2003 Annual Report: President's Report
In reporting this year on the work of the Foundation, I want to concentrate on two developments that can be said to have reached "milestone" status in 2003:
- First, two major digital initiatives, ARTstor and Ithaka, were launched as independent but affiliated entities, following a period of development and preparatory work within the Foundation. They are intended to build on lessons learned through the earlier (1995) launch of JSTOR. ARTstor is a not-for-profit organization that is building a digital repository of images of works of art and related scholarly material along with the software tools that facilitate the searching and active use of its collections. Ithaka (which was established collaboratively by the William and Flora Hewlett, Andrew W. Mellon, and Stavros S. Niarchos Foundations) has the broad mission of accelerating the adoption of productive and efficient uses of information technology for the benefit of the worldwide scholarly community. 
- Second, the legal status of race-sensitive admissions programs in colleges and universities was clarified by the United States Supreme Court's decisions in the University of Michigan cases. During 2003, the Foundation's staff and Trustees considered carefully the implications of these decisions for our programs (most notably MMUF, now named the "Mellon Mays Undergraduate Fellowship Program"). The result was reaffirmation of the importance of MMUF and a modest, but significant, broadening of its mission and eligibility criteria.
In order to accommodate a reasonably full treatment of these major developments in the first two sections of this report, the last section will necessarily be brief, and will consist only of illustrative comments about other ongoing programs. No one should interpret this allocation of space as reflecting any lack of substantial progress in the other programs of the Foundation—which, taken together, accounted for nearly 90 percent of the Foundation's appropriations in 2003. Program staff have continued to work in creative ways with colleagues at grantee institutions to strengthen our core fields of higher education, scholarly communications, research in information technology, museums and art conservation, the performing arts, and conservation and the environment. Previous Annual Reports have described these programs in more detail, and we will return to them on other occasions.
Launching the ARTstor and Ithaka Affiliates
The official launching of these two new organizations is the culmination of nearly five years of active discussion (marked by much good advice from a wide variety of experts in relevant fields, friends in academia, the library and museum communities, and other interested parties), legal research, and experimental "probes" of various kinds. As readers of earlier Annual Reports will know, it was the highly positive reaction of the scholarly community to JSTOR, more than any other factor, that encouraged the Trustees of the Mellon Foundation to sponsor ARTstor and then, in collaboration with the Hewlett and Niarchos Foundations, Ithaka. It is not easy to calibrate the "return" on philanthropic investments, but there would be widespread agreement, I think, that the Foundation's early support of JSTOR was one of the best investments our Trustees ever made. JSTOR has been extraordinarily successful by any measure and can claim, among other things, to have played a not-inconsequential role in changing fundamentally the way students and scholars around the world access journal literature, as well as the ways in which libraries think about their preservation responsibilities. The Foundation's investment in JSTOR has also demonstrated the workability of the basic financial model of "shared responsibility," whereby philanthropic contributions are combined with institutional user charges to help not-for-profit organizations achieve sustainability. As an independent, financially self-sustainable organization, JSTOR continues to generate a return on the Foundation's original philanthropic investment. 
JSTOR's success in turn stimulated consideration of what more the Foundation could properly do to encourage the development of other promising digital initiatives that were consistent with its core commitment to the humanities. As one Trustee put it, "ARTstor was a 'natural' for Mellon." 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. ARTstor offered an opportunity to extend this interest in the fine arts during a new era that is being shaped profoundly by advances in information technology. The "fit" between the new technology and visual images is an unusually promising one. The ability to combine—and make active use of—images, data, texts, and other materials offers the opportunity to bring about a substantial transformation in art-related teaching, learning, and research. 
The subsequent decision to create Ithaka was informed by our experience with ARTstor as well as JSTOR, and was based on: (a) a conviction (shared by our funding partners) that digital technologies offer exceptionally powerful instruments that can serve important educational purposes across many fields of knowledge; (b) careful consideration of the Foundation's comparative advantage in stimulating new uses of fast-evolving technologies within the worldwide higher education community; and (c) the realization that grantmaking foundations are not structured to incubate new organizations like ARTstor, Ithaka, and Ithaka's progeny (discussed later in this report) for extended periods of time. 
From the Foundation's perspective, the case for investing in this domain rests on a combination of capacities, some within the Foundation and some outside it, that we believe we are in an unusually good position to utilize. These include: the considerable knowledge and skills of staff members associated with previous and ongoing projects at the Foundation as well as at JSTOR, including their practical experience in working through legal and business issues; excellent institutional relationships already in place; the financial and organizational resources needed to launch any large-scale initiative that requires a reliable infrastructure and economies of scale to succeed; the credibility associated with the success of JSTOR; the leadership and entrepreneurial skills of Kevin Guthrie, who is now President of Ithaka; and, above all, a reasonably clear sense of the underlying values intrinsic to the humanities and higher education. Technology is seen by our Trustees as a powerful means to an end, but never as an end in itself.
Ithaka can be regarded as a mechanism for "crossing boundaries," "breaking down barriers," and fostering collaborations within the arts, the humanities, and higher education on an international scale. About four years ago, when we asked a group of university presidents to advise us on how the Foundation could invest its resources most wisely, one university president (Richard C. Levin of Yale) said directly that what he wanted most from Mellon is "dollars-plus" grantmaking, where the "plus" is meant to represent some additional conceptual and organizational contribution beyond support for existing identifiable needs. Others present agreed that the emerging technologies are creating system-wide opportunities (and needs) which no individual college or university, or even a group of them, is likely to be in a position to address successfully. A catalytic contribution from one or more respected third parties will be critical in the years ahead, and that contribution is most likely to be made by an organization perceived to be "above the fray" (free of the competitive instincts that often make it hard for individual institutions to assign a high priority to serving broader constituencies), able to marshal the necessary resources, and willing to "stay the course" in terms of operating with a long time horizon. JSTOR is one clear example of what can be accomplished in this way. E-Archive (discussed below) could prove to be another case in which a large scale collaborative effort will yield benefits that would be exceedingly difficult if not impossible to achieve otherwise—in this instance, by bringing together libraries, publishers, and technological expertise to solve the problem of creating a centralized repository or permanent archive for content created and distributed only in electronic form ("born-electronic").
It of course remains to be seen how successful ARTstor and Ithaka will be in serving their stated purposes, but the Trustees are convinced that the potential philanthropic "returns" on these investments are very high. These returns are expected to take the form of (among other things) significant advances in the quality of teaching and learning, worldwide, and the identification of cost-effective ways of achieving these advances. It is highly unlikely, however, that these investments will lead to financial returns that would even cover, much less generate a return on, the very substantial upfront commitments required to launch these enterprises. Therefore, it is difficult to imagine that organizations of this kind will be formed successfully on a for-profit basis. Entities such as ARTstor and Ithaka are mission-bound to pursue objectives that for-profit entities cannot be expected to address, such as the long-term preservation of important scholarly materials, the creation of electronic content that is important from a scholarly perspective but that appeals to a very narrow community, and the broad dissemination of new ideas. Projects like creation of the Dunhuang Archive would be impossible in the absence of large-scale philanthropic support. 
Laurance Rockefeller has long been fond of saying that every foundation, and every philanthropist, should nurture at least a few "tall trees," and it is our hope that the ARTstor and Ithaka seedlings will grow into trees that in their own ways are as "tall" as JSTOR.
As of January 1, 2004, both ARTstor and Ithaka were functioning as independent not-for-profit organizations, classified by the IRS as "public charities," with their own boards, management structures, and start-up funding.  (Status as an independent 501(c)(3) public charity is important for many reasons, one of which is that it permits the organization to pursue the goal of sustainability by attracting funding from users, through institutional user charges, as well as grant revenues.) Staff members who had been on the rosters of Mellon and other organizations were transferred smoothly to ARTstor and Ithaka as the year ended. The one remaining aspect of "the launch" yet to be accomplished involves bringing ARTstor and Ithaka staff together in a space of their own. 
We use the term "Affiliates" to refer to these young entities, as well as to their older JSTOR cousin, to connote the close working relationships among the three.  The benefits of close collaboration are in large part a result of the fact that all of the Affiliates have to deal, in one way or another, with the same broad sets of issues. These include selection of criteria for building and aggregating content; developing relationships with content owners as well as users, and resolving intellectual property rights concerns; creating and maintaining repositories, platforms, and tools that permit digital resources to be distributed and accessed in reliable and effective ways; and building sustainable organizational and business models. The pooling of talent and sharing of ideas and experiences—in all of these areas—have already proved to be of great value. There is no need to repeat mistakes made earlier by one or another of this family of entities with complementary missions.
In building any digital resource, the content that is to be made available is critically important, and ARTstor has devoted a great deal of time to thinking about the selection of its initial collections—recognizing that many, many more collections will be added over time. Its suite of six "Charter" collections reflects the combination of a commitment to provide both a core teaching resource that is broad enough to be considered reasonably comprehensive, and a varied set of "deeper," more specialized, collections that will appeal to users in fields ranging from Buddhist art to modern architecture and design. These Charter collections are illustrated on the ARTstor Web site (http://www.artstor.org), and this is an instance in which it really is true that "a picture is worth a thousand words." We hope readers interested in ARTstor will judge the quality of the resource for themselves. Brief descriptions of each Charter collection follow. 
- The Image Gallery is meant to offer the breadth and cohesion typically associated with an undergraduate art history curriculum or with an art history slide and photograph collection. In its initial phase, it will include over 200,000 images digitized from a well-photographed and professionally cataloged slide library. This collection is expected to grow steadily over time.
- Carnegie Arts of the United States consists of more than 4,500 images, and documents significant aspects of the history of American art, architecture, visual and material culture. It was designed in the 1950s by a panel of internationally recognized scholars and, in its slide format, has been widely used for teaching American art history for several decades.
- The Illustrated Bartsch offers approximately 50,000 images of Old Master European prints (engravings, etchings, woodcuts) from the 15th to the early 19th centuries. The collection is derived from the great 19th-century reference catalog by Bartsch of Master European prints—a catalog that has now been updated in 96 volumes. To the extent possible, ARTstor has worked from the original photographic prints rather than from the published book illustrations.
- The Mellon International Dunhuang Archive (MIDA) is the product of a major multi-institutional, multi-national effort to create high quality digital images reuniting the mural paintings and related art and texts associated with the several hundred Buddhist cave shrines in Dunhuang, China, an important cultural crossroads on the ancient Silk Road at the edge of the Gobi Desert. This Archive includes not only stitched images of original capture photographs of the wall paintings and sculptures in more than 40 of the most important cave grottos, but also high quality images of the sacred and secular scrolls, manuscripts, textiles, and silk paintings once located at Dunhuang and now dispersed around the world, and, finally, more than 2,500 historic photographs from the Lo Archive in Princeton that were taken at Dunhuang in the midst of the Civil War in the 1940s. All told, MIDA will contain more than 10,000 images and will reunite "virtually" a rich body of primary material from a site of great religious, historic, and aesthetic importance.
- The Huntington Archive of Asian Art will initially contain a corpus of approximately 10,000 images and associated metadata digitized from the much larger collection of color slides and black and white photographs taken by Ohio State art historians John C. and Susan L. Huntington over nearly three decades. Works photographed date from 3,000 B.C. through the present, and include images from India, Afghanistan, Pakistan, Bangladesh, Sri Lanka, Nepal, China, Japan, Thailand, Indonesia, and Myanmar. A large number of the photographs document monuments, architecture, and art works that have since been destroyed, stolen, lost, or modified.
