One World: College Libraries in a Digitized Age

Remarks at 10th Anniversary Celebration, The Five Colleges of Ohio
William G. Bowen 
November 10, 2005


It is a very special privilege for me to be back "home" in Granville, with so many old friends—and not just from Denison but from other members of the Five Colleges too. Memories flood back, but worry not: I have no plan to inflict on you stories of tennis matches at Oberlin (complete with jelly beans spread behind baselines), life as co-head resident of a freshman dorm (with Dave Bayley), hours spent in Doane Library working on a mammoth senior thesis (under the tutelage of Leland Gordon), or visits to Denison as a Trustee during contentious Vietnam days. Rather, I want to reflect on more recent events, centered around the effects of new digital technologies on how students and scholars do their work, and on the implications of these technologies for liberal arts colleges going forward. It is particularly fitting that on the occasion of this celebration my watchword is "collaboration."

JSTOR: And Now Portico

My story has to begin with JSTOR, the by-now huge electronic archive of journal literature which was born, as an idea, in a Denison board meeting in late 1993 and then developed more fully—over a cognac—in the study of President Michele Myers' house that evening. (I should note that Michele continues to be a major "player" in the continuing evolution of JSTOR, serving as a board member of both JSTOR and ARTstor.) At that 1993 board meeting, President Myers asked the Trustees to consider the need for more stack space for the library. It occurred to me that since essentially all libraries were facing similar pressures to add space, and since many, many libraries held huge back runs of the same core scholarly journals, it might be possible to serve the needs of everyone by creating a centralized electronic resource. The need was not only to economize on capital spending and on space, but also to improve access—to enable readers to find materials that were often inaccessible and next-to-impossible to locate and use. Our mantra, from the beginning, was "better and cheaper." JSTOR was to become a special kind of library collection: one that was never closed, was accessible from a faculty office or a student dorm as well as from the library, and that contained content that was easily searchable and always available (never "checked out," never defaced). 

The founding and subsequent development of JSTOR has been described in great detail, "warts and all," by Roger Schonfeld in his excellent history, and so I need not retell that story today, fascinating as it continues to be (at least to me). Let me fast forward. Having started out as a pilot project that involved ten journals in two fields of study, and counted less than 200 charter library participants ("early adopters"), JSTOR today serves over 2,500 libraries in 95 countries. At the start of this academic year, there were 528 journals online (over 18 million pages of content), and the year-to-year growth in usage was over 40%. The driving force behind these astonishing figures has been the unflagging support of the library community and the enthusiasm of users, worldwide, for this resource—which has changed fundamentally the way scholars and students teach and do research. JSTOR can justifiably claim to have breathed new life into older journal content and to have permitted users to connect and trace ideas in ways that were difficult if not impossible before. 

The depth of commitment of academics to JSTOR can be illustrated by a recent anecdote. A young Chinese scholar who had just finished a PhD in political science at a US university was contemplating returning to China because she thought that she could make more of a contribution there than by teaching in the US. But she made clear to the Chinese that she would go if, and only if, she would have access to JSTOR. Otherwise, she asked, "how could I do my work?" 

Thanks to this incredible growth in usage, and to the presence of substantial economies of scale, JSTOR has become financially self-sustaining. This is a particularly gratifying achievement in that, from the beginning, JSTOR has adopted a value-based pricing model whereby small colleges such as those in the Ohio Five, and institutions in poor countries, pay only a fraction of the fees charged to larger institutions in the US and Western Europe. Moreover, JSTOR has never raised the fee for any given collection, and the charge per page of content has therefore declined steadily because of the annual addition of content via JSTOR's moving wall. The average cost per article actually "used" (for viewing or printing) has fallen from about $1.40 in 1997 to about 20 cents today. The value proposition seems clear—even without taking account of substantial savings in capital costs and operating costs, and without factoring in the gains in convenience and the time savings for users who enjoy instant access to journal content at any hour and from almost any location. 

In the continuing pursuit of "better and cheaper," the Mellon Foundation is now supporting a new initiative called "Portico" that is in some respects an outgrowth of JSTOR. Portico is an electronic archiving service with a mission that can be stated simply: To preserve the journal literature published in electronic form and to ensure that that these materials remain accessible, under carefully defined conditions, to future scholars, researchers, and students. So easy to state, yet so tricky to accomplish! Portico has been a particularly challenging project from its inception. One key requirement is sophisticated technology that allows Portico to receive deposits of "source files" of electronic journals and then convert the proprietary files to a normalized format that will serve archival purposes and can be migrated as necessary, over time. This is a devilishly difficult task which, fortunately, talented staff at Portico have accomplished. 

