 |
 |
|
|
|
|
|
 |
Feature Article 1 | ||||
|
Six Lessons Learned: An (Early) ARTstor Retrospective Author: Max Marmor - ARTstor (Max.Marmor@artstor.org) |
|
|||
|
ARTstor is a digital library of images intended for educational and scholarly use. Founded by The Andrew W. Mellon Foundation, ARTstor became an independent not-for-profit organization in January 2004. ARTstor started licensing its library to US institutions of higher education in April of that year. We welcome this opportunity to look back at ARTstor’s early development as we mark two major milestones: the passage of ARTstor’s second birthday and the number of institutions participating in ARTstor having recently surpassed 500. We are especially grateful to RLG for this invitation to step back and reflect a bit since, like others active in this arena, we are usually fully occupied navigating the swirling waters—and the occasional rapids—of this complicated and swiftly changing landscape. In this “ARTstor Retrospective,” we look back at the lessons we have learned, with particular attention to six lessons we believe will be of particular interest to others engaged with similar work. We think the key lessons we have learned fall into these categories:
Six Lessons Learned
At the same time, ARTstor is born out of the conviction that the image needs of scholars, teachers, and students across the arts and humanities converge and overlap in important ways. Consequently, while the needs of individuals in any given discipline may sometimes be unique to that field, we believe there is a vast underlying body of art and visual culture—in short, of images—that engages the interest of humanists of all stamps, and indeed of scientists like the engineers just cited. ARTstor’s goal has been to begin to define these shared needs, to address them in developing digital collections, and simultaneously to begin building out from this shared core of content to respond to the more specialized needs of both art historians and scholars in other fields, who are engaged with images in their teaching and research. We believe the value of this approach to building the ARTstor Digital Library has been affirmed over the first two years of ARTstor’s existence, especially as we learn how many users from different fields have come to regard ARTstor as an essential source of images for teaching and learning.
A case in point is our “offline image viewer.” Conceived as an alternative to PowerPoint and other commercial presentation software, it has been designed specifically to support the needs of image users in higher education and museums. The “OIV” in its latest version (2.5) has proved so successful that we will shortly release a basic freeware version to the larger community. And so a second lesson we have learned is that digital image users do indeed need appropriate, full-featured but intuitive software tools. And if one is to be responsive to user needs on this front as on others, it is best to manage software development autonomously—and where feasible and appropriate, to make one’s tools available for use with other resources as well.
The research library’s instinct is—appropriately—to build and to steward local collections. Despite decades of efforts in collaborative collection development and innovative initiatives in the area of interlibrary loan services—and despite the degree to which libraries now invest in, and their users depend fundamentally upon, licensed electronic resources—the urge to “own” rather than merely to “access” resides deep within this community. Unsurprisingly, that instinct is also in evidence when it comes to addressing the increasing need for digital image resources. This natural bent of the research library is reinforced by the traditional approach in visual resources collections—slide libraries and photographic archives in institutions large and small. Slide curators and the faculty they serve are accustomed to (literally or at least meaningfully) owning images, with all that implies for security, trust, and ease of ongoing access. Visual resources collections and visual resources professionals are now making the same transition from “ownership” to “access” that libraries have made so successfully in recent decades and with the same coupling of anxiety and excitement.
And so our third lesson is that libraries and librarians—and other campus officers as well as end users—should and in fact do take into account these challenging questions in assessing the value of a service like ARTstor. At the same time, we recognize that the entire community is still seeking to define the right balance between locally supported and remotely licensed content, tools, and services. We are working to evolve with the community that we serve, and that is one of the many important benefits of undertaking this pioneering journey as a non-profit institution. We have no wish to convince the community of something so that we can cash in on financial rewards; we are here to serve the community as myriad individual institutions find their way through this complicated and shifting terrain.