- The Museum of Modern Art (MoMA) Architecture and Design Collection (DDC) consists of more than 6,000 objects (represented by approximately 8,000 digital images) from one of the foremost collections of its kind. The DDC includes architectural drawings, models, and photographs; graphic design materials such as posters; and three-dimensional objects such as appliances, furniture, tableware, tools, textiles, and sports cars. Only five percent of the works in the collection have been published, the majority is not on permanent exhibition, and many are stored off-site under conditions that render access extremely difficult.
I have described each of the individual Charter collections to provide some sense of the range and reach of ARTstor even in its earliest days. As previously noted, the collection-building strategy has been to surround a large and varied core collection created largely as a teaching resource (The Image Gallery) with a variety of more specialized collections that in some cases contain rare materials heretofore unavailable or exceedingly difficult to access. But of course everyone involved in this undertaking is well aware that this is just the beginning. Over time, ARTstor will both enlarge its core "teaching" collection and add other specialized collections. As Neil Rudenstine, Chairman of ARTstor, likes to say: "In a hundred years, ARTstor may be able to begin to make some claim of being comprehensive."
In 2003, the Mellon Trustees approved several grants intended to move this process along. These grants are supporting several smaller-scale ARTstor projects that nonetheless seem to be of high potential value (such as digitized images of sculptures and artifactual objects comprising approximately 20 percent of the famous Congo and Native American collections of the Museum of Ethnology in Berlin). Support has also been provided to develop new digital collections that range from Old Master Drawings (a portion of the Gernsheim Photographic Corpus of Drawings held by the British Museum), to Islamic Art and Architecture (slide and photographic collections assembled by research university scholars that constitute a rich overview of Islamic arts over many centuries), to The Image of the Black in Western Art (an international collection at Harvard University covering a 5,000-year time period), to Gothic and American Architecture (from the Clarence Ward Archive at the National Gallery of Art), to Native North American Art and Historic Photographs (from the National Museum of Natural History of the Smithsonian Institution).
In the course of building the Charter collections and anticipating the continuing addition of new collections, ARTstor has developed a rich understanding of what is involved in working with a broad spectrum of institutions from around the world, and of the costs and benefits of different image strategies ranging from high-quality but very costly digital photography and scans of original art works (at Dunhuang and MoMA, for example) to the less expensive digitization of excellent 4×5 photographs, to the digitization of 35mm slide collections. Often it is the less glamorous task of collecting and organizing—or even creating—catalog information and metadata that is the most challenging aspect of a project. It is even easier to overlook the enormous amount of work (and expense) involved in establishing and maintaining proper standards in capturing images, exercising quality control in reviewing them, and developing effective data loading processes. The sizes of the raw databases that have to be analyzed and organized are staggering. Images require much more digital space than text, and the Mellon International Dunhuang Archive alone requires 12 terabytes of storage—as compared with 1 terabyte for the entire JSTOR database! It is no easy task to integrate into a single database massive amounts of content from projects spread around the world.
Property Rights and Legal Issues
Gaining permission to digitize content is an ongoing task that requires careful legal analysis, a sympathetic understanding of the goals and concerns of others (including especially those in the museum community), and thoughtful negotiations. ARTstor's statement on intellectual property rights reads (in part) as follows: 
ARTstor recognizes both the importance of respecting intellectual property rights and the significant needs of the educational community to have access to images of art works and other content for teaching and study. ARTstor's approach to intellectual property rights seeks to balance the interests of educational users, intellectual property rights owners, and other content providers....Several principles guide ARTstor's efforts to achieve this balance.
First, ARTstor will limit access to, and the uses that may be made of, the images and scholarly information in the ARTstor database. ARTstor will not distribute the high-resolution images in the ARTstor Library on an open, unrestricted internet Web site. Instead, the only institutions that will have access to the database will be non-profit institutions...[and] these institutions will be required to enter into user agreements with ARTstor, and individual users will be subject to terms and conditions of use, restricting uses of the images to those that are noncommercial, educational, and scholarly in nature. Users who seek to make commercial uses of the images...will be required to seek permission from the appropriate rights' holders. Additionally, ARTstor will employ technology to facilitate authentication (assuring that users are authorized members of a participating institution) and authorization (confirming that legitimate users have appropriate access to any particular resource) before users can make certain uses of the database.
Second, ARTstor itself will clearly not make commercial uses of the content of the ARTstor Library....While ARTstor will charge institutions fees for use of the repository, these fees will only help to defray ARTstor's significant operating costs.
Third, ARTstor will rely on a variety of approaches to obtain content....In some instances, ARTstor will collaborate directly on projects with owners and rights' holders. In other instances, it may only seek permissions for use. In still other instances, it will rely on fair use and other educational exceptions to copyright laws.
Extensive consultations with legal experts as well as a wide variety of content owners, including leading museums, have led us to conclude that the ARTstor approach is workable, especially when accompanied by continuing conversations with museums, artists, and other rights holders so that there is both mutual respect and much shared understanding. Looking ahead, ARTstor has sought to assure institutions accessing its database, as well as those who create and care for works of art, that it stands ready to work with them in addressing any concerns that may arise. The willingness of ARTstor to pursue questions in this area as the law continues to evolve should be reassuring to both providers and users of content. We believe that working through these legal issues, inside and outside the US, is itself a major contribution to nonprofit institutions responsible for scholarship and teaching, to faculty and students, and of course to the rights holders themselves. 
Tools and Technology
In 2003, ARTstor decided that it made sense for it to build up its own capacity to create the tools and software environment needed to work with its collections, using off-the-shelf components when possible. ARTstor's Chief Technology Officer, Bill Ying, has assembled a highly capable team, and one of the major achievements of 2003 was designing ARTstor's repository, and a browser-based, easy-to-use set of tools for viewing, storing, and presenting images. ARTstor's interface incorporates icons and navigational tools that are already familiar to many users. The tools that have been created provide a great deal of functionality. They allow users to search across both individual collections and all of the collections taken together (to find, for example, all of the Van Eyck paintings in the database, or all of the images of tigers, or all of the images associated with the Russian Revolution). The user can then "open" an image by clicking on a thumbnail, zoom in on any feature of special interest (to see, for example, the tile work employed by an architect in joining one detail of a building's exterior to another, or the clarity of form and feature in the face of a Dunhuang bodhisattva), examine the catalog data associated with an image, compare the image with other images on the screen at the same time, create groups of images (adding, subtracting, and "dragging and dropping" images as desired), add personal annotations, and store groups of images for presentations or future reference. As testing proceeds, other components are being added, including a personal collections capability (which will permit users to integrate their own images into a group of ARTstor images). Also, thought is being given to ways of linking ARTstor to JSTOR, and especially to its collection of art history journals.
As always, there is more that can and should be done—there are new ideas to explore, connections to be built to learning and content management systems, including SAKAI, which is being developed by an important Foundation-sponsored university collaboration mentioned later in this report, and so on. Right now, however, the need is to resist all such temptations and concentrate on providing as good an experience as possible for early users of ARTstor.
Paying for ARTstor
It should surprise no one that creating and sustaining ARTstor is a daunting task from yet another standpoint—paying the bills! To date, the Foundation has treated the operating costs of ARTstor (which now has a staff of 16 full-time employees and approximately 20 part-time employees and consultants) as direct charitable expenditures. In 2003, these expenses including allocations, totaled about $9 million. In addition, in 2002 and 2003, the Foundation appropriated $10 million to fund a large part of the ongoing operating expenses that ARTstor will incur now that it is no longer being incubated within Mellon. Finally, grants in support of specific content development projects have totaled approximately $14 million since 1999, when the MoMA and Dunhuang digitization projects were launched as probes into a complex new domain. 
In thinking about the costs of ARTstor—which far exceed the costs of JSTOR—it is necessary to consider the nature of the resource that has been created and is now to be offered to interested parties (beginning in the fall of 2004). ARTstor is a resource comprised of "primary" materials for users of visual images, laboriously assembled from a very large range of different sources. JSTOR, by contrast, is a curated archive of "secondary" materials (the journal articles produced by scholars who have worked with such materials, as well as with original documents, records, and databases), and these journals are then digitized directly from the journal copy for read-only purposes. Put another way, ARTstor deals not with the end products of scholarship, as JSTOR does, but with the visual raw materials of scholarship and of teaching. These raw materials are the common "stuff" of the humanities and have very broad potential appeal; but they have to be collected, checked, assembled, and then made available—"presented"—in ways that facilitate their use in different contexts. The high-resolution images cannot just be "thrown out on the table" if their value is to be realized. If a user cannot locate an image, cannot study it appropriately with the help of excellent technology and reliable metadata, it might as well not be there. Thus, ARTstor images have to be accompanied by well-organized descriptive data, a new tool kit (embedded in ARTstor's Web site) of considerable independent value that will permit users from licensed institutions to study images in detail, and a well-conceived set of legal agreements with users. It is much more than "just" a repository.
In the lexicon of the marketplace, those institutions that obtain full access to ARTstor will receive four interconnected "products":
- A very large (and growing) primary resource—the database—containing carefully selected collections of high-quality images, including many readily available for the first time, and metadata describing them;
- A large inter-institutional network that links non-profit educational institutions, offering them ready access to all the "pooled" (or aggregated) collections, from many different sources, in the database;
- A new tool kit of considerable independent value, that will permit users to study images in detail, to use them for classroom presentations, and take full advantage of their capacity to advance both teaching and research; and
- A set of legal protections that it would be difficult, if not impossible, for individual institutions to craft on their own.
ARTstor hopes that all interested parties will understand the substantial complexity and the very high level of costs associated with this venture. In all likelihood, ARTstor's participation fees will necessarily be modestly higher than JSTOR's, simply because all aspects of its work are more labor-intensive, technologically complex, and difficult to sustain. The capital costs associated with building, maintaining, and making active use of images are very real, and the extent to which ARTstor "displaces" these costs for institutions should be recognized, even though ARTstor may seem at first glance to be just another "content subscription." It is expected that ARTstor will be available in the fall of 2004, and fee schedules are being set now. Every effort will be made to keep them as low as they can be, consistent with the need for ARTstor to generate enough revenue to allow it to cover most of its ongoing operating (but not capital) costs. At the same time, as I have said in earlier Annual Reports, we have never believed that the economics of ARTstor would turn out to be as positive as the economics of JSTOR. We have understood from the start that ARTstor, like education itself and most other worthwhile not-for-profit activities, will almost certainly need to seek continuing philanthropic investments.
Much will depend, of course, on the value that institutions and individuals attach to ARTstor and on how willing they are to help cover some part of the operating costs. The imputed savings for institutions that are potentially active users of ARTstor are considerable, and this fact, if it is communicated effectively and appreciated at the appropriate decision-making levels, could make a considerable difference. Eventually, if ARTstor continues to grow and develop as we think that it will, it could save educational institutions the substantial capital and operating costs associated with digitizing much of their own content and attempting to build software tools comparable to those that will be provided by ARTstor. Over the longer course of time, institutions may be able to realize cost savings associated with the upkeep of their slide libraries—and save faculty and students immeasurable amounts of time and effort in assembling and working with visual materials. In addition, institutions would not have to invest in attempting to create a protected inter-institutional network that allows them to access one another’s visual collections (and data) and to "pool" their content in order to build a continually growing database. Finally, ARTstor may save nearly all institutions the complications—and heavy costs—involved in finding solutions to many complicated intellectual property rights issues which ARTstor is prepared to address on behalf of the larger scholarly community.