An even more challenging task has been to design an operating-business model that will encourage all of the relevant parties, publishers and libraries, to join the party—cheerfully and with enthusiasm, we hope. This model has undergone various iterations, and it is now at a stage where it seems to meet the basic requirements of publishers and libraries alike. The current model emphasizes long-term preservation over short-term access. It is more like a long-term insurance policy than anything else. Unlike JSTOR, the archive will be highly inclusive and open to all peer-reviewed journals. Access to content in the Portico archive will be far more restricted than access to JSTOR: access will be provided only in the case of such established "trigger events" as when a publisher goes out of business, ceases to publish a title, removes back issues from its site, or otherwise fails to provide consistent access. Publishers may also choose to rely on Portico to meet perpetual access obligations, and there will be a standard "auditing" procedure that would allow designated individuals to verify the ongoing integrity of files. 

The success of Portico depends utterly on the willingness of publishers to deposit electronic content in it. The decision to emphasize preservation over access has been reassuring to publishers, and recent conversations lead us to believe that large numbers will in fact see the advantages to them of participating. At the Washington meeting of the ARL, I was very pleased to report a late-breaking piece of news that is highly promising. The leadership of Elsevier has told us that it is fully on board in support of Portico. Elsevier agrees that archiving is a critically important issue for the academic community and has committed not only to contribute its content for preservation through Portico, but also to do whatever else it can to help develop momentum for this approach. The leadership of Elsevier recognizes, and indeed has emphasized in our conversations with them, that the benefits of archiving are best realized if many publishers and libraries join together in supporting a common solution. Another important publisher, the American Mathematical Society, has also committed to participate and will encourage other publishers to do so. Of course libraries are in a pivotal position to encourage widespread publisher participation by, in effect, demanding archival deposit by publishers as a condition of licensing electronic content. 
The sustainability of the Portico archive will depend on the willingness of all elements of the "system" to help cover its costs. Charitable foundations such as Mellon and organizations such as JSTOR have already invested significantly in developing Portico's infrastructure, and it was very encouraging to see the recent announcement of the award to Portico of $3 million by the Library of Congress. Publishers can be expected to make annual financial contributions as well as deposit their content, and libraries should be expected to support such an enterprise by making modest annual payments. Fortunately, the economies of scale are so pronounced that if the community as a whole steps forward, cost per participant (scaled to size) should be manageable.

The time to act is now, and I liked very much the title of the recent statement describing the consensus reached on this subject at a meeting of academic librarians (from small colleges as well as large universities), university administrators, and others: "Urgent Action Needed to Preserve Scholarly Electronic Journals." In its first paragraph, this statement notes:

...As the creation and use of digital information accelerate, responsibility for preservation is diffuse, and the responsible parties—scholars, university and college administrators, research and academic libraries, and publishers—have been slow to identify and invest in the necessary infrastructure to ensure that the published scholarly record represented in electronic formats remains intact over the long run. Inaction puts the digital portion of the scholarly record—and the ability to use it in conjunction with other information that is necessary to advance knowledge—increasingly at risk... For electronic journals, the academy has as yet no functional equivalent in long-term maintenance and control over the scholarly record that 'owning a copy' provided for printed journals.

The case for solving this problem is straightforward, as well as time urgent. We have here, within our grasp, a classic "win-win" opportunity. For the end users, scholars and students, the advantages of accelerating the movement to electronic formats, especially to facilitate search and retrieval, are evident. For publishers, the existence of an established archive can reduce or eliminate internal archiving costs, meet the demand of its "customers" for a trusted third-party archive, satisfy demands by libraries for perpetual access without a negative impact on operating revenues, and facilitate the inevitable movement to electronic-only publishing. For libraries, the existence of Portico creates a mechanism for acting upon their traditional preservation mandate. Moreover, having in place a well-functioning third-party archive, which they can trust to be around for the long term, is essential to achieving a smooth transition to electronic-only journal publishing—with all of the system-wide savings and benefits associated with such a shift. In the absence of such an archive, libraries will be reluctant to give up the processing and storage of print copies of journals, even though moving to an electronic-only mode of operation offers enormous potential savings. Portico reduces system-wide preservation costs by providing a cooperatively designed preservation infrastructure which eliminates the need for each library to archive electronically published literature independently. 

