linking the backfiles of journals in the JSTOR database to the electronic publication of current issues. If such linkages can be created in ways that protect publisher revenues (as we believe they can be), the benefits to scholars and to libraries will be tremendous. The scholar interested in a particular topic, such as conceptions of "culture" in German philosophy and literature or the urbanization of South Africa, is served best if he or she can search across the entire run of a journal, from the first issue to the most recent, without having to consult both the backfile and then (separately) current issues. For the library, such linkages could simplify user services and provide large cost savings. Libraries would still need to subscribe to current issues (albeit sometimes solely in electronic form)—otherwise publishers could not continue to afford to publish them! But if the librarians could be confident that their users would always have ready electronic access to the full run of a journal, they could then contemplate saving money not only on storage costs (as growing numbers of libraries with access to JSTOR are beginning to do now), but also on the costs of binding current issues and, conceivably, on the costs of handling paper copies in any form.
These and other aspects of the economics of journal publishing and library operations deserve continuing study. Fortunately, some "natural experiments" are underway that will allow us to analyze the effects of linking current issues to backfiles. There is no longer a gap between JSTOR’s coverage of the backfiles of Eighteenth Century Studies and the electronic editions of current issues provided through Project MUSE. Other publishers are shortening their "moving walls" (from five years to three years in the case of the American Economic Association) and several, including the University of Chicago Press, are interested in working with JSTOR to link their current issues directly to the JSTOR backfile. From the beginning, the Foundation’s objective in creating JSTOR was to provide a resource that would be both "better and cheaper." The long-run savings to libraries should be substantial, and the Foundation remains committed to promoting cost-effective approaches to the electronic archiving of journal literature and to demonstrating the savings that can be achieved.
Broadening Access to JSTOR
In addition to digitizing and adding content, foundations can play an important role in broadening access to scholarly resources such as JSTOR. In 1999, for example, both the Bush and Mellon Foundations made grants to the Minnesota Private College Research Foundation so that it can underwrite the start-up costs of providing access to JSTOR to a number of small and medium-sized colleges in Minnesota and the Dakotas. The Foundation’s earlier grants of a similar kind to the Appalachian College Association have had extremely beneficial effects on teaching and scholarship at the member institutions.
Other foundations with special "country interests" have helped to underwrite the costs of making JSTOR available in Ireland, Greece, and Israel. The Foundation continues to hope that efforts in South Africa to enhance connectivity and reduce its costs will make it possible to provide access to JSTOR in that country. It is only a matter of time before it will become feasible to make JSTOR available in other developing countries, many of which have virtually no access to scholarly literature.
Back in this country, the Sherman Fairchild Foundation is underwriting a pilot project to explore the value of making JSTOR (or parts of it) available to selected secondary schools; and JSTOR is also considering ways of making its archive available to public libraries. Scholarly associations, such as the American Sociological Association, are experimenting with programs designed to give individual members, including those who have no affiliation with a major library, access to their parts of the JSTOR database. More generally, the American Historical Association has suggested that consideration be given to ways in which the "independent" or unaffiliated scholar can be helped to gain access to JSTOR. In the long run, it is possible that JSTOR will need to focus more on relationships with end users (individual scholars) rather than rely almost exclusively, as it does now, on relationships with libraries and other intermediaries.
Monographs and Other Archival Projects
The Foundation has also sponsored a variety of projects intended to develop and test techniques for digitizing and distributing electronically materials of many kinds, and it is only limitations of space that prevent me from discussing a number of them here.
Special mention should be made of monographs. Following earlier grants in support of the Cornell-Michigan Making of America project, Columbia University’s Online Books project, and the Early American Fiction project at the University of Virginia, the Trustees made several grants in 1999 for projects intended to explore further an important but complex arena for digitization.
The American Council of Learned Societies (ACLS) received an appropriation of $3 million to carry out an ambitious collaborative project involving five learned societies and seven university presses. A unique feature of this project is that it intends, over six years, to produce 85 newly written electronic monographs, all in historical studies. Unlike the American Historical Association’s plan to produce electronic versions of doctoral dissertations (which is also being funded by the Foundation), the ACLS project, which is led by John D’Arms, aims to attract established authors who will want to write broad-gauged books that take advantage of the electronic medium. In the words of the historian Robert Darnton (a participant in the ACLS project), an electronic book, unlike a conventional one, can be "structured in layers that are arranged like a pyramid:" it can be read in a "linear"way for the development of its argument, can be plumbed for back-up data, bibliography, and the whole texts of source documents, and can be linked with correlative material located on other Web sites.