Initial efforts to project costs and revenues, and to build a business plan, lead us to be cautiously optimistic that there can be a fair sharing of the costs of ARTstor, especially over the longer run. The Foundation's Trustees have made clear their intention to provide sufficient resources to ensure that ARTstor will have a good opportunity to attract supporters (including not only users but perhaps other funders too). It will need to demonstrate its value convincingly, and we believe it will do just that—perhaps even appealing to classes of users that are not even recognized now.
This is a time of major technological transition in higher education, and Ithaka  is committed to accelerating the most productive applications of information technology by: incubating new entities with highly focused missions, such as E-Archive and Aluka (described below); carrying out a research program that informs these initiatives and others; and assisting all of the Affiliates to operate in a cost-effective manner by providing a common infrastructure, a common software platform, and a suite of shared services. Later, Ithaka may also provide strategic advice and assistance to selected projects.
Perhaps the most important function of Ithaka is to stimulate the creation and development of promising ideas/projects that can meet targeted needs. One of Ithaka's first "incubation" efforts—provisionally called “E-Archive”—has a mission that is simply stated, but very challenging: "To preserve scholarly literature published in electronic form and to ensure that these materials remain available to future generations of scholars, researchers, and students."
The initial focus is on journal literature that is "born electronic."  When a library subscribes to a journal in print format, it can, if it chooses, store the paper copy on its own shelves so that it will always be available—or it can rely on some other library or library consortium to take responsibility for this preservation function. But what should happen in the electronic world? It makes no sense for each library to attempt, on its own, to solve the problem of storing electronic content and migrating that content to new platforms as technology changes. It would certainly seem that a common, coordinated solution to this problem is needed, but to date no such solution has emerged.  E-Archive is seeking to fill this void.
As Eileen Fenton, Executive Director of E-Archive, has explained: 
The academic and publishing communities have moved into the 21st century with ever-increasing reliance on digital content, but the infrastructure for preserving this content has not been created. If electronic resources are to form a permanent part of the record of scholarly achievement, the long-term preservation of these resources must be assured. While there is an increasing level of concern about this situation, none of the individual constituents in the scholarly communication system is in a position to act independently to solve this problem.
The benefits of finding a solution are considerable. First, ever improved access to electronic content benefits students and scholars, who generally find it much more convenient to access electronic resources (using increasingly powerful searching and linking capabilities) than to find what they want in traditional periodicals on library shelves—or, for that matter, in personal copies in their own offices. JSTOR and other collections of online journals have demonstrated dramatically the appeal of a system that allows users to find and print the scholarly literature that they want from any location, at any time of the day or night (the library for online journals is never closed, and nothing that is wanted is ever out or ever defaced).
In addition to the evident scholarly advantages, there are also large potential cost savings. Academic libraries face growing financial pressures, and switching to electronic-only access to a growing part of the corpus of scholarly literature (as more and more journals are available in electronic formats) could lead to savings in subscription fees, storage costs, and the operating expenses associated with handling paper versions of journals (receiving them, processing them, reshelving them, binding them, preserving them, and so on). A recent study of 11 academic libraries (led by Roger Schonfeld, who is coordinating the research function within Ithaka), found that the non-subscription costs of print periodicals, measured per title on a discounted "life-cycle basis" (i.e., over 25 years), are consistently higher than the comparable costs for periodicals in electronic format.  The magnitude of the difference varies by size of library, the nature of the journal, and many other things, but the pattern is utterly consistent. Annual savings in the hundreds of thousands of dollars seem potentially achievable. However, such operational savings can be realized in a responsible way only if a preservation solution is found to ensure that electronic journal publications can be stored and accessed permanently.
Thus, it is hardly surprising that concerns about insuring the longevity of electronic journals have prevented many libraries from making a large-scale shift to electronic-only subscriptions. Journal publishers are feeling pressure from the library community to create archiving arrangements with trusted third parties since it would be hard for the publishers themselves to "guarantee" that they would provide perpetual access to their own journals— especially if a journal or even a publisher ceases to exist.  Moreover, libraries and users would presumably prefer a repository that is comprehensive and permits searching across the content provided by many publishers.
It is in this setting that E-Archive proposes to offer an archiving service for electronic scholarly journals that will focus on the core activities essential to archiving itself—data ingest, verification, normalization, storage, delivery, and migration. In this way, publishers would be assured that the intellectual content of their journals will be preserved, along with selected functionality such as reference linking. Libraries would receive a guarantee that a vast collection of scholarly e-journals will be preserved for the long run. Libraries would also like some level of ongoing access to the content of the repository, both to assure them that the content is really "there" and to provide a useful resource for teaching and research. E-Archive itself expects to maintain an operational focus on archiving, but it certainly recognizes the great importance of searching and access. (To this end, E-Archive is exploring the possibility of entering into a partnership agreement with one or more organizations, such as JSTOR, that possess the ability to deliver access.)
To date, the E-Archive staff have been engaged in building a prototype archive; obtaining commitments from a group of publishers to participate in a pilot, developmental phase of the project by providing e-journal data that can be evaluated within the archive; talking with libraries and other representatives of the academic community about their e-archiving needs and how to meet them; and attempting to build a realistic business plan that offers some prospect of sustainability. In large part because of the "public goods" aspect of an archive of this kind, and because of the complex interconnections between the economics of scholarly publishing and the economics of library operations, creating a workable business model is a daunting challenge. E-Archive's present plan is to seek funding from multiple sources, including publishers, libraries, government agencies, foundations, and potentially JSTOR itself (which might use E-Archive to process electronic journals and as a back-up repository for its own archived content). E-Archive's centralized approach should make it possible to spread the costs of creating and maintaining a very large archive of this kind across a broad set of beneficiaries.
Looking further ahead, it is possible to imagine E-Archive taking responsibility for many other kinds of content besides journal literature. Organizations of many kinds, including especially libraries such as the New York Public Library, are interested in preserving digital records of holdings of various kinds, and yet they may not see themselves as best positioned to provide and maintain a suitable repository. But this is a large question for another day. In the near term, E-Archive's goal is to generate enough support from interested parties to begin to give a reality to its vision—and then to learn more by "doing."
Last year's Annual Report identified a second candidate for incubation, an international project under the leadership of Thomas Nygren, that was then referred to as "NIDR," or "Network of International Digital Resources." The basic idea, which was reaffirmed in 2003 as a result of a great many conversations with scholars from around the world, is to build over time an interconnected set of broadly conceived scholarly resources ("clusters of content"), including valuable primary materials, that would focus on materials from and about other parts of the world, especially developing countries. This content would be distributed over the Internet, following the general principles and practices employed by JSTOR (and, starting in the fall of 2004, by ARTstor). Each cluster will have enough breadth and depth to be of genuine scholarly interest in both its region of origin and in the US and many other parts of the world.
Aluka has chosen to start its collection-building in Africa, and principally in sub-Saharan Africa initially (with work in other regions to follow shortly if the broad concept continues to appear promising). This choice was made because of: the clear need for such resources in Africa; the strong interest of scholars and students in other parts of the world in African-based content; and the Mellon Foundation's promising earlier experience in working on related projects in South Africa, in particular, an experience that has served to identify able partners in the region. There is great interest in projects of this kind in Southern Africa—a deep desire to contribute indigenous content as well as to benefit from the aggregation of related content that resides in libraries, archives, and centers for advanced study all over the world. 
A new name has been chosen for this rapidly evolving entity: "Aluka," which is derived from a Zulu word that means "to weave" and thus reflects Aluka's mission of joining together in a single place scholarly resources from around the world that relate to a single theme, very much like the Dunhuang project in ARTstor.  Major progress was made in 2003 in defining the first two sub-Saharan clusters—"Struggles for Freedom in Southern Africa" and "African Plants and Their Uses"—and in laying the groundwork for their relatively rapid development.
Struggles for Freedom in Southern Africa. Allen Isaacman, Regents Professor for History at the University of Minnesota, an expert on Mozambique, and a former President of the African Studies Association, is serving as Content Advisor for this cluster. One key collection within this cluster is a digital archive focused on literature published during South Africa's struggle under Apartheid. Much of this collection has already been digitized by the Digital Imaging South Africa Initiative or DISA, which has been sponsored by the Foundation and is based at the University of Natal. Additional grant support for DISA was approved by the Foundation's Trustees at their December meeting, and early reactions to the content being assembled by this group (much of which is hard to locate and in danger of being lost) have been highly positive. Major progress was made in 2003 in exploring other opportunities in southern Africa, especially in documenting the history of Mozambique in moving from Portuguese rule to independence. This has involved identifying key materials in Mozambique (including the personal papers of Mozambique's founding President, Samora Machel), and key sources of related content in the US, and in Europe (including Lisbon).
African Plants. African plants are of great interest to scientists in Africa and around the world. Thanks in large part to the leadership provided by William Robertson, Program Officer at the Foundation for Conservation and the Environment, great progress has been made in a very short period of time not only in defining an African Plants cluster for Aluka (see below), but also in enlisting the enthusiastic participation and support of leaders in the field from around the world, including Peter Crane, Brian Huntley, and Peter Raven (respectively, Director of the Royal Botanic Gardens, Kew, in the United Kingdom; Director of the National Botanic Institute, South Africa; and Director of the Missouri Botanic Garden). The Foundation has made appropriations thus far to eight leading botanical gardens, institutes, and herbaria in South Africa, the United Kingdom, Ethiopia, France, Germany, Belgium, and the United States in order to pay for the digitization of a large majority of the known "Type specimens" of African plants. (A "Type specimen" is the original dried representative of the actual plant preserved on a sheet of paper and stored in an herbarium that is associated with determining that the plant is a distinct species; questions of plant identification, relationships, and related information ultimately refer back to the Type for the species.)
Africa's history of colonial governments has meant that African Type specimens are today held mostly in developed countries, and a major contribution of this digitization project will be to create a highly organized and highly accessible information resource that will allow researchers to view and compare Type images at their home institutions rather than traveling to one or more distant herbaria to accomplish the same tasks. Creation of this resource will also mean that information about African plants will be accessible at institutions in Africa itself for the first time. The African Plant resource is organized around Types and their associated names. It will contain: high quality images of the Types; data from the Type sheet and other sources; images of related botanical illustrations, artwork, and photographs; information on use from published sources; and archival material such as the field notebooks of the collectors associated with the Types. In time, Aluka intends to add related content, in order to extend the usefulness of the cluster to a variety of disciplines beyond botany. This cluster, no less than "Struggles for Freedom," seems to us to illustrate again the appeal of the basic idea of "weaving" together (aggregating) various kinds of content from all over the world—a theme that plays through much of the work being done by all of the Affiliates.