These potential savings redound to the benefit of the entire college or university (for example, by reducing dramatically the need, over time, for new capital investments), and thereby offer at least a partial response to the growing public demand for greater efforts within higher education to achieve efficiencies—a demand that we ignore at our peril. For this reason, the cost of putting in place a service like Portico and sustaining it should be understood to be a claim on institution-wide pools of resources. It should not be thought of as merely a charge to an already strained acquisitions budget within the library. Savings in funds that otherwise would have had to be spent to "handle" and archive paper journals should be part of the Portico financial equation. The potential productivity gains—which can be translated in this context into dollar savings—are evident and evidently substantial.

And Now Images and an Image Management System

Of course one idea often leads to another, and the success of digitizing the text content of journals led us to wonder if there were similar advantages to be gained by creating a central repository of high quality images of art—intended, again, to serve a broad range of colleges and universities. "ARTstor" was created to seize this opportunity. Under the leadership of James Shulman and Neil Rudenstine, it is, as many of you know, busily engaged in building a vast digital repository of images of works of art and related scholarly material. The charter ARTstor library will include 500,000 images, associated cataloging information, and other data. The content ranges from a broad and deep university slide library at the University of California at San Diego (UCSD) to remarkable specialized collections such as The Illustrated Bartsch; the Mellon International Dunhuang Archive (which is a resource essential for the study of Buddhist art and the history of Asia that consists of cave paintings and manuscripts originally located at Dunhuang, on the Silk Road in China, and now dispersed all over the world); an initial part of the Gernsheim corpus of 182,000 old master drawings from photographs taken over a 70 year period; and extraordinary new images of the recently restored sculpture of Ghiberti's Gates of Paradise. 

Superficial similarities notwithstanding, the "resource" that ARTstor offers is very different from journal literature. ARTstor is comprised of primary materials for users of visual images. Unlike JSTOR, ARTstor deals not with the end products of scholarship, but with the visual raw materials of scholarship and teaching. These raw materials are the common "stuff of the humanities and related social sciences 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. There are intellectual property issues and sensitivities aplenty, as I'm sure all of you understand. In order to manage the risk (a risk which ARTstor takes off of the institutions' plates by indemnifying users for authorized uses), and to facilitate active use of the images, ARTstor delivers its content via an active online and offline software environment that is expensive to develop, enhance, and maintain. ARTstor software functions in many ways as an instructional technology. 

The software in which ARTstor images are imbedded is seen by many institutions as a wonderful resource—as a way they can study and present digital images of many kinds, including institutional holdings and the personal collections of faculty members, without having to incur the substantial ongoing costs of building and maintaining their own software for presenting and manipulating images. ARTstor is experimenting with providing a "hosting" service that will allow institutions that are so inclined to utilize its software to store and display a wide variety of images owned by the institution itself (including, for example, a collection of images of insects at Denison). Because of this possibility, and because images are of great pedagogic value in many fields outside art history—such as area studies, classics, history, anthropology, political science, literature, and even entomology—it is essential that ARTstor be understood to be a campus-wide resource. Obviously, the library has a key role to play—the key role, I would say—in both emphasizing this point and helping the campus community learn how to make optimal use of ARTstor images and its imaging software. 

On some campuses, the fact that the content "within" ARTstor ranges from the externally licensed ARTstor images to local hosted content to individual images added by instructors highlights one of the profound changes resulting from the application of digital technologies. Collections are now conveniently (some might say infinitely) replicable and customizable, instead of being held exclusively by a single steward. This development shifts, in a sense, the question of what "collection acquisition" means and what the role of the library is in this context. By gathering up and sharing across the campus—in fact, across many campuses—what one of our colleagues calls "random digital acts of progress," a platform like ARTstor helps us understand and redefine the role played by the library, which can become, as it were, a highly cost-effective nexus of inflows and outflows of local and remote content. The potential for productivity gains and cost-savings is, once again, very considerable. These kinds of advances would be unthinkable in the absence of what are really powerful collaborations that are organized around centralized resources.