The ACLS and its collaborators are also committed to digitizing 500 backlist titles in related fields, along with major reviews of these titles, and to mounting the resulting collection of old and new monographs in a searchable database that will be made available to libraries on a subscription basis. While many of the practical details remain to be worked out, an important goal of the project is to test the economic viability of publishing electronic monographs. To that end, the project aims to find ways to standardize and streamline production processes; develop an infrastructure that will address the archiving issues that exist here as well as in the journal field; ensure that commercial vendors do not dominate the electronic dissemination of the results of scholarly research; promote the publishing of more monographs of high quality (but low market potential) by reducing costs and simplifying distribution; and, finally, encourage more presses to experiment with electronic publishing.
Grants were also made in 1999 to the University of Pennsylvania library, so that it can work with Oxford University Press (OUP) to create a large digital library of approximately 1,500 volumes in all fields of history, and to the University of Virginia to complete the digitization of a large group of early editions of American literature.
The decision to support this particular set of projects, focused as they are on scholarship in the fields of history and literature, reflects both the Foundation’s commitment to the humanities and its conviction that new developments in information technology should include such core subjects. We also recognize, however, that "the jury is out" on the question of how successful electronic publication of monographs will prove to be. In the case of journal literature, the enthusiastic responses of faculty and students to JSTOR demonstrate that very important substantive gains have been achieved already through the application of digital technology. In the case of monographs, no such persuasive evidence exists as yet, and it is clear from the design of several of the projects described above that not even an initial assessment of their "value added" will be possible for at least five years. (Note 5)
The Digitization of Works of Art: Possible Creation of an "ARTSTOR"
The most important early-stage investments made by a foundation do not always take the form of dollars appropriated. During 1999, staff and Trustees devoted a great deal of time and thought to ways in which the Foundation might encourage the creation of electronic archives of art and related scholarly materials. This area is in many ways a "natural" for the Mellon Foundation. Paul Mellon’s longstanding interest in works of art has been reflected in the Foundation’s own history of grantmaking in support of museums, the arts, and art conservation. There is an opportunity now, the Trustees believe, to extend this interest in the fine arts into a new era that is being shaped profoundly by advances in information technology. The Foundation’s interests in information technology and its long-term commitment to the arts and to scholarship in the humanities intersect in this rapidly developing area. The hope is that the Foundation’s experiences with the application of information technology in other fields, and especially with the development of JSTOR, can be translated into useful work in the fine arts.
In order to advance its broad interest in the digitization of works of art, and in order to learn more by "doing," the Foundation decided in 1999 to provide substantial support for two large and very different projects holding great scholarly promise—one focused on the architecture and design collection at the Museum of Modern Art (MoMA) in New York and the other on the cave art of Dunhuang in Western China. Consideration is also being given to the broader possibility of creating a JSTOR-like entity that might provide a common, coordinated system of distributing electronically high-quality "collections" of visual materials, including some that are otherwise inaccessible. (We refer to this hypothetical entity as "ARTSTOR," but this designation is only a temporary shorthand; if we proceed with this concept, as seems likely, it will probably be given a different name.)
There is already a great deal of activity in this broad field, but much of it consists of relatively small-scale, generally uncoordinated, efforts to digitize particular collections or parts of collections. On the basis of a recent survey of projects currently underway, we believe that a larger and broader effort focused on building a research-quality scholarly resource could have a significant leveraging effect—especially if launched relatively soon. Few of the current efforts to digitize art images are focused primarily on the needs of the scholar, and there appears to be insufficient emphasis on intellectual coherence. (As our colleague, Donald Waters, puts it: "The overwhelming first impression in this arena, as in many other parts of the Web, is its ‘wild west’ qualities...." Another observer refers to it as a "cottage industry.") But before saying more about the larger organizational questions, it may be helpful to provide a clearer sense of what is envisioned by describing the two pilot projects now underway.
The Design Collection of the Museum of Modern Art
At their March 1999 meeting, the Trustees appropriated $1.7 million to digitize a large part of MoMA’s architecture and design collection. It was established in the early 1930s and is one of the foremost collections of its kind: international in scope, representing substantial coverage of the best in modern design from the mid-19th century to the present. Although its existence is well known, the collection is virtually inaccessible. Only five percent of its works are published, and opportunities to study the material are scarce.