Other Clusters. During 2003, considerable thought was also given to other potentially interesting sub-Saharan clusters that might complement "Struggles for Freedom" and "African Plants." One intriguing large-scale possibility of a longer-run nature is a cluster (or even several clusters) that would focus on literature and other forms of artistic and humanistic expression related to the African Diaspora. Another idea that has emerged more recently, and that appears to offer more immediate prospects, is to focus on a limited number of key archaeological and cultural heritage sites across sub-Saharan Africa, documenting them as thoroughly as possible with digitized images of photographs, maps, archaeological records, excavation reports, scientific papers, books, and so on. A cluster of this kind might also be a logical place to include projects to digitize and assemble images of African rock art (the earliest record of human expression in Africa) that the Mellon Foundation is already funding.
Production Strategies. At the same time that Aluka staff and their advisors have been seeking to identify promising content, progress has been made in defining a "distributed production strategy" (whereby digitization and assembly of content will take place in many different locales) that is similar in many respects to the strategy being used by ARTstor—and very different from the centralized processes used by JSTOR. The need for a distributed approach is driven largely by the unique and fragile nature of much of the relevant content—in most cases content owners will not permit the materials to be taken off site, or if they will, they may prefer a local, trusted vendor. If the content owner does not have its own digitization facilities, Aluka may have to help set up small scanning labs on the premises of the institutions and train staff in their use. The present plan is to have a relatively small, and mobile, production team based in the US, responsible for establishing technical standards, assessing new collections, developing production plans, coordinating the various remote projects, and ensuring quality control. Fortunately, Aluka has DISA as a partner in South Africa for the "Struggle for Freedom" cluster, and we expect DISA staff to be very helpful in working with other groups in the region. This approach, while unquestionably more complicated and more expensive, has the advantage of helping to spread the skill set involved in carrying out such projects, thereby "training the trainers" for future projects. 
Space constraints allow only the briefest mention of a third nascent organization that Ithaka intends to incubate. As readers of previous Annual Reports will know, the Foundation's Liberal Arts College program, under the leadership of Pat McPherson, has invested considerable funds in supporting a set of regional technology centers that oversee and carry out a range of programs for the benefit of colleges in different parts of the country. We now believe that there could well be value to these centers, to the colleges they serve, and to Ithaka itself (as well as ARTstor) if a somewhat more centralized organizational structure could be created and housed within Ithaka. The current name of this entity is "NITLE" (which originally stood for "National Institutes for Technology and Liberal Education," and is pronounced "nightly"), and it may now have a life of its own as an acronym!
To date, the regional centers have been very responsive to the needs of their constituent colleges. They have designed a broad array of programs which use technology to address general challenges faced by liberal arts colleges or that seek to remove obstacles to the effective adoption of technology within these colleges. Programs range from small, specialized workshops (e.g., a seven-day Music Learning Objects Team workshop with five participants) to larger scale programs such as conferences (e.g., Rethinking Teaching and Learning Spaces for 220 participants) and curriculum development (for example, an Arab Culture and Civilization site). Programs have also been created to meet less direct needs, such as Web resource development, software development (a project to enhance search engine technologies), non-technical research projects, and representation in national policy debates. Finally, there are programs to meet very specific needs such as listservs, discussion boards, and newsletters. Now that considerable experience has been gained with this range of programming, the time seems right for a more focused approach to be developed through the leadership of Jo Ellen Parker, former President of the Great Lakes Colleges Association, who has agreed to serve as Executive Director of NITLE.
One challenge for Ms. Parker and others involved at both national and regional levels in shaping the future of NITLE is to determine not only specific "offerings" that will be attractive to the colleges, but also to develop a business plan that will allow NITLE to sustain itself over time. Thus far, the Mellon Foundation has provided all of the funding for NITLE, and this support will continue for another year and a half. The colleges themselves will then need to contribute resources in return for the benefits they receive. A related question is how a new NITLE organization and Ithaka's Affiliates might help each other. One possibility is that NITLE would include training or other special programs related to ARTstor, E-Archive, and Aluka as one of its "participation benefits."
In addition to incubating new entities, Ithaka has also made progress in serving a less publicly visible aspect of its mission—in particular, to develop and administer mechanisms for sharing infrastructure and other services among JSTOR, ARTstor, Ithaka itself (collectively, the "Affiliates") and Mellon. The basic objective is simply to operate as cost-effectively as possible. There is no reason why each of the Affiliates (and Mellon) needs its own internal IT organization, or its own Human Resources staff, or its own library, and so on. A Master Services Agreement has been developed that provides the framework for collaborations in these areas and sets out in detail what each entity will provide and what each will expect. For example, responsibility for the provision of IT services is centered in Ithaka, and Mellon and the other Affiliates will rely on Ithaka for these services, with each entity paying its fair share of the allocated costs. This kind of "pooling" will permit, we believe, a higher level of service to be delivered (with more specialization of staff than would be possible otherwise), at a lower cost. 
Minimizing the costs of infrastructure is always important, but in some ways it is more important in the foundation world today than at any time in recent memory. Concern has been expressed in Washington and elsewhere that some foundations spend more "on themselves" than they should, relative to the funds they invest in advancing their charitable purposes. Whatever the merits of such charges, the best way to rebut them is by operating in as cost-effective a mode as possible, even though sharing services can (and does) add its own form of complexity. Trust is required, as well as good performance, but the effort seems well worth making. More generally, not-for-profit entities as a class need to be alert to opportunities to hold down administrative costs through service sharing and pooling arrangements.
One other activity in which Ithaka is making a substantial investment is the development of a shared software platform that is intended, in the first instance, to benefit all the Affiliates. Although each Affiliate has specialized requirements of its own, there is reason to believe that some real economies of scale can be achieved by working collaboratively. A core software development team is now being established within Ithaka, which will seek to build and implement common service components that each Affiliate can deploy to meet its specific requirements. It is expected that these tools will include a production work-flow framework, the hardware and software specifications to support archival repositories, a set of platform services (security, statistics, logging, monitoring, etc.), and a set of software tools for administration and management. If valuable service components are developed, it may also be possible to make them available to other not-for-profit entities with similar needs, but it is much too early to be at all sure that this is a realistic goal.
These last paragraphs suggest a more general point. Ithaka differs from JSTOR and ARTstor in one fundamental respect. Although it too, through its incubated entities, is seeking to make valuable digital resources available to the worlds of scholarship and teaching, it operates at a somewhat different level. Ideally, entities such as E-Archive and Aluka will make their own ways in the world, and at some point they could become large enough and established enough to function as independent affiliates, as JSTOR and ARTstor do today. Ithaka will work hard to support them, and to do all it can to promote that outcome, at the same time that it explores (judiciously) other opportunities to stimulate and help to develop promising ideas. In addition, by supporting research, providing advice to others seeking to meet related needs, creating new organizational models, and managing shared services, Ithaka may be able to serve a broader, more "integrative" function within the higher education sector. If it does evolve in this way, it will need to secure funding from a variety of philanthropic sources, and to be seen as a useful partner of foundations and others that have similar aims. This sounds ethereal, I realize, and it is unavoidably speculative. But it seems desirable to state Ithaka's prospective mission in these broad terms as we conclude this account of what it is doing at present and where it stands today.
For those who believe, as many of us associated with the Foundation do, in the importance to our country of continued, determined efforts to reduce racial disparities, 2003 was a very important year. The Supreme Court's holding in Grutter v. Bollinger was heartening in that it provided strengthened judicial support for the nuanced consideration of race in admissions (as one among other factors). At the same time, the Court's decisions in the two University of Michigan cases reinforced concerns about racially exclusive programs and highlighted the need to reexamine the structure of programs such as MMUF—an effort led by Michele S. Warman and Lydia L. English, to which other staff members and Trustees also devoted a great deal of time. The main purpose of this section of the Annual Report is to describe the conclusions we reached and the reasons for them. First, however, some history and context are needed.
History and Context
When I arrived at the Mellon Foundation in 1988, the first new program that we launched was what was then called the "Mellon Minority Undergraduate Fellowship" program, or simply MMUF. At its inception, MMUF was led by Henry Drewry, an extraordinarily able African-American teacher, scholar, and administrator, with whom I had worked closely at Princeton. It was evident to both of us, on the basis of our immediate Princeton experience as well as on the basis of incontrovertible national data, that colleges and universities desperately needed more faculty members from minority groups if they were to succeed in recruiting and educating properly an increasingly diverse student body. Princeton's considerable success in recruiting larger numbers of highly talented minority students (especially in the undergraduate college) was not matched by anything like equal success at either the graduate school or faculty level. Nor was it easy to persuade very bright minority undergraduates to enter PhD programs with the goal of becoming faculty members. 
As Mr. Drewry put it, in reviewing the early history of MMUF in the 1993 Annual Report of the Foundation:
The principal short-term objective is to increase the number of highly qualified candidates for PhDs in core fields within the arts and sciences who come from minority groups that are seriously underrepresented in these fields (African-Americans, Hispanic Americans, and Native Americans). The longer-term objective is to increase the diversity of faculties at colleges and universities throughout the country in order to bring a wider range of experiences and perspectives to teaching and scholarly discussion. By providing increased opportunities for all students to work with minority professionals, diversity serves the related goals of structuring a campus environment more conducive to improved racial and ethnic relations and of providing role models for all youth. 
At the time of that report, African-Americans constituted 11.5 percent of the undergraduate population but earned less than 4 percent of all PhDs. Similarly, Hispanics constituted 7.6 percent of the undergraduate population but earned only about 3 percent of all PhDs.  By 2002–2003, the general situation was essentially the same: African-Americans and Hispanics constituted, respectively, 10.9 and 6.6 percent of the undergraduate population; each group earned less than 4.5 percent of all doctorates in the arts and sciences. 
Mr. Drewry went on to discuss the factors responsible for the limited numbers of minority candidates for PhD programs:
Several causes are often cited to explain the dearth of minority students enrolling in doctoral programs in the arts and sciences. Most basic is the small number of minority students at the earlier stages of the education pipeline....At the college level, many of the most talented minority undergraduates, who come disproportionately from economically disadvantaged backgrounds, have opted for careers in business, law, or medicine. These professions offer better income prospects and also are seen by many as offering clearer opportunities for upward mobility. Moreover, minority students find few role models on arts and science faculties, and they are also less likely to have grown up knowing academicians as members of their families or communities. 
It would be hard to improve on this account today—even though real progress has been made, in part because of MMUF and complementary efforts supported strongly by other foundations, corporations, individual colleges and universities, and government agencies. Since its inception, MMUF has had over 2,000 participating students, and the attrition rate continues to be extremely low. Over 100 MMUF participants have already obtained their PhDs, approximately 500 are currently in graduate school (with many more completing their undergraduate studies or taking time off before going on to graduate school), and the number earning PhDs each year continues to accelerate. Today, MMUF "alums" teach in a variety of fields at a wide range of colleges and universities. 
Useful as statistics are, personal references are often at least as compelling. One early MMUF student from Oberlin credited his mentor with "bringing my mind to life." Just this spring, Dartmouth College noted with pride the academic achievements of two seniors who are MMUF fellows: Khristina Gonzalez, who is investigating the portrayal of women in Old English and contemporary Italian literature; and Jerome Green, Jr., who is exploring the influence of the Baha’i faith on American race relations. The same Dartmouth publication called attention to the presence on the Pennsylvania State University faculty of Associate Professor of Latin American History Ben Vinson, a ’92 graduate who was also an MMUF Fellow. Numerous other vignettes could be presented.