And Now Primary Source Materials from All Over the World

If JSTOR is about text and ARTstor is about images, a third entity created even more recently by the Foundation has yet another focus: this enterprise is called Aluka (after the Zulu verb "to weave"), and its general purpose is to partner with key libraries, archives, scholars, and museums "to build and support a sustainable, online digital library of scholarly resources from and about the developing world, beginning in Africa, for research and teaching worldwide." Its content is not to be limited to one medium and will include not just journal literature and images, but also manuscripts, other primary source documents, oral histories, and perhaps [later on] music and video. The decision to focus first on Africa is due to the need to know more about a continent of great intrinsic importance, to an appreciation of the need for scholars in African countries to have access to their own materials (which otherwise may be lost forever), and to the Foundation's considerable experience working in South Africa.

The first Aluka content cluster to which I want to draw your attention is the African Plants initiative. Built through the leadership of Bill Robertson, a Mellon Foundation program officer with long-standing ties to botanists the world over, the objective is to present over the internet, in an easily searchable format, high quality images of plant type-specimens of every known African species, along with information about the plants and their uses. Already, through an absolutely unprecedented international collaboration (that word again), 47 herbaria from 26 countries around the world, including the Royal Kew Botanical Gardens, the South African Biodiversity Institute, and the National Museum of Kenya, are contributing content and participating in the development of the database. An initial demonstration of the resource was a highlight of the International Botanical Congress in Vienna this summer. The reception was so positive that thought is already being given to including plants from other regions of the world, perhaps moving next to Latin America. 

A second collection that is being developed by Aluka is called "Struggles for Freedom in Southern Africa," and it illustrates the importance of broad international collaborations in a field very different from botany. A guiding principle is that Aluka must bring to life rich resources from Africa (such as the personal papers of the first president of Mozambique and of other leaders of liberation movements)—and then link digitized copies of these primary source materials to valuable collections of related content that are in other parts of the world, such as the Herskovits Library of African Studies at Northwestern University and Rhodes House at Oxford. Taking advantage of the knowledge of a large network of scholars from many countries, Aluka will aggregate (weave together), content in Africa, much of it in precarious condition, and related materials elsewhere so that students and scholars can make necessary intellectual connections that otherwise would be difficult if not impossible even to imagine. Scholars in Africa want very much to contribute their ideas and content to the building of a world-wide resource—they do not want to be just "takers" of what scholars outside Africa will give to them, and they do not want their resources to be used solely by others elsewhere.

There is a third initial Aluka cluster, called African Cultural Heritage Sites, which I do not have time to describe. Interested parties can consult the Ithaka web site to find information about this fascinating use of laser technology to "map" such important sites as Kilwa Kisiwani in Tanzania, Axum and Lalibela in Ethiopa, Elmina Castle in Ghana, the mosques of Djenne and Timbuktu in Mali, and to combine these maps with historical commentary, including journal literature, pertaining to the sites.

Aggregation, Collaboration, and Scale: Seen from on High

I have imposed all of this detail on you for two reasons. First, I think the initiatives are interesting, in and of themselves, and I wanted to introduce you to some of the newer ones. Second, I wanted to provide enough of a sense of some of the things actually being done, right now, that when we next consider broader lessons and implications, it will be against a background of real projects, not just vague hopes. Looking down on this family of Mellon "Affiliates" from "on high," we see a connecting thread, or theme, running through all of them. It is the power of aggregation, collaboration, and scale—all made possible by advances in digital technology. No one scholar or even a single educational institution could conceivably mount such projects, and it is also hard to imagine a single consortium mobilizing the resources and establishing the decision-making machinery needed to "execute" efficiently in the context of what is close to a venture capital model. It is clearer and clearer to me that many of these initiatives and others yet to come (I will mention one more in just a moment) have to occur on a system-wide basis. They generally require new organizational arrangements and, in many instances, the catalytic contribution that a trusted third party can provide. 

The advantages of these broad-based digitally-driven collaborations are apparent. Quite apart from the economies that they permit (building fewer feet of stack space, moving to electronic-only journals, and rationalizing slide library holdings and operations), they allow faculty and students to study content in new ways. The Dunhuang collection, for example, allows scholars of the history of Asia to view images of bodhisattvas painted on cave walls on the edge of Gobi desert to be compared, at incredibly high resolution, with images of the same bodhisattva on a rare silk painting at the Guimet in Paris. The African Plants Initiative is another example, and Aluka's Struggles for Freedom cluster will "connect" fugitive anti-apartheid literature in South Africa to rare materials in Oxford and Chicago. JSTOR itself (the "parent," if you will) allows a user interested in a topic such as affirmative action to search for, retrieve, and view peer-reviewed articles in many disciplines. As compared with Google, all of these resources have the advantage of containing carefully curated content, which is far richer than an amalgam of unsorted materials. 