Unlike many museum collections, these works, many of which are three-dimensional, often quite large as well as fragile, are stored in various locations, mostly off-site, under conditions that render access extremely difficult even for local scholars; access for the international research community is virtually non-existent. The digitization of over 6,000 works from this collection will make these holdings available, for the first time and in unprecedented ways, combining images of the highest resolution and appropriate text with user interfaces and exceptionally flexible search mechanisms developed by Luna Imaging.
This project is conceived primarily to meet the exacting needs of scholars. It addresses the form that visual collections need to take in a digital environment if they are to support research and scholarly practice. From the start, scholars in the field have endorsed this particular perspective, not only emphasizing the intrinsic appeal of making this important collection available, but also drawing attention to the project’s potential value as a methodological model for future initiatives in the digital field. Professor Barry Bergdoll of Columbia University strongly endorsed the multiple advantages of providing access to "the oldest and most complex collection of design in the world, together with the highest quality of catalogue information, and a powerful set of search capabilities hitherto unknown." Peter Rowe, Dean of the Faculty of Design at the Harvard University Graduate School of Design, predicts a broad scholarly constituency for the resource:
I think it is safe to say that the very mode of representation embodied in the project will quickly become expected in research quarters, much as the hard-copy library files of earlier times formed, shaped, and nurtured the scholarly habits of our predecessors. . . .The real power and intellectual usefulness of the image resources lie not only in the availability of the data per se, although that is certainly impressive, but in the interactive capacity that we as scholars will have as a consequence of the data system design.
In the ten months that have elapsed since approval of the grant, great progress has been made in capturing images at a very high level of resolution, supporting them with appropriate catalogue data, and developing associated information systems that will enrich the scholar’s search capacity in a variety of ways. It will take at least two more years, however, for the particular characteristics of this digital design collection to be effectively tested within the scholarly community and to be made available for wider distribution.
Wall Paintings and other Art from the Dunhuang Cave Shrines
The second pilot project offers an unusual opportunity to digitize art of great importance in a remote part of the world—Dunhuang, in northwestern China on the edge of the Gobi desert—that then can be linked, it is hoped, to manuscripts, paintings, and other materials that were once at the site but have since been widely dispersed. Dunhuang was a major stopping point for the caravan trade between China and the West, which moved along what is now called the "Silk Road." Between approximately the 5th and 10th centuries, Buddhist art was introduced to Dunhuang, a great many cave shrines were built there, and extraordinary collections of manuscripts were amassed in what was called the "hidden library" (because it was walled off in 1002 to protect it from invaders).
These treasures were undiscovered, and protected by the dry air of the desert, for almost a millennium. The "hidden library" was reopened by a Chinese monk in 1900. In the following 25 years, various European adventurers and archaeologists came upon the caves and transported a considerable number of their finds back to sponsors in Britain, France, Germany, India, Russia, and other countries. The Chinese stopped this unauthorized export of their art in the mid-1920s, moved some objects to Beijing, and have since attempted, as their resources have permitted, to protect the very considerable amount of cave art that remains.