ARTstor means to be a bridge between the international community of content owners (archives, libraries, museums, photographers, et al.) and the community of end users, in higher education especially, but also in museums and other cultural organizations. That these two communities do not always see eye to eye when it comes to the educational and scholarly use of digital resources is well known and was amply demonstrated by the CONFU process some years ago.[2] Our goal has been to work with the community of content owners, on an international scale from Berkeley to Beijing, to build a digital image library that will be highly valued by scholars and teachers and, indeed, shaped fundamentally by the needs of scholars and teachers, while also advancing the missions of collecting institutions. This has entailed compromises at both ends of this spectrum. Put metaphorically, we have concluded that this essential “bridge” must in important respects remain a “covered bridge”—at least for now. By that we mean that in order to balance the concerns, interests, and needs of content owners with those of end users, we have felt obliged to create a secure network on the Internet, within which digital content can be used in appropriate ways by educators and scholars, and without, for the most part, allowing that content to be removed from the digital library for use in other environments. We have, in short, wrapped ARTstor content in the ARTstor software. And we have thereby placed real limits on our ability to interoperate with other systems and services. We have taken this approach for two reasons: First, we believe that this is the only way we can build the kind of valued collections our users say they most want from a service like ARTstor; and second, we believe it is important to keep these two communities in dialogue—a mission-driven goal we would jeopardize if we fully accommodated the interoperability interest some institutions and individuals have expressed.
While content owners and end users will continue to face points where they might not agree completely, we are pleased that the “covered bridge” we have erected is now providing for some early passage—and that it is becoming a two-way street. As we continue to establish relationships with archives, libraries, museums, and other collection development partners here and abroad, we are beginning to ease some of the restrictions on content in ARTstor. We expect, for example, to allow for significantly larger downloads of some ARTstor images later this year, especially for use in teaching. We also anticipate working with supportive content owners to enable fuller use of ARTstor images in the larger context of scholarly communications.
This already complicated picture becomes still more ambiguous, complex, and contested when the underlying work of art is still under copyright. This is, of course, the case with contemporary art and much of 20th century art as well. ARTstor is making great efforts to provide its users with a substantial body of modern and contemporary art images; we are, for example, just launching a project to digitize more than 100,000 images of works of contemporary art shown in galleries throughout New York City in the last third of the 20th century. In some instances we pursue such projects with some dependence upon the US exception to copyright law of fair use. But we are also making great efforts to reach out to artists, estates, and artists’ rights organizations. And because we are committed to making the ARTstor Digital Library available internationally (it is now available in the US and Canada, and pilots are underway in the UK and Australia/New Zealand), we want to secure a firm foundation for providing access even in countries where there are no traditions comparable to fair use. What lessons have we learned from these explorations? The obvious lesson is that this is, first and foremost, a profoundly complicated terrain, especially in the international arena, with many risks to be managed on the part of all concerned. We have also learned that artists and those who represent them care about the noncommercial, educational, and scholarly use of digital art images and that these individuals and institutions will often lend their support to the effort to facilitate such uses. And finally, we have learned that understandings can be reached that address mutual concerns without jeopardizing or compromising fair use—for ARTstor and its users or for others active in this community. The perspective we have adopted on this set of issues is a long-term one. We believe that the community is determining today whether the effort to define and enable educational use will be a collaborative effort or a confrontational one, and we are doing what we can to try to keep the exploration collaborative.
Finally, as ARTstor prepares to enter its third year as a “live” digital library with expanding collections and services, we often ask ourselves how well we are performing. Are we building the “right” kind of collections? Do our software tools lend themselves to the uses we hope to support? How is ARTstor being used, by whom, and in what ways? And of course libraries must ask themselves—and us—these questions as well. One thing we have discovered is that conventional metrics for measuring the use and value of online resources are less helpful than one might hope or expect. We attribute this to the fact that ARTstor is not a typical online information resource. It is both a reference/research tool and a tool that enables course support via shared image groups and classroom presentations. It is an educational technology, providing tools supporting classroom applications (lecture preparation and presentation), as well as student study (there are definite spikes in usage during midterms and finals!), and other pedagogical uses. It lends itself to integration with learning management systems due to the provision of stable URLs for all ARTstor images as well as for all “image groups” created by users. And above all, ARTstor exists both as an online resource among many others and as a personalized resource that has both online and offline versions. Thus ARTstor users perform a range of activities that are difficult to track and difficult to assess and interpret and that—above all—bear little resemblance to the ways in which the vast majority of online information resources are used. And thus our final “lesson learned” is that we—as a community—need better ways to assess the use and value of electronic resources that are conceived to pioneer new paradigms for teaching and learning. We are still learning further lessons, both by getting some things right and by making mistakes, and we look forward to hearing from, and learning with, all those who have a stake in how this fascinating and important exploration plays out. Notes: |