These success stories can be attributed in no small measure to the excellent leadership the program has had from its inception, with the current Director and Assistant Director of MMUF (Lydia English and Carma Van Allen, respectively) matching the high standard set by Mr. Drewry and reinforced by his successor, Jacqueline Looney. Equally important have been the distinguishing features of the program itself: delegation of responsibility for personalized selection of candidates to individual campuses, where campus coordinators are chosen and put in charge; targeted recruitment of promising students in their sophomore year; careful identification of mentors who pay close attention to their MMUF students; provision of opportunities for academic involvement (teaching or research) during both the academic year and the summer, with modest financial subventions provided by the Foundation; annual conferences which lead to intellectual stimulation, practice in writing and presenting papers, and the strengthening of support networks; the promise by the Foundation that it will repay, in a staged manner, undergraduate debt incurred by MMUF students of up to $10,000, as they progress through their PhD programs; and, more recently, targeted support for research by MMUF students during graduate study and after they obtain initial faculty appointments. But much experience has taught us that the quality and commitment of the campus coordinators are most important of all.
MMUF Today: Mission Statement, Selection Criteria, and Name
After an internal review of MMUF, the Foundation's Trustees voted in June to reaffirm in the strongest terms their unequivocal commitment to the fundamental goals of the program. To date, the Foundation has invested a total of nearly $50 million in MMUF, and we expect to continue to make substantial investments for the foreseeable future. At the same time, following extensive consultation with presidents and other representatives of participating colleges and universities, with campus coordinators, and with outside counsel, we decided that we should modify the program’s official mission statement, its selection criteria, and its name.
The present-day legal and political environment was without question one stimulus for these revisions. The widely publicized challenge to MIT's summer enrichment-recruitment-mentoring program for minority students, MITE²S, by the Center for Equal Opportunity and the US Department of Education's Office for Civil Rights, made clear that simply "standing pat" raised real risks that any race-exclusive program might be shut down by injunction. I know of no university in the United States more committed to addressing racial disparities than MIT, and President Charles Vest and others at MIT were determined to preserve the essentials of their highly successful program. After interminable discussions, huge investments of time, and what can only be construed as threats to shut down the program, MIT continued its program in all essential respects—but only by modifying it to make disadvantaged students from other groups also eligible. MIT's experience made clear the desirability of acting proactively, and of finding ways to continue to serve the basic purposes of such programs without exposing individual colleges and universities to unnecessary legal risk. Several of our own MMUF institutions received letters similar to the one that precipitated the debate over the MIT program, and they naturally and properly turned to the Foundation for advice. Our goal was to maintain the integrity of the MMUF program—strengthening it if possible—while reducing the risk that participating colleges and universities would be unable to survive potential legal challenges. 
The most vulnerable aspect of MMUF was of course its race exclusivity, which was reflected in its original mission statement, its selection criteria, and its name. Only members of "underrepresented minority groups" were eligible to apply—a direct approach that seemed to make a good deal of sense at a time, 1988, when the focus of the program was so squarely on increasing the number of members of underrepresented minority groups who would apply to PhD programs. In the intervening years, we have concluded that our real objective is deeper and more far-reaching. Therefore, we broadened the MMUF mission statement to read:
"The fundamental objectives of MMUF are to reduce, over time, the serious underrepresentation on faculties of individuals from certain minority groups, as well as to address the attendant educational consequences of these disparities."
The rationale behind this revised statement of objectives is:
- Reducing racial disparities among faculty members and including faculty members committed to eradicating such disparities will facilitate efforts to increase the diversity of students enrolled in undergraduate and graduate programs, in part because the presence of more faculty role models will encourage more minority students to apply and to enroll.
- The increased presence of minority faculty members and faculty members committed to eradicating racial disparities can also be expected to increase retention of minority students, through improving the educational climate and enhancing the support structure for these students—in part by giving them the opportunity to form meaningful intellectual and personal relationships with faculty members from a wide variety of backgrounds.
- Larger numbers of minority faculty members and faculty members committed to MMUF's mission will enhance the educational experience for all students by: a) introducing a wider array of backgrounds, which is likely to broaden the perspectives voiced in classroom discussions and in campus life in general; and b) breaking down stereotypes and promoting cross-racial understanding through exposure to distinguished scholars from previously underrepresented groups.
- Faculty members are themselves important leaders in American society, serving in many influential roles both on and off campus, including in business, civic life, and government, and more minority faculty members and those committed to MMUF's mission will benefit society at large by bringing more diverse leadership and understanding to our increasingly diverse population.
At the same time, we adopted more inclusive criteria for eligibility, so that students of all races and ethnic backgrounds will now be eligible to apply for participation in MMUF. But we also made clear that special efforts will continue to be made to recruit large numbers of students from underrepresented minority groups, and that MMUF's goals can be achieved by both increasing the number of students from these groups who earn PhDs and by supporting the pursuit of PhDs by students of all races who have demonstrated a commitment to the goals articulated in the mission statement.
This line of thinking led to the adoption of a revised (and more explicit) set of selection criteria that will be put into effect at all participating institutions. These criteria are:
- Academic promise.
- Interest in pursuing an academic career in Mellon-designated fields of study.
- Potential for serving as a mentor and teacher for a wide variety of students.
- Race and ethnicity (in relation to underrepresentation in designated fields of study).
- Demonstrated commitment to increasing opportunities for underrepresented minorities, breaking down stereotypes, increasing cross-racial and ethnic understanding, and enabling others to better understand persons of different races and ethnicities.
- Commitment to participating fully and enthusiastically in all aspects of the MMUF program, including attendance at conferences and meetings.
- Status as US citizen or permanent resident (except for students at the University of Cape Town, South Africa).
Selection committees on individual campuses will evaluate all candidates in light of this full set of criteria. As in all selection processes that take into account more than a single criterion, no one candidate will have to satisfy every criterion, and decisions will be made on a case-by-case basis by comparing each individual's qualities with those of the rest of the applicant pool. There is no "goal" of enrolling a specific number of minority or non-minority participants, although we expect that a very large fraction of MMUF participants will continue to be minority students. Each participating institution will have to ensure that its selection procedure is open and fair, and that it respects the stated criteria. Thereafter, the numbers of participants from different racial groups should be allowed to go where they will—this is not a quota program.
The final change was in the name of the program. To celebrate the success to date of the program, and to signal both its restatement of objectives and its high hopes for future contributions, the Foundation has renamed it the "Mellon Mays Undergraduate Fellowship" program. The intent is to honor Dr. Benjamin E. Mays, the noted African-American educator and former President of Morehouse College, who exemplifies so many of the goals of the program.  The "MMUF" acronym, which has become so important in its own right, remains unchanged.
The changes just described were discussed at length with the presidents and other representatives of the 34 colleges and universities (plus the UNCF schools) that participate in MMUF, both before they were adopted and afterward. Lydia English and Carma Van Allen worked closely with campus coordinators to ensure that the reasons for the changes were well understood and to address details of implementation (a process that was helped greatly by the work of two task forces of coordinators); they also communicated with MMUF students, past and present, to minimize risks of misunderstandings and to enlist support for MMUF going forward. This was not an easy process—in large part because of considerable resentment of the aggressive tactics used by well-financed groups of anti-affirmative action advocates. It is not surprising that there was often an initial tendency to say: "No way; we are staying right where we are!" Full disclosure: this was my own first reaction.
Further reflection, and much discussion, led to a more thoughtful, more constructive response. The outcome of the University of Michigan cases without question strengthened the case for continuing to use race as one criterion in making selection decisions of various kinds. Unlike Bakke, five justices signed the same opinion supporting the race-conscious admission process used by the University of Michigan Law School. Also, the Court now recognized, for the first time, not only the on-campus educational benefits of diversity but also the broader benefits to the professions and to society at large. From the perspective of MMUF and affirmative action broadly conceived, this was—and is—very good news. At the same time, although the holdings of the Michigan cases apply only to admissions decisions, the arguments by the Court in these cases (in particular those emphasizing that race can be only one factor among others considered in admitting students) make race-exclusive programs such as MMUF in its original form increasingly vulnerable to legal challenge by detractors. As this reality became clearer to more people, the desirability of modifying MMUF on our own terms, and in advance of potentially debilitating attacks, became widely accepted.
In addition, as we considered in detail how we might modify the mission statement, the selection criteria, and the name, we saw opportunities for strengthening the core values of the program. It is, after all, both the existence of gross racial disparities on faculties and the attendant consequences of these disparities that are profoundly troubling. Adopting this broader, more fundamental view, of what really matters has been helpful, we believe, in refocusing our thinking. At the end of the day, it is going to take a concerted effort by dedicated, committed individuals from every racial group to bring about the changes that are needed. The experience of MMUF has already demonstrated that non-minority mentors—and there have been many from the inception of MMUF—can be enormously effective participants in achieving the aims of the program. The cross-racial interactions, and friendships, resulting from many of these mentor-student relationships have been valued by all concerned, and there is every reason to believe that this same principle can apply within the ranks of MMUF students. There are surely non-minority graduate students (prospective faculty members), firmly committed to the objectives of MMUF. They can help reduce racial disparities by encouraging future minority students of their own to pursue academic careers and by demonstrating their own commitments through the ways in which they teach, write, advise students, and serve in leadership capacities within higher education. It is easy to think of examples of non-minority academics who would have been excellent MMUF fellows. 
The new challenge for MMUF will be to identify such candidates. The explicit emphasis in the selection criteria of having already "demonstrated commitment to increasing opportunities for underrepresented minorities, breaking down stereotypes,...and enabling others to better understand persons of different races and ethnicities" should be very helpful in this regard, as should the next criterion, which is "commitment to participating fully and enthusiastically in all aspects of the MMUF program, including attendance at conferences and meetings." These criteria should be important in the sorting process, first by encouraging the right kinds of self-selection among non-minority candidates, and then by providing faculty members responsible for choosing among the candidates with attributes to consider in making individualized choices.
The application of these criteria, combined with the explicit recognition that members of underrepresented minority groups are to be given special consideration and that targeted recruitment of promising students from these groups is to be encouraged, will in all likelihood produce new MMUF cohorts that contain very large numbers of minority students. Moreover, fears that the more inclusive approach to eligibility now in place will "dilute" the cohesiveness of the MMUF group need to be tempered by recognition that MMUF is by now so strong, and has such a pervasive "fellowship feeling," that it would be difficult indeed to change its ethos—even if anyone wanted to do that. As one experienced person put it, "It would have been one thing to be racially inclusive when the program was first launched; it is quite another to add non-minority students to the mix now that the program is so firmly established." Only time will tell how the numbers work out, and if the ethos and accomplishments of the group change, but we believe there is reason to be optimistic that the revisions to MMUF will position it well to go from success to success. We are encouraged by the excellent spirit and continuing commitment of the MMUF coordinators as well as the presidents of the MMUF institutions.
Near the end of her opinion in Grutter, Justice O’Connor emphasized her hope, shared by others, that steady progress in closing racial divides will, in time, make it unnecessary to give explicit consideration to race in making admissions decisions. She suggested that 25 years from now, the use of racial preferences will no longer be necessary to further the compelling interest approved in the Court's decision. There has subsequently been much debate about what this part of her opinion really means and whether it is at all realistic to imagine that in another quarter century we will no longer need affirmative action. These are enormously important questions, to which several of us at the Foundation plan to return in other research that is underway. In the present context, the relevant point is that programs like MMUF are more necessary than ever if we are to have any chance of meeting the O’Connor timetable. This much-debated element of her opinion provides added impetus for working as hard as we can to achieve the goals of MMUF. This is not a time—if ever there is one—simply to sit back and assume that events will somehow turn out fine on their own.