Dunhuang is arguably the most important site of Buddhist art and culture in Asia.Wall paintings and sculptures chart the transmission of Buddhism—from India to China—through the crucial transportation routes of Central Asia. Moreover, the thousands of documents originally stored in Dunhuang’s library cave provide one of the most extensive records available of the exchange of goods, people, ideas, and languages across the world’s largest land mass. The creation of an online Dunhuang Archive that re-connects the remaining cave art with the most important manuscripts and objects now dispersed all over the world would produce an invaluable scholarly resource for the study of the history, art history, archaeology, religion, and culture of medieval China. It would also serve to introduce East Asian murals more broadly to the art historical community and make them available for detailed study in ways not possible with traditional photographic techniques or even on-site visits. The Archive has the potential to transform the way that Chinese art is viewed and discussed and to serve as a model for reuniting materials now physically dispersed in many libraries and museums around the world. (Note 6)
Professor Sarah Fraser at Northwestern University first brought the possibilities of this project to the attention of the Foundation, and the Trustees approved three separate appropriations to Northwestern in 1999 in support of the photography, digital image processing, and cataloguing of selected portions of the cave art in Dunhuang. Experimental photography in two caves (carried out by Harlan Wallach of Northwestern University) has been extraordinarily successful in producing images of a quality that would allow scholars to see some areas of the caves far more clearly than is possible on the site. However, agreement still has to be reached with the Chinese authorities on continued access to the caves and the handling of intellectual property rights (see below) before the major part of the digitization of the cave art can be undertaken. Representatives of the Foundation have made two visits to China to discuss these complex matters, and we have been encouraged by the clearly expressed desire of our Chinese colleagues to join with us in crafting an agreement that will meet the needs of all parties. Current estimates suggest that a full program of cave photography and database construction would cost well over $3 million, and the Trustees have committed the Foundation to provide this funding, assuming that suitable legal agreements are in place. In addition, the Trustees have agreed, in principle, to support the digitizing of manuscripts and other works of art that now reside in museums outside China, and the integration of these images into a Mellon International Dunhuang Archive for which the Foundation would take responsibility. Finally, an appropriation was made in 1999 to the Dunhuang Research Academy to support activities associated with the commemoration of the centennial of the reopening of the Library Cave.
This is obviously an ambitious, costly, and highly complex project. It also presents an unprecedented opportunity to demonstrate the potential value of a truly international collaboration that addresses:
• scholarly issues (selecting the content to be captured and organizing it appropriately);
• technical challenges (creating, cataloguing, and storing images of a wide variety of objects, some of them three-dimensional and very large, as well as manuscripts written in many languages, and then developing a database architecture that facilitates searching, "zooming," and manipulating the contents of the Archive to serve research and teaching needs);
• legal issues (defining intellectual property rights as well as rights and obligations of other kinds) in an international context;
• political sensitivities (related in large part to the dispersal of an important part of the Chinese cultural heritage and the corresponding need today to reach agreements across many jurisdictions if these scholarly materials are to be reunited electronically);
and, finally
• organizational /business issues (how the Archive is to be assembled and made available in a manner that facilitates scholarly use while assuring the sustainability of the project).
Daunting as they are, the need to address these issues is an intriguing opportunity to work through basic issues likely to affect the viability of many other art digitization projects.
The Broader "ARTSTOR"Concept
In considering whether to make the large appropriations necessary to create MoMA’s Digital Design Collection and the Mellon International Dunhuang Archive, two important questions were: (1) whether these scholarly resources could be made available to users in a way that met their needs while simultaneously protecting the rights of the content providers; and (2) how the archives could be sustained and could continue to evolve without permanent underwriting by the Foundation. It would have been hard to justify the expenditure of the funds and the substantial efforts needed to conceptualize and carry out these projects, absent reasonable confidence that the resulting archives could be accessed by scholars worldwide in some satisfactory way—and that resources would be available to make these collections available on a continuing basis, which includes providing for their migration to new technological environments.
The very positive experience of the scholarly community with JSTOR led us to believe that an analogous approach might be useful in addressing these central issues, and the current plan is to use an "ARTSTOR"-like entity to manage and distribute the MoMA and Dunhuang archives. Such an entity might also offer libraries, museums, and other potential users several other significant advantages:
• the convenience of being able to obtain access to more than a single digital art collection from one repository;
• a standard user interface and a single mechanism for providing user services;
• economies of scale in pursuing technical questions, including the continuing development of database architecture;
• a recognized method for monitoring usage and encouraging compliance with legal agreements pertaining to intellectual property rights (and, perhaps, a centralized mechanism for collecting and remitting publication fees);
• a cost-effective way of housing whatever infrastructure is required for these and other purposes; and, finally,
• a "meeting place" for scholars and others interested in these subjects.
The Trustees discussed the "ARTSTOR" concept at length at their retreat in the fall of 1999, concluded that it is promising, and expressed their strong support for investing resources in its further development. At the same time, all involved in these discussions recognize that this new territory is uncharted, that uncertainties abound, that "ARTSTOR" differs in fundamental respects from JSTOR, that the ideas under discussion are highly preliminary, that there is a need to proceed incrementally and cautiously, and, finally, that it is important not to overreach. We also believe that others need to be involved in thinking with us about these concepts.
The comments that follow, which identify a number of questions that are still very much under discussion, are intended, therefore, to inform interested parties of our thinking and to encourage comments and suggestions.
