As the image library turns one year old, it is finding an expanding audience across disciplines The scenario is a familiar one: Professor Smith teaches in the history department on campus (but it could be anthropology, classics, literature and languages, philosophy, religious studies, or sociology). She is winding her way through the unfamiliar precincts of the art history department on a mission. She wants to start using images in her teaching, having concluded that by requiring her students to take into account art, architecture, and visual and material culture she can truly bring her subject to life for a new generation. Her own department does not offer a slide library, let alone curatorial assistance for cataloging and managing teaching images. So she is in search of the elusive art history slide library she has heard so much about from her art historian colleagues. Professor Smith finally reaches the slide library - only to discover that using the collection is not going to be easy. The visual resources curator, who clearly would like to be helpful, has no support staff and is more than fully occupied with the needs of his primary clientele in the art history department, who now require him to manage both slides and digital images. There is no other entity on campus that offers comparable collections and services shaped around the work of teachers. This scenario is playing itself out on campuses large and small, from Berkeley to Berlin. As teachers, scholars, and students in disciplines well beyond the arts attempt new approaches to teaching and learning - approaches that require the integration of visual materials into their curricula and research - they are encountering organizational and financial barriers. Traditional approaches to the development, management, and delivery of institutional collections and services cannot support the evolving and expanding need for images. Visual resources departments at literally hundreds of institutions are striving to take advantage of the promise of digital technologies to solve this organizational conundrum. In the process, they often discover that not everything desirable is doable. This discovery takes many forms, as challenges emerge on all sides. Among the foremost of these are:

• daunting obstacles on the copyright and intellectual property front
• serious handicaps on the intellectual access, cataloging, and metadata fronts
• staggering infrastructure costs on the technology front
• profound challenges when it comes to developing scalable and sustainable user services and support.

Toward a cultural community

In the face of these roadblocks, many institutions of higher education have begun to reexamine and rethink traditional approaches to building image collections and to test new ways of bolstering an expanding range of image users. New partnerships among libraries, instructional technology units, and visual resources collections are being explored; conventional organizational boundaries are beginning to blur and shift. A role is being envisioned for a trusted third party whose place in this evolving ecosystem would be both to foster cross-institutional collaboration and help alleviate organizational redundancies in order to contain systemwide costs. After all, innumerable campuses are independently trying to digitize and catalog the same visual materials and then investing - again, redundantly - in managing these newly created digital assets, developing or licensing often expensive software in order to facilitate their use. Eventually they are discovering that the cost of all this is prohibitive for individual institutions and, in the aggregate, wasteful of the larger community's resources. All of this argues for the existence of a central role for an independent partner in this expensive enterprise. At the same time, there seems to be a key spot for such a trusted third party in helping to bring together, on an international scale, archivists, artists, librarians, museum professionals, photographers, publishers, and visual resource professionals, all of whom have a role to play in building the library of digital images we all need in the cultural community - and in helping to mediate the interests and concerns of this wide range of stakeholders.

A trusted third party

The Andrew W. Mellon Foundation created ARTstor precisely to be this kind of "trusted third party." This new member of the arts and higher education community would, from the outset, work with all of these constituencies on an international stage. ARTstor's mission is, in a nutshell, to help higher education institutions respond to the complex and expensive set of challenges described above via the ficticious Professor Smith. How shall we build the kind of digital image library required by scholars, teachers, and students in the 21st century? How can the key aspects of this emerging digital library - its collections and services, its provision for intellectual access, its software tools, its technical infrastructure, its licensing and legal framework - be made scalable, sustainable...and affordable? How can we, as a community, address the challenges that have, so far, prevented these clearly desirable aspirations from becoming attainable goals? ARTstor (www.artstor.org) went live in July 2004, and we're still answering some of these questions. We hope to work with archives, libraries, museums, teachers, and scholars to help higher education institutions achieve together something that has proved elusive so far. The roles ARTstor hopes to play are several:

• As a nonprofit organization with roots in both higher education and the museum community, ARTstor strives to bring the international areas of archives, libraries, and museums together around a set of shared values and common goals focused on teaching and learning. Essential to this effort is a commitment to identifying, understanding, and balancing the concerns and interests of content and rights owners with those ofend users and their institutional representatives.
• As an expanding digital library offering (even at this early date) hundreds of thousands of digital images and related data, ARTstor seeks to provide scholars, teachers, and students with the kinds of collections and software tools they need to make the pivotal transition from slides to digital images. This effort again requires close collaboration with a wide array of content owners both in the United States and abroad, as well as creative approaches to managing heterogeneous cataloging data.
• As an online resource available only to nonprofits, ARTstor seeks to create a secure, trustworthy space on the Internet for the educational, noncommercial use of digital images. This space is defined by a licensing framework that embraces - and seeks to accommodate the concerns and interests of - content owners, participating institutions, and end users.

Each of these key goals poses significant challenges, ones that can only be addressed collaboratively.

Balancing stakeholder interests

One of ARTstor's greatest challenges is to balance the concerns, interests, and needs of content owners (archives, libraries, and, especially, museums, both here and abroad) and those of potential users of digital content, predominantly in higher education. Most museums, as well as many other institutions that collect and preserve cultural objects, depend on revenues generated through commercial reproduction of their collections, whether on T-shirts and coffee mugs or in scholarly articles and books. Conversely, the purely noncommercial, educational use of images of museum objects in teaching and learning has never been a source of significant revenue; the majority of museums do not even actively make their collections available for these kinds of efforts. Only by understanding and underscoring this important distinction can the ground be cleared for developing a shared, mission-driven commitment to strengthening the educational use of digital images. ARTstor hopes to make a real contribution here. In much the same vein, it is seeking to arrive at international understandings with artists' rights groups, with the goal of defining a similar "safe haven" for the educational use of digital versions of works of 20th- and 21st-century art still under copyright in varying jurisdictions worldwide. Balancing the concerns of content owners and end users often requires compromise. For the content owner, it means remaining mindful of and actively endorsing the noncommercial, educational, and scholarly use of digital collections, even in times of dwindling resources and corresponding pressure to focus on the bottom line. For the end user, as well as for librarians representing the interests of end users, it means - for now at least - accepting that ARTstor's ability to "interoperate" with local resources, software, and systems will be circumscribed. In an attempt at nonpartisanship, ARTstor is placing limits on the individual's ability to download its images for use in other software environments and on the participating institution's ability to integrate these images into local software and systems. Many content owners and providers have gradually come to trust ARTstor to enable and manage the judicious educational use of their collections, but this trust does not - at this early stage - seem to be transferable at large. To enable the active use of ARTstor "content" outside its software environment would jeopardize its ability to develop the kinds of collections it has been able to offer. That said, we are pursuing a variety of promising approaches to interoperability. These include enabling "metasearching" across our collections in tandem with other online resources and the ability to use other resources within the ARTstor software environment. Users may already easily integrate local images into ARTstor's offline presentation tools, and shortly they will be able to do the same online. We are also conducting an "institutional hosting" pilot project involving a dozen facilities. We hope to offer a service through which we will host already-digitized image collections on ARTstor servers, so institutions can use their own digital collections alongside ours in a seamless way that also promises to relieve the burden on libraries to offer software tools for teaching and learning with images. The effort to understand and equalize the sometimes conflicting interests of a range of stakeholders is perhaps most challenging in the international arena; the international intellectual property and copyright landscape is singularly complicated and can vary significantly from country to country. The educational "fair use" exceptions to copyright that exist in U.S. copyright law do not exist in most countries. Similarly, cultural heritage and patrimony attitudes and laws vary markedly around the world. It is no small task to address these complexities on an international scale, and here, too, the effort is likely to require compromises on all sides.