The Court's decisions in the University of Michigan cases also left many important questions open for interpretation and debate. The Foundation received a number of requests for assistance in thinking through questions about particular admissions criteria and processes, financial aid, summer enrichment programs, curricular matters, and faculty recruitment. In response, the Foundation's General Counsel, working closely with Paul Smith and others at Jenner Block in Washington, convened a well attended meeting in the fall in which presidents, general counsels, and other officers at colleges and universities shared questions and concerns, and had an opportunity to hear directly from outside counsel on these topics.
As the end of the meeting, several of us concluded that the issues we had been discussing are deeply imbedded in what Gunnar Myrdal, writing in 1944, so prophetically called "An American Dilemma."  The problems associated with race in America remain daunting; they cannot be wished away.
Other Foundation Programs
As important as the ARTstor, Ithaka, and MMUF initiatives are, they accounted for only about 13 percent of all appropriations made by the Mellon Foundation in 2003 (with ARTstor and Ithaka together accounting for about 10 percent). All told, the Foundation's Trustees approved appropriations totaling approximately $182 million, with three-fourths of this total falling within the broad field of higher education and scholarship, defined to include not only grants to research universities, liberal arts colleges, centers for advanced study, and independent libraries, but also grants in scholarly communications and research in information technology. As always, a full listing of grants made during the year appears in the back of this report, along with financial statements. The following "snapshots" are intended to illustrate some of the principal developments and emphases in the Foundation's core program areas in 2003.
Research Universities and Liberal Arts Colleges
Following the completion of the Foundation's decade-long Graduate Education Initiative in 2000–2001 (in which the Foundation invested approximately $80 million), the Program Officers with principal responsibility for research universities, Harriet Zuckerman and Joseph Meisel, led a sustained planning effort that laid the groundwork for three sets of new initiatives; the first aimed at providing substantial resources for fellowships (approximately $15 million per year) which would sustain scholars in the humanities at various stages of their careers;  the second, at funding institutions devoted to strengthening the humanities generally;  and the third, at encouraging promising developments in the core disciplines comprising the humanities. It was not until 2003 that the last of the fellowship programs, the one intended for Emeritus professors, was launched.
Like other fellowship programs the Foundation supports, Emeritus Fellowships are intended to benefit both individuals and institutions. In this instance, they enable exceptional senior scholars to continue their research after they have retired, while also allowing them to retain an active affiliation with their home institutions. With the elimination of mandatory retirement, a growing number of older faculty members have chosen not to retire. This pattern is especially evident at the major research universities and has led to significant increases in the share of faculty at these institutions who are 70 years of age or more. While the problems associated with the "uncapping" of retirement cut across all disciplines and are far larger than the Foundation could contemplate resolving, it nonetheless seems worthwhile to offer some modest reward to excellent scholars who are willing to retire—and thus to create opportunities for younger colleagues—but wish to continue active research programs.
Emeritus Fellowships are awarded to the fellows' institutions and consist of a research fund and modest additional resources those institutions may use for recipients' office space, secretarial support, and access to other services. In this first round, nominations were sought from more than a dozen research universities (13) and a smaller group of liberal arts colleges (5). From these, 16 fellows—with research projects in fields as diverse as Chinese art history, Renaissance history, American literature, religion, and musicology—were selected by a panel of scholars chaired by Phillip A. Griffiths, former Director of the Institute for Advanced Study and now a Senior Advisor to the Foundation. The majority of recipients plan on continuing work along lines they have already established, but others will use their awards for research on subjects altogether different from those on which they have spent their scholarly careers.
Now that the various fellowship initiatives are "in train," staff intend to focus more intensively on assisting institutions that strengthen the humanities on a broad basis (especially centers for advanced study), on promising institutional collaborations, and on projects aimed at developing the core humanistic disciplines themselves.
The Foundation has also continued its research on selected aspects of higher education—including studies of the outcomes of its own programs. Data are still being gathered from the institutions that participated in the Graduate Education Initiative, since the Foundation was intent, from the outset of the program in 1990, on assessing the Initiative's effects on graduate programs, and on time-to-degree and attrition rates. As these data accumulated, it became clear that they should be supplemented with direct reports from graduate students who were part of this large-scale experiment. In 2003, with the assistance of the Princeton University Survey Research Center and Mathematica Policy Research, Inc., a survey was mounted of all 15,600 doctoral students in the ten universities which participated in the program and received Mellon funds and of some 3,000 students in addition who were studying for the PhD in the same fields but were not part of the Mellon experiment. Aimed at detailing the passage of doctoral students through graduate school and later into their careers, the survey will make it possible to compare the experiences of those who completed their degrees and those who did not (to learn more about the correlates of graduation and attrition); to test some propositions about graduate education, including the motivating effects of targeted student aid; to learn more about the nature of the job market and its effects on students' completion of degrees; to identify the effects of programmatic changes as they were perceived by graduate students; and not least, to learn more about the early professional careers of PhDs in the humanities. Ronald G. Ehrenberg, Irving Ives Professor of Industrial and Labor Relations and Economics at Cornell University, has agreed to oversee the larger research project that will ultimately bring together the institutional and the survey data. In the end, these data will provide the most comprehensive record available anywhere of the experiences of graduate students in the humanities; after they have been "cleaned," they will be incorporated into the Foundation's growing data archive on higher education for eventual use by qualified researchers. 
In addition to its work with research universities and centers for advanced study, the Foundation has long supported a variety of programs at liberal arts colleges. Current projects, being carried out under the leadership of Pat McPherson and Danielle Carr Ramdath, are designed to increase educational effectiveness (often by modifying curricula and encouraging the sharing of resources); promote faculty career enhancement at different stages of a faculty member's life cycle; improve the educational contribution of study abroad programs; build a stronger sense of academic community on campus by involving faculty more fully in residential life; and encouraging new forms of library collaboration, especially those that take advantage of digital technologies. Special attention also continues to be given to the Appalachian colleges (through the Appalachian College Association) and to institution-building by selected Historically Black Colleges and Universities.
Creation of regional technology centers has been the most ambitious of the collaborative initiatives sponsored by the Liberal Arts College program, and the organizational future of NITLE and the individual regional centers has now been entrusted to Ithaka (as explained earlier in the report). The Foundation will, however, continue to provide some financial support—probably in the form of incentive grants—as NITLE and the centers make the transition to a more self-sustaining, independent status.
An even more ambitious project—designed to address the widely-felt need for better post-retirement health benefits—has also grown out of the Foundation's Liberal Arts College program, and it, too, is now planning a transition to life as an independent, self-sustaining entity. Readers of previous Annual Reports will know that this project grew out of research centered at Union College and that it has elicited strong expressions of interest from a wide variety of institutions. Originally known as the College Retirement Project, its fundamental purpose has been (and is) to permit the funding and joint purchasing of supplemental medical insurance for retiring members of not-for-profit institutions. The innovative qualities of the program are derived from three key, integrated elements: the collective buying power of higher education and allied not-for-profit institutions; a strategically defined contribution-funding approach that has definite tax advantages; and a competitive and comprehensive benefits package. With continuing Foundation support, the co-directors, Linda and Kenneth Cool, have begun to move beyond their initial focus on liberal arts colleges to include universities and other not-for-profit institutions.
The project, now officially named "Emeriti Retirement Health Solutions" (or simply "Emeriti"), is being incorporated as an independent not-for-profit organization and has a Board of Trustees chaired by John H. Biggs, former Chairman and Chief Executive Officer of TIAA-CREF. (Pat McPherson, who has done so much to get Emeriti to where it is today, also serves on the Board.) Fidelity Investments and PacifiCare have been selected as collaborators in the delivery of financial services and insurance products to participating member institutions. Informational materials have been sent to nearly 200 colleges, universities, and other not-for-profit organizations that have signed letters of interest in the Emeriti program, and four national meetings were held this winter to present the final program design. A series of regional workshops are being conducted this spring to introduce the Emeriti program to new organizations. A complete schedule of workshops can be found on the Web site, http://www.emeritihealth.org.
As always, "we will know more later" (an observation attributed to John Doar, when he was in charge of the Nixon impeachment proceedings). But there do seem to be grounds for being at least moderately optimistic that the considerable time and money invested in developing this novel plan will yield positive results. At the minimum, much will have been learned about the willingness—and the capacity—of this set of not-for-profit institutions to address an evident problem facing many individual faculty and staff members and their employers.
In addition to its extremely active involvement in all aspects of the Foundation's support of ARTstor and Ithaka, program staff in Scholarly Communications (Donald J.Waters and Suzanne M. Lodato) advanced a wide range of other initiatives in 2003. Two broad areas deserve special mention: the development of scholarly resources in history, archaeology, and music; and support for scholarly publishing.
In December, the Foundation appropriated $801,000 to the University of Oxford for continued development of the Electronic Enlightenment project. Existing scholarly editions are being digitized to create an online, searchable database of the correspondence among European and American participants in intellectual debates during the 18th-century Age of Letters. The database, which one prominent scholar has predicted will prove to be of "staggering importance," is expected to be widely available by the end of 2005. In archaeology, the School of American Research received a planning grant to develop a Center for Digital Archaeology, and the University of Virginia was awarded $683,000 to plan and create a major online research archive that collects and integrates archaeological resources pertaining to the study of the Pueblo Indians who occupied the Chaco Canyon. This archive would build on the integrated information system for the archaeology of slave plantations in the Chesapeake region developed with Foundation support at Monticello by the Thomas Jefferson Memorial Foundation. Together, the Monticello and Chaco projects would serve as models for those that the Virginia Center for Digital Archaeology expects to sponsor and maintain over time.
In the field of music, the Foundation continued its support for the Recorded Anthology of American Music's Database of Recorded American Music, a collection of digitized recordings that is currently being field tested and will soon be made widely available to educational institutions. The Foundation also made an award of $390,000 to the University of Illinois at Urbana-Champaign for the development of a testbed that would stimulate research in the field of music information retrieval. This new resource would provide a common, standardized collection of music materials and a set of standardized retrieval tasks, which researchers could then use to systematically evaluate the relative success of newly created music retrieval systems.
The Foundation's continued commitment to scholarly publishing is illustrated by two grants designed to help improve the infrastructure for the production of scholarly journals and monographs. The American Anthropological Association (AAA) received $756,000 to introduce digital technologies into its publication process and to create AnthroSource, an Internet portal that would provide access initially to current issues of AAA's 18 peer-reviewed journals. The Foundation also awarded the University of Chicago Press $1.252 million to continue the development of its Bibliovault repository and related services for short-run digital printing, which make it possible for participating university presses to keep scholarly monographs in print while realizing significant savings on the costs of warehousing physical copies.