In thinking about how an "ARTSTOR"-like entity might evolve, the most important issues involve the selection of content. There are major differences between the content of JSTOR (which contains standard journal literature that is, among other things, widely held, used in reasonably predictable ways, and added to on a regular basis through the publication of current issues) and the far more amorphous field of art (in which museums and other institutions own a diverse array of unique objects that are of interest to scholars and others for a wide range of reasons). The Trustees are unanimous in their view that the content of the proposed "ARTSTOR" archive should be driven by careful consideration of the needs of scholars, who in turn will be relied on to give the Foundation disinterested advice on the importance of particular fields and collections. The present intention is to be highly selective and not comprehensive. The field is too broad and the costs of creating content are too high. In thinking about the kinds of content that might merit inclusion over the long run, certain basic principles can be proposed. First, a standard of high quality (for example, in image capture, cataloguing, and system design) should be established and maintained. Second, it would be important to avoid the "greatest hits" approach. That is, we would want to avoid a situation in which museums and other content providers alone decided what images to contribute to the database. Broadly speaking, the objective would be to assemble a series of "digital collections," each of which would have a clear scholarly integrity and coherence and enough depth as well as breadth to be genuinely useful to scholars. (Note 7)
As a practical matter, we are concentrating first on developingthe MoMA and Dunhuang "collections" as successfully as possible. As the earlier discussion indicates, this is no modest charge. (Note 8) These two initial "stakes in the ground" satisfy the principles outlined above, and they should appeal to very different scholarly audiences.We are also in the preliminary stages of examining the possibility of developing a "third stake," which might consist of an ancient studies project. A well-chosen project of this kind might address some of the same objectives as the first two projects (especially the desirability of enhancing access to hard-to-study monuments, spaces, and related materials that are widely dispersed). An ancient studies project has the added appeal of providing an opportunity to link monuments and art objects to the vast journal literature on classics and classical sites that JSTOR plans to include in its expanded arts and sciences collection. If the proposed "ARTSTOR" experiment were to succeed, we would anticipate adding other suitable "collections" over time.
There are also a great many business and organizational issues to which we would like to direct attention, but with the understanding that many of them are far from resolved. It is difficult to know with any precision what level of costs would be involved in establishing and maintaining a mature "ARTSTOR" entity. Obviously much would depend on the amount of content to be included. We can be sure, however, that the initial costs of digitizing art collections will be substantial, largely because of the unique nature of each work of art and the need to capture fine detail (i.e., nature of material, shadow, brush work, color registration, etc.). Thus, the cost of digitizing art objects is likely to be significantly higher than the cost of digitizing text—as our two pilot "stakes in the ground" illustrate so clearly. Moreover, the wider variety of works of art means that there are likely to be fewer economies of scale. As our colleague, Angelica Rudenstine, pointed out in recommending the appropriation to MoMA:
The nature of the materials in the visual arts presents inherent difficulties that are not shared by the print-matter field. Works of art [unlike books and periodicals] are unique to each institution or individual collection, and photography of each item to the requisite level of quality (which may or may not exist for any group of targeted works of art) will therefore always be cumulatively expensive.
It is possible, however, that an entity like "ARTSTOR" could encourage more rapid development of common solutions to some of the main technical problems including image resolution, which, if successful, might yield their own economies. At present, there is a risk that individual scholars and projects will duplicate both the work and the mistakes of others, without adding a great deal of new value. Creating baseline measures of cost for various kinds of digitization might also be useful. In recognition of the "research and development" function that needs to be served in this broad area, the Trustees recently made a grant of $1 million to the Institute for Advanced Technology in the Humanities (IATH) at the University of Virginia. The grant is intended to accelerate the development of the tools, methods, and procedures needed to manage two- and three-dimensional imagery in the fields of architecture, archaeology, and the visual arts.
