I wondered if any other universities or university libraries could boast of anywhere near as many useful resources, but when I went hunting, I found many others fully as remarkable. The University of Texas, for instance, harbors the Perry-Castaneda Library Map Collection [http://www.lib.utexas.edu/Libs/PCL/Map_collection/Map_collection.html], the World Lecture Hall [http://www.utexas.edu/world/lecture/], Tenet Web: The Texas Education Network [http://www.tenet.edu/], and a host of Digital Library Projects. The University of Tennessee-Knoxville site hosts Dr. George Hoemann's comprehensive American Civil War page [http://sunsite.utk.edu/civil-war/], the Mathematics Archives [http://archives.math.utk.edu/], and the library's digital projects, including the Great Smoky Mountains Regional Project [http://www.lib.utk.edu/refs/smokies/]. 

Cornell University's site is home to many important digital collections, including the invaluable Legal Information Institute [http://www.law.cornell.edu/], Computer Science Technical Reports [http://encompass.library.cornell.edu:8081/], Core Historical Literature in Agriculture [http://cdl.library.cornell.edu/chla/], and its own half of the Making of America Project [http://moa.cit.cornell.edu/moa/]. The University of Pennsylvania hosts a number of sites, including the Penn Humanities Forum [http://humanities.sas.upenn.edu/], the African Studies Multimedia Archives [http://www.sas.upenn.edu/African_Studies/Home_Page/GIF_Images.html], and Oncolink [http://oncolink.upenn.edu/], while its library offers such digital projects as the On-Line Books Page [http://digital.library.upenn.edu/books/] and A Celebration of Women Writers [http://digital.library.upenn.edu/women/]. 

The University of Washington offers a gold mine of useful Web sites, such as the Seismology and Earthquake Information Center [http://www.geophys.washington.edu/SEIS/], the Orthopaedics and Sports Medicine page [http://www.orthop.washington.edu/], and the Polar Science Center [http://psc.apl.washington.edu/]. Meanwhile, the UW Library operates a Digital Initiatives Program [http://content.lib.washington.edu/] that "works with faculty members to create online collections of interest and value," such as the Cities and Buildings Database, the American Indians of the Pacific Northwest collection, and papers and documents related to the World Trade Organization protests. 

How Do They Do That?

When a university produces one or two really good Web sites, it may simply mean the institution has a couple of really talented people on its staff, provides the server space for their work, and stays out of their way. When a university's Web server hosts dozens of outstanding resources, however, you have to think the university itself is doing something very right. I set out to find out what that was. 

After reading innumerable "about this project" pages and articles on the creation of digital library resources, I interviewed creators of the Web sites and talked to digital library project directors and campus administrators. I learned that projects came in three flavors: 

• The work of lone individuals who carry their work with them when they move on, like the On-Line Books page, which moved with its creator, John Mark Ockerbloom, from Carnegie-Mellon to the library at Penn, and the Physics Preprint Server, started in 1991 and formerly hosted by Los Alamos National Laboratory, and acquired by Cornell in September 2001, where it is now known as the ArXiv.org E-print Server [http://www.arxiv.org] 

• The work of individual faculty researchers or research teams. 

• The work of librarians in digital library projects, both following their own ideas and assisting faculty in their projects. 

A Research Environment

Wendy Lougee, assistant director for digital library initiatives at the University of Michigan Library, and Eric Shulenberger, head of Multidisciplinary Research Development at the University of Washington, both made a point of telling me that the creation of educational Web sites is in many cases incidental to the real point of the exercise, the research itself. 

Even when university libraries digitize collections specifically for dissemination on the Web, the collections chosen usually reflect the research needs and interests of university faculty and programs. According to Kody Janney, Digital Initiatives coordinator at the University of Washington Libraries: 

We like to focus on strengths of the UW. For instance, we worked with the Center for Labor Studies on the WTO History Project. We have mounted quite a lot of Seattle and Pacific Northwest material, as that is an area of unique strength for our institution. An example of research support is the Mt. Saint Helens Succession Collection, which contains post-eruption photos taken of the same site over a period of years. We are working with an archeology professor on images from a dig. We support both research and curriculum. 