|
|||
Lesson One: The feasibility—and importance—of building a campus-wide resource that engages users across a range of disciplines without being balkanized into narrow discipline-specific collections
The “ART” in “ARTstor” requires some unpacking if our name is not to mislead. For despite its name, ARTstor was conceived from the outset as what participating libraries have encouraged us to call a “campus-wide” resource. ARTstor is intended to provide for teachers, students, and scholars throughout the arts, humanities, and social sciences—and indeed even beyond their boundaries—the kind of image library, and associated services, that have traditionally existed only in academic slide libraries serving departments of art history and related programs. We know from long experience that most academic programs outside the arts have needed, but notoriously lacked, such substantial image resources, and ARTstor is intended to redress that imbalance and to break down the organizational barriers that created and continue to sustain it. We have recently learned, for example, that two structural engineering faculty at Johns Hopkins University are using ARTstor to teach about bridge construction. This strikes us as opening the door to exciting possibilities that traditional visual resources collections have rarely had reason or opportunity to explore before locking the doors at night.[
Lesson Two: The importance, when it comes to digital images, of providing tools for teaching and research
Users want to do things with digital images. “Read only” is not enough. They want to assemble images, often in huge numbers, to shuffle and re-shuffle them into unpredictable and unanticipated permutations, to sift and filter them in sometimes indiscernible ways, and then to actively use them in teaching, learning, and research. All of these activities demand tailored, “bespoke” software tools. At ARTstor, we concluded early on that we ourselves needed to provide such tools for our users, both to support appropriate uses and simultaneously to help address the concerns of content owners about potentially inappropriate uses. And we concluded rather reluctantly that we had to build these tools ourselves, since we felt that those that were available elsewhere were either too elementary to support this range of activities or not intuitive and “usable” enough to enlist the engagement of a wide range of users, particularly as many of those users were unfamiliar with or even averse to learning new ways of doing familiar things. But, we also believed that it was in the best interest of our users if we developed our own tools, since only in that way could ARTstor be maximally responsive to user needs in an ongoing way.
Lesson Three: The ramifications for “buy vs. build” decisions on the part of libraries and other campus entities 
ARTstor is intended to enable and foster that transition by providing both a very large core body of digital images capable of supporting shared curricula in the arts and humanities and, increasingly, the kind of deep “special collections” of primary resources that alone can turn an instructional resource into an online information resource capable of supporting and fostering advanced research and scholarship. This transition clearly has implications for libraries as well as visual resources curators and image users. What is the library’s role in this development? When should image collections be developed and managed locally and when should they be licensed? Should libraries or other campus entities redundantly invest in sustaining digital image archives? Is it more cost-effective and scalable to invest in local infrastructure (and staffing) for digitizing, cataloging, archiving, and supporting the use of digital images—or to depend upon a trusted third party for many of these activities and services? These are all questions that ARTstor poses, sometimes implicitly (simply by building collections meant to be of broad value) and sometimes explicitly (as through our nascent hosting service through which we “host” image collections on behalf of participating institutions). These questions implicate libraries—and library budgets—in the areas of collections, technology, and services. And indeed, since ARTstor is both a library resource and an educational technology service, these questions have larger implications for academic programs and those who plan and administer services to them. 
Lesson Four: The trade-offs entailed by building valued, “user-driven” collections while also striving to accommodate a strong interest in interoperability with other collections and services
Having said this, we have also learned that there is much we can do to address these conflicting interests in appropriate ways. Part of our rationale in building software tools and in developing a hosting service and a personal collections tool, for example, has been to enable the individual end user to easily integrate their own images (whether personal or institutional) with ARTstor images, both offline and online. Similarly, we are now offering a growing range of interoperability solutions, beginning with federated searching via a recently released ARTstor XML Gateway.[
Lesson Five: The (perhaps unique) complexities surrounding contemporary art
Digital images rarely travel unaccompanied by rights issues, issues that are frequently ambiguous, invariably complex, often contested, and always exigent. These issues cannot—and will not be!—ignored. And these issues are shape shifters, appearing differently from one context to the next and one country to the next. 
Lesson Six: The challenge represented by the lack of appropriate assessment metrics for online resources that support both research and teaching