In thinking about the revenue side of the equation, a starting assumption is that institutional users (who, initially at least, are expected to be mainly college and university libraries, centers for advanced study, and museums) would make some payment to gain access to the collections in "ARTSTOR," as they do now to use JSTOR. It is impossible to know what level of user fees would be generated, but it seems almost certain that the revenue-generating capacities of this kind of entity are more limited than are those of JSTOR. (Note 9) There is, however, also the distinct possibility that overtime parts of this kind of a repository might be of interest to a broader public, including community learning centers and publiclibraries. Other future possibilities include the development of specialized applications (in distance learning courses, for example) and the granting of access on a controlled basis to some share of the enormous population of individuals who go to museums. The rapid development of Web sites and dot.com enterprises warns against ignoring such possibilities, hard as it is to calibrate their potential role in meeting the costs of maintaining such archives. The central conclusion we draw from this discussion is that "ARTSTOR" cannot be expected to be self-sustaining if all costs are counted, including the high initial costs of creating digital collections. It seems clear that start-up funding is essential, that some subsidization of even recurring costs could be required, and that if new collections are to be added, significant philanthropic investments in creating content would be required. A major advantage of this model is that decisions to add content can be linked to the availability of funding. Foundations and philanthropically minded individuals exist, of course, precisely to pay for worthwhile things that would not be possible otherwise. All the uncertainties and open questions notwithstanding, we are optimistic that the "ARTSTOR" concept will prove viable.
Intellectual Property Rights
It has been evident for some time that almost every Foundation project involving the application of information technology raises intellectual property rights issues of one kind or another and that in important instances the ability to resolve these issues determines the very viability of the project. It has been much less evident how these issues should be approached, in part because the law in this area is so unsettled. Even when guidelines exist in the print world, their application in the electronic arena is often far from clear—and hotly contested. (Note 10)
Last year’s annual report listed broad principles adopted by the Foundation to govern the ownership of intellectual property produced in part with Foundation funds. That first effort to establish ground rules was stimulated mainly by grants to Middlebury’s Center for Educational Technology to improve language teaching and, more generally, by grants made as part of the Foundation’s program to support Cost-Effective Uses of Technology in Teaching (CEUTT). (Note 11) More recently, much time and effort have been devoted to thinking about what principles should govern access to the rapidly growing array of digitized text and images that can be made available on the Internet. In general, the Foundation favors making access to such materials as broadly available as possible, but in ways that are consistent with the need to meet the ongoing costs of maintaining and enhancing databases. Costs need to be recognized explicitly and covered, in some instances by subsidies and in other cases by site licenses or other kinds of user fees. Many of the world’s greatest libraries have a long and honorable tradition of absorbing these costs and making their resources freely available to users, but this model of complete subsidy is not universally applicable. The appropriate structure and mix of subsidy and user charges will vary depending on the circumstances—including the nature of the scholarly resource, the expense of maintaining it, and the sources of funding used to create it. When the Foundation makes grants to assist libraries, museums, and other owners of intellectual property to create digital resources, it expects to reach agreement with the grantees on policies governing the use of whatever new resources are produced. To put the point another way, the Foundation does not make grants to create new intellectual property that the grantee can then exploit in whatever way best serves its own self-interest without regard to broader societal objectives, including the reasonable expectations of scholars.
Since the publication of last year’s report, and as a result of discussions with many experts and a number of grantees, we have learned a great deal about whether, when, and how to reserve intellectual property rights. In light of this experience, the Foundation has adopted a new set of operating principles that include explicit understandings that:
• The Foundation will not assert intellectual property rights when: (a) its investment in a project is de minimus; (b) the Foundation has no special interest in the subsequent pricing or distribution of a product developed with Foundation funding; and/or (c) the assertion of intellectual property rights would create harmful disincentives for grantees.
• The Foundation will reserve rights with respect to pricing, use, distribution, and/or revenue-sharing when: (a) its investment is substantial; (b) an objective of the grant is to promote wide and equitable access to the intellectual property being developed; (c) the property in question can be usefully incorporated in, or linked to, other projects the Foundation is funding; (d) the Foundation is contracting out for the development of a specific product or is the principal architect of the product; and/or (e) significant revenues are anticipated which may be usefully applied to other projects.
We have also learned some "process" lessons. In particular, we now undertake to identify and resolve intellectual property rights issues early in the grantmaking process. In the case of some early grants related to the development of JSTOR, we were slow to address issues related to the ownership and control of software that was being developed, and the result was much confusion and the need later to spend considerable time fixing problems that never should have arisen in the first place. Grantees clearly deserve to know the Foundation’s intentions and requirements at the outset of an agreement to undertake a Foundation-sponsored project.