Research in Information Technology
Two unusual projects supported through the Foundation's program in Research in Information Technology, Chandler and SAKAI,  deserve special mention. Both projects are potentially of great substantive importance to a wide range of colleges and universities. Yet, as the Foundation's Program Officer with direct responsibility for this area of activity (Ira Fuchs) has emphasized, what may be most noteworthy about them is that both involve collaborations of dozens of institutions that have each made a commitment to contribute significant funds over a multi-year period. This degree of commitment to a collaborative software development effort is, to our knowledge, unique. These institutions have become partners with the core developers and their involvement will help assure that the final products both meet the needs of higher education and are supportable in the long term.
"Chandler" is a next-generation open-source personal information manager. It is being developed by the Open Source Applications Foundation (OSAF), led by Mitch Kapor, and over the next two years OSAF will receive a total of $2.75 million to develop a version of this advanced information management software that is especially tailored to the needs of higher education. Almost half of these funds will come from the 25 members of the Common Solutions Group (CSG), a group of information technology specialists from major universities. Beginning last month, the CSG schools have formed a formal advisory group to help OSAF during the development. Also encouraging is the fact that since the announcement of this effort, several other institutions from outside the CSG have asked if they could participate and contribute $50,000 each to the effort. The CSG is now considering how to expand the Chandler support group to accommodate these requests.
The SAKAI collaboration represents an equally unusual, although arguably more complex, arrangement. This project's purpose is to develop an open source course management system that would serve a broad range of institutions of higher education in a much better organized, much more cost-effective way than they are being served now. The Foundation is contributing $2.4 million over two years, and four universities (Michigan, Indiana, MIT, and Stanford) are making an in-kind contribution, primarily staff time, totaling more than $4 million. These universities are joining forces to integrate and synchronize their enormous investments in educational software to create an integrated set of open source tools that would draw upon the "best-of-breed" from among existing open source course management systems and related tools. If successful, the result would be an economically sustainable approach to high quality open source learning software for higher education that would overcome the two main barriers that have consistently impeded such collaborative efforts: (1) unique local architectures, including heterogeneous software, software interoperability requirements between systems, and diverse user interfaces that hinder the portability of software among institutions; and (2) timing differences in institutional funding and mobilization that reduce synergy and result in fragmented, often incomplete offerings and weak interoperability.
Unlike Chandler, where the funds are being used to pay a third party (OSAF) to do the development with advice from university staff, SAKAI is an effort in which the core schools are doing the development themselves (with partial matching support from the Foundation). Each core institution has agreed to provide full-time technical staff that will be under the direction of the SAKAI Board. In addition, a SAKAI Education Partners Program (SEPP) has been formed to permit interested schools to participate in the discussion of the strategic directions for SAKAI. Each of the partner schools will commit to contribute $10,000/year for 3 years. In order to get the SEPP launched, the William and Flora Hewlett Foundation has agreed to provide seed funding of $300,000. So far, without any formal announcement of SEPP, 13 schools have agreed to join, and staff expect that we will attract more than 100 in the next 6 months.
In a related area, the Teaching and Technology program led by Saul Fisher has continued to support studies of the actual uses of instructional technology. One recent grant to the University of California at Berkeley is supporting a study that will compare the use of digital resources that are unrestricted and thus freely accessible with the use of resources for which charges are levied; by collecting data on usage where the access involves no subscription fees, this study may help us understand whether there are significant non-financial barriers to classroom usage of such resources. In another study, at the University of Maryland, investigators are researching the extent and impact of digital rights management software tools on online education—a subject of great interest to ARTstor and Ithaka.
The Foundation has continued to support its ongoing programs in theater, dance, and music. In 2003, the Program Officer responsible for the performing arts, Catherine Maciariello, also began a series of meetings with professionals in dance to investigate emerging needs in and across the field. The purpose was to collect data as well as anecdotal information about how dance artists and support organizations are adapting to the changing economic, political, and social climate. An important enhancement to the symphony orchestra program also occurred in 2003, when the Foundation made a grant of $1 million to the American Symphony Orchestra League to initiate a post-graduate fellowship program for talented early-career conductors.
In addition, the Foundation made several significant endowment grants, almost all with matching requirements. These grants were made primarily to help ensure the continuation of important programs initiated by institutions through earlier operating grants from the Foundation. For example, Jacob's Pillow Dance Festival received a grant of $750,000 for its endowment in honor of the organization's 70th anniversary. The Brooklyn Academy of Music received a grant of $1 million to support its opera and music theatre program. The American Symphony Orchestra League received an endowment grant of $1.45 million for its professional development services to the field—a grant intended specifically to help the League complete its matching grant from the Helen F. Whitaker Fund. The Alvin Ailey Dance Foundation received a grant of $1 million to support the company's New Works Fund.
Museums and Art Conservation
In the field of art conservation, the Foundation's Program Officer Angelica Z. Rudenstine, has continued to focus research and resources on strengthening the scientific component of conservation, an initiative begun in 2000, and one that remains an important priority for the Foundation. To this end, the Foundation has renewed its longstanding support of the Mellon Institute's Research Center on the Materials of the Artist and Conservator, at Carnegie Mellon University. In addition, it has supported the establishment of new positions for scientists at various levels. Notable among the latter is a senior position in science at the Freer Gallery of Art/Arthur M. Sackler Gallery within the Smithsonian; and, for the first time, the establishment of full-time professorial positions for scientists at two of the country's leading conservation training programs (Buffalo State College and the University of Delaware). These grants were made following extensive curriculum review that provided compelling evidence in each case of the need for stronger scientific representation among their faculties. Earlier support intended to strengthen the role of science in art conservation has been awarded in the past three years to the Art Institute of Chicago, Harvard University, the Los Angeles County Museum of Art, Metropolitan Museum of Art, the Museum of Fine Arts Boston, the National Gallery of Art, the Walters Art Museum, and the Worcester Art Museum.
In the museum field, the Foundation has continued to focus on enhancing curatorial scholarship. This has been accomplished through the strengthening of existing endowments for scholarly publications; through the creation of postdoctoral curatorial fellowships for young art historians at selected museums;  and through the establishment of securely endowed senior positions. In 2003, multi-million dollar matching grants for such purposes were made to the Morgan Library (to endow the Department of Medieval and Renaissance Manuscripts) and the Philadelphia Museum of Art (to endow the Senior Curatorship of Costume and Textiles and the Associate Curatorship of Arms and Armor).
Conservation and the Environment
The most important new initiative during 2003 in this program area was the active support of the "African Plants" cluster within Aluka/Ithaka (see earlier discussion). It is no exaggeration to say that without the direct involvement of the Foundation, and its highly respected Program Officer, William Robertson, this project would never have been undertaken. As a result of a long history of support for herbaria and botanical gardens, and for other projects intended to share digital information about plants, the Foundation was exceptionally well positioned to work with Aluka in developing a new resource of great promise. This is an excellent example of how extensions of earlier Foundation initiatives, that build on prior relationships with grantees, can benefit the scholarly community in entirely unanticipated ways. Synergies of this kind are invaluable.
In concluding this report, I would like to repeat my thanks to the Trustees and staff of the Foundation for their interest and unstinting support. This past year witnessed the addition to the Board of Trustees of Drew Gilpin Faust, Professor of History at Harvard and Director of the Radcliffe Institute, and Lawrence R. Ricciardi, former General Counsel of IBM who has a longstanding interest in books and the humanities. At the officer level, the Foundation was fortunate to enlist the services of Patricia L. Irvin as Vice President for Operations and Planning.
William G. Bowen
The Andrew W. Mellon Foundation
140 East 62nd Street
New York, NY 10021
April 2, 2004
 Ithaka takes its name from the Cavafy poem describing Odysseus's journey home. The poem notes that getting home entailed visits to many "harbors seen for the first time." Ithaka is exploring the emerging digital landscape in this same spirit. Return to text.
 This assertion is buttressed by an impressive body of evidence. Now under the leadership of Michael Spinella, who has succeeded Kevin Guthrie as its Executive Director, JSTOR continues to thrive. By the end of 2003, the JSTOR database of scholarly journals—which contains content back to inception of about 360 journals, over 13 million pages in all, and still growing—was being accessed by authorized users at about 1,900 institutions in more than 75 countries. Approximately 15 million articles were accessed during 2003 and just over 2 million articles were printed from the database. The usage rate continues to rise annually, with greater than 50 percent increases in each of the last two years. To date, JSTOR has not had to raise the annual access fee for any of its collections following their introduction, even though the content increases each year as a result of the moving wall feature (explained in last year's Annual Report). Even so, JSTOR's revenues have grown to the point where recurring revenues cover all of the current operating costs (though new content is still funded largely through outside sources). Return to text.
 As Neil Rudenstine, Chairman of ARTstor, has put it: "We all recognize that there is no substitute for direct engagement with original works of art or actual archival study. But the special opportunities presented by digital technologies constitute the most fundamental development in the potential for increased access and flexibility of use since the advent of photographic reproduction." Return to text.
 One very practical consideration is that these new entities, in order to become sustainable, need to raise revenues from the users of the products and services that they offer, as well as funds from a variety of philanthropic sources. These are "business" and "development" functions that a grantmaking foundation does not perform. Nor is a large grantmaking foundation well equipped to make—and execute—the myriad day-to-day decisions that any "start-up" has to make constantly. This is one of the lessons to be learned from the early days of JSTOR (see Roger Schonfeld, JSTOR: A History [Princeton: Princeton University Press, 2003], especially chapters 4 and 5). Return to text.
 The basic underlying question is why, and in what ways, not-for-profit organizations can be expected to do things that for-profit organizations or unaided market mechanisms will not do at all, or will not do very well. This is a huge subject all its own, which deserves more space than it can be given in this report. (I attempted to address some of these questions, at least in part, in the Romanes Lecture that I gave at Oxford in October, 2000. In that talk, I examined JSTOR's decisions to include scholarly content in a field such as classics that is of interest to only a relatively small number of scholars and to set fees so that small, impecunious institutions such as the Appalachian colleges and Historically Black Colleges and Universities (HBCUs) could afford to participate.) More generally, there is of course an important role for for-profits in many of the same fields in which not-for-profits are active, but the two sectors usually address quite different needs. In the field of higher education, for example, a for-profit such as the University of Phoenix serves a very different mission than a liberal arts college such as Bryn Mawr, a large research university such as the University of Michigan, or a center for advanced study such as the National Humanities Center. Return to text.
 Both ARTstor and Ithaka have excellent leadership at staff and Board levels. James Shulman is Executive Director of ARTstor and Neil Rudenstine chairs its Board. The other ARTstor Trustees are Anthony Appiah, Lewis Bernard, James Cuno, Anne d'Harnoncourt, Kevin Guthrie, Carol Mandel, Michele Tolela Myers, Peter Wendell, and me. Kevin Guthrie is President of Ithaka and its other Trustees are Paul Brest (Vice Chairman), Lynne Brindley, Charles Exley, Kenneth Frazier, Mamphela Rampele, Lawrence Ricciardi, Charles Vest, and me (Chairman). Andreas Dracopolous and Phillip Griffiths are Board Advisors. See the ARTstor and Ithaka Web sites for a full listing of Trustees, officers, and senior staff members. Return to text.
 Three town houses on 61st Street in New York, connected to the Foundation's space on 62nd Street via our back garden, are being renovated and will be occupied (we hope!) in the latter part of 2004 by ARTstor and Ithaka. Return to text.