But let us not allow Pollyanna, wherever she may be lurking, to obscure how hard all of this can be for colleges such as the Ohio Five we are celebrating today. In charting new ways of accessing, managing, and storing information, colleges and universities are going to have to learn to be better than many are today at making trade-offs and in recognizing that "the best may be the enemy of the good." Advances in information technology and the potential new collaborations pose organizational challenges for these institutions, which need to be more willing to exorcise old demons and more capable of insisting that an institution-wide perspective be brought to bear on resource allocation decisions. It can be so tempting to just keep doing things "locally," even if that is often more expensive (certainly over the long run) and less likely to lead to real breakthroughs in scholarly technique. 

My own view is that liberal arts colleges have no real choice but to seek their place at the table when system-wide approaches to common problems are being discussed. Otherwise, they risk being left behind in the competition to find more effective ways to teach and to learn. Prospective students are going to want to know that by attending a small, selective college, which offers so many advantages in terms of opportunities for personal growth as well as intellectual development, they are not sacrificing other opportunities to benefit from constant advances in information technology. Similarly, the most able faculty members are going to want to be confident that they will have access to the same general kinds of resources as those who teach at larger places and as they themselves became used to in their graduate schools. 

One other new organizational entity has been created by the Foundation precisely to stimulate the sharing of resources and ideas among liberal arts colleges. It is called, as you all know, "NITLE" (National Institute for Technology and Liberal Education), and one of its key purposes is to gain for liberal arts colleges access to new developments in technology and the benefits of economies of scale that it might otherwise be very hard for them to realize. Jo Ellen Parker recently of GLCA, is the Executive Director of NITLE, and having worked with the faculty and staffs of the Ohio Five institutions and other colleges, she brings much good experience with her. As we all recognize, liberal arts colleges face great challenges in adopting instructional technologies and even greater risks if they fail to do so. In many cases, time for faculty and professional development is scarce. Small academic departments and teaching-intensive schedules make it difficult for faculty to seize opportunities to develop skills or redesign courses to incorporate new technologies. At the same time, today's students and newly-trained faculty come to campuses like those of the Ohio Five expecting to learn and work in technologically sophisticated ways. Institutions that fail to meet these expectations risk seeming antiquated, irrelevant, lacking in rigor, and out of touch. 

But if training is an important part of the NITLE mission, so too is the encouragement of a culture of innovation and exploration of major new concepts. I mention just one. A rapidly growing movement in higher education is the development of Open Source Software, which promises to provide to educational institutions more flexible (and probably less expensive) approaches to a myriad of needs/opportunities than are available through for-profit proprietary mechanisms. Open source applications range from software that can be used to create customized portals to ways of managing course content to ambitious efforts to redo the management of business systems. But can small colleges participate effectively in the Open Source movement, since some considerable local expertise may be required? Well, the hope is that by facilitating inter-institutional collaboration, organizations such as NITLE may be able to assist in this process. The potential is enormous.

Let me end by saying that this is such an exciting area for all of us because the landscape is ever changing. You certainly see the changes, and the pace of change, from your own angles of vision. For my part, I can say only that I have learned so much, though never enough. If I had known ten years ago what I know today, I might have chosen a different life for myself—but I don't think so. (You may have pondered the same question.) All the complications and aggravations notwithstanding, these projects are tremendously exhilarating; they allow us to do what would have been unimaginable only a few years ago.

Risk-taking is definitely required, and this includes both a willingness to "learn by doing" and a willingness to trust others. Libraries need to be right at the center of these campus-wide developments. I recognize that having grown up in libraries, and having "lived," as it were, in a number of them (and I think especially of the Doane library at Denison), I may not be entirely objective. Nonetheless, let me state my categorical view that a new age for libraries is dawning, and that those that enjoy the best leadership will have an unparalleled opportunity to work across their institutions, and beyond them, in developing new modes of research, teaching and learning. It is a great and exciting time to be a librarian, and I salute all of you who have worked so hard to get us where we are today—and who are now poised to help us cross new boundaries.