These operating principles and an effective process for applying them were developed by Michele Warman, the Foundation’s recently-elected General Counsel and Secretary, working in concertwith other Foundation staff. Intellectual property rights issues are proliferating at the Foundation, and this is clearly an area in which there is simply no substitute for having excellent staff members "in house." It is no exaggeration to say that the Dunhuang project, for example, could not have gone forward without the day-today involvement of Ms.Warman and the Foundation’s Assistant General Counsel, Gretchen Wagner. (Note 12)
The art digitization projects in which the Foundation is involved have raised the most complex kinds of intellectual property rights issues, and one useful by-product of the extended discussions with the Chinese authorities has been valuable experience in thinking through issues which lack off-the-shelf solutions and which are certain to arise in many other contexts. In the case of both the MoMA and Dunhuang projects, the Foundation indicated from the start that it would be able to provide funding only if the grantees and other participating entities were comfortable giving the Foundation (or a designee such as "ARTSTOR") rights of two basic kinds:
• First, the royalty-free right to incorporate the "collections" of digitized images and other content in an archive of the Foundation’s creation (for which the Foundation would seek a "compilation copyright").
In the case of the Dunhuang project, this means the right to include in the Mellon International Dunhuang Archive not only images of the Dunhuang cave art, but also images and text obtained from the many other places that hold materials that once were at Dunhuang, and to be able to provide the tools that will allow scholars to link or "reconnect" these materials to the images of the cave paintings. More generally, the Foundation, acting reasonably, has to have the right to assemble collections of various kinds under a single "ARTSTOR" umbrella.
• Second, a royalty-free license to distribute the digital archive worldwide to civic, educational, and charitable entities on terms that it regards as appropriate and consistent with its educational objectives.
The Foundation itself is ill-equipped to serve the distribution function, and it needs therefore to be able to select a distributor and to assure the distributor that it will not have to pay royalties for electronic content that the Foundation has helped to create.
At the same time, the Foundation also seeks to protect the legitimate rights of the content providers (MoMA and the Dunhuang Research Academy, in these cases). Specifically, MoMA and the Dunhuang Research Academy will be the sole owners of the master images of their art, which they are free to use as they see fit, for educational or commercial purposes. In addition, the Foundation (or its designee) will take appropriate steps to protect the rights of the content providers: (1) only licensed users will be given access to the "ARTSTOR" archive; (2) carefully constructed user agreements will regulate the ways in which the images can be used and will facilitate the normal payment of fees for the reproduction of an image in a scholarly publication; (Note 13) and (3) "ARTSTOR" will monitor(as JSTOR does now) the electronic use of its archive in order to detect any unusual patterns of use that might suggest, for example, the improper downloading of large numbers of images.
While working out the details of agreements of this kind is difficult and time consuming, experience to date is encouraging. When there is a shared set of objectives, trust, and mutual respect, it is possible to reconcile the multiple interests of participants. This is perhaps the most important lesson we have learned through these negotiations.
Academic Libraries in a Digital Age
The library component of this "program" is, if one counts its antecedents, the oldest one, dating back to the 1970s, when my predecessor, Jack Sawyer, led an initiative designed to encourage collaboration and resource sharing among libraries. The Foundation’s much more recent investments in reinvigorating and linking library consortia in Eastern Europe and South Africa are a quite direct extension of what was at its time path-breaking work in building incipient library networks. Partial evidence of the extent of this activity in the US is provided by the history of the Foundation’s support for The Council on Library and Information Resources (CLIR) and its predecessor organizations, the Council on Library Resources and the Commission on Preservation and Access. These entities have been grantees since 1976, and had received a total of over $16 million prior to the most recent grant of $2.6 million.
Since its inception, CLIR’s primary audience has been the research library community. CLIR has sought to assist leaders at universities and other research institutions in understanding library and archival issues and to enable them to make well-informed decisions about the future. Those activities continue, at the same time that CLIR has adopted a broader agenda.
In 1999, the Foundation’s principal focus in the library field was on liberal arts colleges. In December alone, the Trustees approved appropriations totaling $3,475,000, which was designed to strengthen the role of the library on liberal arts college campuses. Grantees included several consortia (the Five College group based in Amherst, the Associated Colleges of the South, the Five Colleges of Ohio, based in Gambier, and the Bryn Mawr/Haverford/Swarthmore group), several individual colleges (Carleton, Reed, and Vassar), and one service organization (CLIR).