Campuswide collections

ARTstor also faces considerable obstacles on the collection development front. We are building a digital library that already holds approximately 300,000 images. The ARTstor "Charter Collection" will, by the end of 2006, include half a million images. But numbers tell only part of the story. Our key collection development goal is to build the kind of "campuswide" image library Professor Smith requires. There is a core body of visual cultural heritage materials that is of profound interest all across the arts, humanities, and beyond. ARTstor does not seek to build, say, an "art history collection," a "classics collection," and a "social sciences collection." Rather, we seek out coherent and cohesive image archives that will have wide relevance and value for these areas and beyond. In this respect, ARTstor's name may be misleading. We are committed to promoting the image needs of art historians, architectural historians, and practitioners of art and architecture, but ARTstor is equally committed to identifying the intersections and points of convergence that define the academic and scholarly landscape and to building image collections that have profound cross-disciplinary value. Achieving this goal is as challenging as it is compelling. A model of this kind of collection is "The Illustrated Bartsch." Based on a 100-volume reference work of the same name, it offers no fewer than 55,000 images of European prints from the 15th to the 19th century, drawing upon the print collections of scores of archives, libraries, and museums around the world. This is a collection that - presented online in a searchable database format - provides unique access to a body of visual materials that will be of enormous value to art scholars as well as all students of early modern European history and indeed to all parties interested in Western visual culture. ARTstor's focus on assembling and integrating heterogeneous image archives creates a range of mundane but no less pressing issues. Foremost among these, perhaps, is the need to manage and massage the data associated with these collections in such a way that the end user might search and browse intelligently. In its effort to address this, ARTstor and its metadata team have benefited particularly from the Getty Research Institute (a program of the J. Paul Getty Trust in Los Angeles), which has made both its essential controlled vocabulary resources (the Art and Architecture Thesaurus, Thesaurus of Geographic Names, and Union List of Artist Names) and the expertise of its staff readily available to us.

More than read only

On the software front, ARTstor staff concluded early on that it would be essential to develop new software. This choice reflects our strong commitment to offering software tools that, like our collections, will respond first and foremost to our users. They need a software environment that, while web-based, is more than merely "read only." Those who work with visual materials must "drag and drop" images around the screen, select and store groups of images to which they wish to return later, share images with students and colleagues, present images dynamically in the classroom - online and, when a safety net is desired, offline as well. Only by controlling its own software fate can ARTstor respond with agility to the evolving needs of its users. The decision to "bundle" content and software has proven controversial. While many institutions place great value on having a "sole source" answer to both sets of needs, others (especially research universities that are committed to developing institutional digital library repositories and services) would sometimes prefer to host and manage ARTstor content in their own local software environment.

A new space

This leads us to ARTstor's effort to create a secure, trusted space on the Internet for the educational, noncommercial use of digital images. As indicated, this space is defined by a licensing framework that embraces - and harmonizes the concerns and interests of - content owners, participating institutions, and end users. ARTstor pursues this goal by focusing its outreach efforts exclusively on the large community of nonprofits. ARTstor is presently available for participation by U.S. and Canadian nonprofits: schools, colleges, universities, museums, and other institutions dedicated to education, research, scholarship, culture, and the arts. This poses its own difficulties. Independent scholars, for example, play a significant role in the arts and humanities and are too often deprived of access to licensed information resources. We currently seek to address their needs by helping "walk-in" users where participating libraries themselves accommodate them. We are also interested in following in JSTOR's footsteps by exploring other avenues that lead to the same goal, e.g., by making ARTstor available via individual membership in scholarly organizations or through public libraries. As faculty from economics to engineering and from English to ethnology start to use images in their classrooms and studies, the challenges will become even more acute. We hope that ARTstor's efforts will help the community begin to address the roadblocks encountered by the Professor Smiths of the world.

---

Barbara Rockenbach, formerly an Instructional Services Librarian at Yale University, is ARTstor's Assistant Director for Library Relations. Before becoming ARTstor's Director of Collection Development, Max Marmor was the Director of the Arts Library at Yale