 There is also a formal legal connection. The Affiliates have closely related missions, and in the terminology of the law, ARTstor and Ithaka are supporting organizations, and JSTOR is the supported organization. Although a different kind of creature altogether (a grantmaking foundation), Mellon sees its mission as encompassing the work of the Affiliates and expects to continue to support them. Return to text.
 These descriptions are based largely on information presented on the ARTstor Web site, which also contains a wealth of information about ARTstor: its origins, goals, and plans, as well as frequently asked questions. Return to text.
 Excerpted from the ARTstor Web site, which contains a much longer discussion. Return to text.
 Similar issues pertain to the work of JSTOR and Ithaka, and we are fortunate in that the attorneys working on these questions for the Affiliates and the Foundation comprise a very strong team: Gretchen A. Wagner, General Counsel for ARTstor; Nancy A. Kopans, General Counsel for JSTOR; Patricia L. Irvin, General Counsel for Ithaka; and Michele S. Warman and Jacqueline D. Ewenstein, General Counsel and Assistant General Counsel, respectively, for the Foundation. This is a good example of the value of collaboration among the Affiliates. Return to text.
 The Mellon International Dunhuang Archive has been by far the most expensive single project. To date, over $8 million has been spent in China and around the world capturing, documenting, and reuniting Dunhuang content. Because of the importance of Dunhuang as a center of Buddhist art and a major source of information about the history of Asia over many centuries, this will prove to be, we believe, an excellent investment on its own terms. But there are only so many such investments that the Foundation can make! Return to text.
 Since Ithaka, unlike JSTOR and ARTstor, was founded as a collaborative project of three foundations, which together are providing $10 million of start-up funding, this part of the Mellon Annual Report may also be read as a commentary on what the support of the Hewlett and Niarchos Foundations has made possible—without, however, implicating their Trustees or officers in the language I have chosen. Both Paul Brest of the Hewlett Foundation and Andreas Dracopolous of the Niarchos Foundation are active participants in the work of Ithaka, with Mr. Brest serving as Vice Chairman of the Board of Ithaka. Return to text.
 Examples include the Bryn Mawr Classical Review, the Bryn Mawr Medieval Review, and the Journal of Statistical Mechanics:Theory and Experiment. Return to text.
 Starting in 2001, the Foundation has supported several pilot initiatives, and it continues to support the LOCKSS project ("Lots of Copies Keeps Stuff Safe") based at Stanford as well as E-Archive. These two projects have seemed the most promising initiatives. LOCKSS is especially well suited to capturing and preserving the content of journals as they are presented on the Web, while the objective of E-Archive is to capture and preserve the source files. Initially, E-Archive existed as a JSTOR-sponsored project and so it has had a slightly longer life than its new parent, Ithaka, to which E-Archive has now been transferred. Return to text.
 Excerpt from a position paper prepared in January 2004. Return to text.
 Roger C. Schonfeld, Donald W. King, Ann Okerson, and Eileen Gifford Fenton, The Non-Subscription Side of Periodicals: Changes in Library Operations and Costs between Print and Electronic Format (Washington: Council on Library and Information Resources, forthcoming 2004). An abbreviated version has already been published as Roger C. Schonfeld, Donald W. King, Ann Okerson, and Eileen Gifford Fenton, "Library Periodicals Expenses: Comparison of Non-Subscription Costs of Print and Electronic Formats on a Life-Cycle Basis," D-Lib Magazine 10, no. 1 (January 2004), available at http://www.dlib.org/dlib/january04/schonfeld/01schonfeld.html. Return to text.
 Publishers have not traditionally been responsible for "preserving" the content they create, and libraries and scholars naturally are concerned that if the publisher runs into financial difficulty, sells or discontinues publication of a journal, or encounters other difficult circumstances, the publisher's priorities may not include a willingness to spend scarce money maintaining an electronic archive, which might include the costly process of migrating it to a new platform. Presumably the best way a publisher could provide a credible preservation guarantee would be by transferring the data to a trusted third party or by establishing an in-house endowment sufficient to cover the potential costs of maintaining the archive in perpetuity. JSTOR is such a trusted third party for digitized print publications and is building an "archive capital fund" as an endowment, but of course archiving is a central mission of JSTOR (a top priority), whereas it has not been a central mission of publishers. Return to text.
 A great deal of thought has been given to content selection strategy and to the criteria used in choosing a region (or regions) of the world in which to work, the substantive themes that define the "clusters" of content about and from the region, and, finally, the individual "collections" of materials that together comprise a cluster. There is not space in this report to describe these principles and criteria, but interested readers should feel free to consult the Aluka section of the Ithaka Web site, which contains Thomas Nygren's explanation of these principles and criteria. Return to text.
 The Aluka staff have recently learned of a second meaning of the word that also works well. In the Ovambo language groups of northern Namibia and southern Angola, "aluka" means "to return" or "to repatriate; in the context of the liberation struggles, it connotes the return of exiles." Return to text.
 There is a clear precedent in the work done in China, in connection with ARTstor's Dunhuang project, to train staff of the Dunhuang Research Academy and also Chinese staff based in Beijing to digitize and stitch images. Return to text.
 In commenting on an early draft of this report, Tom Nygren, Executive Director of Aluka, emphasized: "the huge benefit we get from the shared services aspect of being an incubated entity. Although seemingly mundane, I can assure you that the human resources, financial services, IT support, legal advice, etc., are incredibly important to a start-up entity. I can't imagine how much time I would have to spend on those aspects were we not being incubated within Ithaka, which instead I can spend on the core mission of Aluka...I want to be sure that the value of this aspect of Ithaka—including help with business plans and our thinking about sustainability—comes through loud and clear." Return to text.
 If I may be allowed a personal reminiscence, one of the ablest black undergraduates at Princeton in the 1970s was Randall Kennedy, who, among other things, was ferociously independent intellectually and loved to argue about anything and everything—a heaven-sent candidate for academia, I thought, and I tried hard to convince Randy that his highest and best calling was as an academic. Others also noticed Randy's capabilities, and he won a Rhodes Scholarship. He then elected (my protests notwithstanding!) to go to Yale Law School and subsequently served as law clerk to Skelly Wright (U.S. Court of Appeals) and to Thurgood Marshall. Not that many years later, Randy accepted a faculty appointment at the Harvard Law School. Today he enjoys an outstanding reputation as a scholar, teacher, and academic leader. I still have the note Randy sent me when his appointment to the Harvard faculty was announced. It read simply "You win." Yes, Randy did become a highly accomplished academic, but his path to that destination was not a direct or easy one. Return to text.
 1993 Annual Report, p 22. Mr. Drewry's essay in this report is an excellent source of information concerning the early years of MMUF and the rationale for investing the program with the special combination of features that it still possesses today—along with additional attributes intended to facilitate the completion of the PhD and to encourage early success within the professoriate. Return to text.
 Rosalind R. Bruno and Andrea Curry, "School Enrollment: Social and Economic Characteristics of Students," U.S. Bureau of the Census, Report P20-487, September 1996, Table G, p. xiv. Return to text.
 Data from Chronicle of Higher Education Almanac, August 30, 2002, Vol. XLIX, No. 1. Return to text.
 1993 Annual Report, p.24. Return to text.
 The two fields with the largest numbers of MMUF faculty are English/literature and history. But there are also significant numbers in fields such as anthropology, physics/chemistry, mathematics, music, and art history. The 40 colleges and universities that have recruited former MMUF students to their faculties constitute a veritable "Who's Who" of American higher education. The roster includes Bryn Mawr, Harvard, Queens, Clark-Atlanta, and the Universities of Chicago, Michigan, and New Mexico, to name just a few to illustrate the range of institutions served. Return to text.
 It is the colleges and universities participating in MMUF that were primarily at risk of being found to have violated Title VI of the Civil Rights Act of 1964, not the Foundation itself (since the Foundation receives no federal funds). But of course the Foundation did not want to create problems for its participating colleges and universities through the structure of the MMUF program, and without the active participation of the individual colleges and universities, MMUF could not continue to operate as effectively as it has in the past. Return to text.
 For a short biography of Dr. Mays, see Orville Vernon Burton, "Mays, Benjamin Elijah," American National Biography Online, Feb. 2000. Dr. Mays seems an excellent model for MMUF in part because it is said that "at an early age [he] developed an 'insatiable desire' for education." Overcoming obstacles of every kind, he earned an undergraduate degree from Bates College and, at age 40, a PhD from the University of Chicago. He was a great teacher, and one student inspired by Mays was Martin Luther King, Jr. At King's funeral, Dr. Mays gave the eulogy. Dr. Mays was President of Morehouse College for 27 years. Return to text.
 Anthony Marx, recently elected President of Amherst College, comes immediately to mind. I first met Tony when, as a graduate student at Princeton, he was part of a multi-racial group of students determined to convince the president to change the University's investment policies in order to hasten the end of Apartheid in South Africa. I was unpersuaded that divestment was the appropriate strategy, but I was greatly impressed by this articulate and committed graduate student. Tony has demonstrated his commitment to the goals of MMUF, and he certainly will continue to play a leading role in working to reduce racial disparities. Return to text.
 Myrdal, Gunnar. An American Dilemma: The Negro Problem and Modern Democracy. New York and London: Harper & Brothers, 1944. Return to text.
 In addition to the Distinguished Achievement Awards which honor scholars who have made significant contributions to the humanities and enlarge opportunities for scholarship at the institutions with which they are affiliated, the Foundation now supports an array of fellowships aimed at meeting the differing needs of scholars at various stages of the scholarly life cycle. These include fellowships for junior faculty members (ACLS/Andrew W. Mellon Fellowships for Junior Faculty and Charles A. Ryskamp Research Fellowships); support for scholars at mid-career (Sabbatical Fellowships overseen by the American Philosophical Society; Frederick Burkhardt Residential Fellowships for newly tenured faculty members; New Directions Fellowships for scholars seeking training in fields other than those in which they earned their degrees); and Emeritus Fellowships for senior scholars. The Foundation also supports postdoctoral fellowships, which assist institutions seeking to enrich their curricula while not enlarging the size of their permanent faculty and which simultaneously provide promising young scholars with teaching experience and time to develop their research programs. Previous Annual Reports provide detailed descriptions of these programs. Return to text.
 See Harriet Zuckerman and Joseph S. Meisel, "The Foundation's Programs for Research Universities and Humanistic Scholarship," essay in the Annual Report for 2001, pp. 37–56. Return to text.
 Foundation staff have also continued to do research themselves, and in 2003 Sarah Levin and I published Reclaiming the Game: College Sports and Educational Values, a study of the effects of college sports on selective colleges and universities that do not offer athletic scholarships. The study has engendered wide debate, and efforts are now underway to see if there is sufficient interest in the core ideas proposed in the study to justify continuing efforts to institute fundamental reforms. Return to text.
 The name SAKAI comes from the name of the chef (Hiroyuki Sakai) who appears on a TV show on the Food Network, the "Iron Chef." The project name is a play on the acronym for the University of Michigan project that SAKAI is built upon, CHEF, which stands for "CompreHensive collaborativE Framework." Return to text.
 The Foundation's program of postdoctoral fellowships for art historians, initiated in 1995 and currently active in six leading museums, offers promising scholars the opportunity to gain experience in the museum field, while strengthening the scholarly capacity of important institutions which are not in a position to expand the size of their permanent curatorial staffs. Return to text.