This set of grants grew directly out of a series of meetings Foundation staff had with presidents, provosts, and librarians from selective liberal arts colleges, who expressed concerns about the changing role of the library on their campuses. Broad, philosophical concerns are joined to two highly practical issues.
• First, many institutions continue to run short of space for books and are interested in designing useful off-site storage facilities that might be shared with other institutions. Individual institutions (especially large universities) have used off-site storage for a long time, but now there is growing interest in collaborative approaches to this problem. The Five College group based in Amherst has converted a former US Air Force Strategic Command headquarters (a two-story bunker, built 32 feet under the surface of the Holyoke mountain range) into a Five Colleges Depository for off-site storage.
• Second, ways clearly need to be found to ensure that library and technology staff maintain the necessary skills to perform well in this new electronic environment. The grant to CLIR will allow promising mid-career librarians and technology staff at liberal arts colleges to attend the Frye Leadership Institute at Emory University. This Institute, named most appropriately in honor of Billy Frye, a long-time champion of libraries and librarians, exists to educate leaders to meet the challenges of managing hybrid systems of print and electronic information, establish programs supporting both traditional and new modes of scholarship, and communicate effectively about the choices available to the campus community.
More fundamental challenges arise from the advent of the technologies themselves, the availability of information in electronic form and the growth of personal computers on campuses—all of which make it possible for students and faculty members to conduct research from their dormitory rooms and offices. Notwithstanding this wealth of new information and the new modes of access to it—and perhaps even because of these "enhancements"—presidents report that faculty members believe that their students’ research skills have deteriorated. The presidents feel frustrated in their efforts to involve faculty, students, librarians, and technology staff in redefining the role for the library in a fast moving electronic environment. Many librarians express their own frustrations, derived principally from their view that faculty have not been effectively engaged in the new opportunities the library can provide to enhance teaching and learning on campus.
The colleges and consortia that received these grants are addressing this range of issues in a predictably wide variety of ways. Getting people to "cross boundaries" can be surprisingly difficult, even in small, rather intimate settings; what works on one campus may not work at all in a seemingly comparable institution. Perhaps the most ambitious of these efforts to give the library a new intellectual role on campus is occurring at Vassar, which is nearing the completion of a major renovation and addition to its library. This renovation will include new settings for instruction using electronic technologies that are designed to strengthen the intellectual connections between the classroom and the library and to foster active collaboration among faculty, librarians, and technology staff within the library. The new facilities include a large computer classroom, an electronic seminar room, five electronically equipped group study areas, and a media cloister. The cloister will house the college’s existing Center for Electronic Teaching and Learning and will be used to integrate technology resources across the campus as well as provide professional support for the use of advanced technologies for teaching, learning, and research. Vassar’s overarching objective is to reestablish its library as the intellectual center of the campus in the era of new technology. (Note 14)
The presidents and librarians say over and over again that exchanging ideas and experiences with one another is extremely valuable. The staff member at the Foundation organizing these discussions is Pat McPherson, and she and her associates intend to work closely with these institutions to bring key participants together to assess progress and consider future directions in an environment that is far from stable.
Some commentators believe that the forces that serve to "liberate" individuals from dependence on physical locations such as libraries are gaining strength rapidly. They cite as evidence the development of new Web sites and portals aimed directly at students and other consumers of academic information, and the belief of some commercial entities that there is money to be made in these ways. The Foundation itself has encouraged a kind of "liberation" by developing resources such as JSTOR that reside on servers at centralized locations and are just a mouse click away. At the same time, there is the danger that the ready availability of electronic resources will lead to more intellectual isolation. A central challenge for libraries and librarians, some of us believe, is to become ever more effective in managing access to the new digital resources.
This major responsibility involves deciding how to evaluate various digital offerings, how to consolidate and synthesize them, and, in effect, how to turn them into knowledge. Even more so than in the past, librarians must become both effective teachers of research methodologies and active participants in the shaping of courses and curricula.
Libraries should also, in this view, remain both physical and symbolic centers of intellectual life on campus—places where it will still be possible to hold a book that has become an old friend in one’s hand and to contemplate how little of the accumulated knowledge of times past any of us can command. Libraries should remain places where faculty and students with diverse interests come together to celebrate shared enthusiasm for that greatest of all gifts, a new idea.
William G. Bowen
President
The Andrew W. Mellon Foundation
140 East 62nd Street
New York, NY 10021
(212) 838-8400
Notes