Technical Support

The more the institution encourages and supports research, the more useful Web sites it will likely host. Some of the research support services are formal and campus-wide, such as technological resources (the Texas Advanced Computing Center, for example, or University of Tennessee's Applied Visualization Laboratory), and assistance in grant-writing. Most of these institutions offer many venues, formal and informal, for faculty to learn to design instructional multimedia. Most of them also offer support with statistics and datasets, like the University of Tennessee-Knoxville's Statistical and Computational Consulting Center. 

Many of the libraries have digital initiative programs, where faculty can propose ideas for digital collections and get both the funding and technical assistance to bring them into being. [To get an idea of the extraordinary range of such projects, see the Digital Library Federation's guide to Public Access Collections at http://www.hti.umich.edu/cgi/b/bib/bibidx?c=dlfcoll.] 

Perhaps even more important, though, is informal support within and across departments and schools. On most of these campuses, each school has its own servers, system administrators, and technical support. In many departments, faculty have been using the Net for years; many of the most productive universities on the Web started their online networking as part of the ARPANET and NREN networks. That means there are enough faculty who are very comfortable with the technologies that the idea of putting your knowledge on the Web is pervasive, and help in doing it right at hand. 

The Habit of Collaboration

In many ways, universities are a natural place for digital projects to bloom, because faculty are in the habit of working collaboratively with researchers in their field at other universities or government agencies. Cross-boundary collaboration is part of the stated mission of many projects, such as the Penn Humanities Forum. Many Web projects at the University of Washington, which is the largest single recipient of federal grants, are natural outgrowths of extensive work with the Department of Defense, the U.S. Geological Survey, the National Weather Service, and numerous other federal and state agencies. The University of Michigan has partnered with the University of Illinois to create an "academic Hotbot" — a search engine to retrieve material from digital libraries — and with Cornell in the Making of America collection. [For more information on the Open Archives Initiative Metadata Harvesting Project, check out http://oai.grainger.uiuc.edu/michigan.htm.] The University of Tennessee Center for Information Studies has a stated mission of "building interdisciplinary research teams to bring the full range of University research talent to bear" on "problems involving the intersection of information technology, knowledge-based systems, and user behavior" [http://www.sis.utk.edu/cis/]. 

University faculty, who often have scant respect for stupid artificial boundaries, also frequently collaborate across disciplinary lines; in fact, some institutions actively encourage them to do so. The sole job of Eric Shulenberger at the University of Washington is to provide a kind of one-man "skunkworks," to help faculty create cross-disciplinary projects, find logical research partners, and design and seek funding for them. Among the research projects he's helped nurture is the Urban Ecology interdisciplinary education and research Web site [http://www.cfr.washington.edu/research.urbaneco/main2.html], which draws on faculty from Forestry, Urban Planning, Geography, Public Policy, and Computer Science. 

Many research universities have embraced a relatively new and evolving concept called a "collaboratory" — "a center without walls, in which the nation's researchers can perform their research without regard to geographical location — interacting with colleagues, accessing instrumentation, sharing data and computational resources, and accessing information in digital libraries," according to the article by Kouzes, Myers, and Wulf that started it all ("Collaboratories: Doing Science on the Internet" in IEEE Computer, August, 1996). 

One such project from the University of Michigan's School of Information is a Collaboratory for Research on Electronic Work [http://www.crew.umich.edu/], which "focuses on the design of new organizations and the technologies of voice, data, and video communication that make them possible." This collaboratory draws on experts in computer, information, cognitive, and social sciences. In fact, one of its projects examines The Science of Collaboration itself [http://www.scienceofcollaboratories.org/html/AboutSOC/Activities.html], which plans to compile a knowledge database about the process. 

The result of such collaboration seems to be a wonderfully generative cross-fertilization of ideas. When I spoke with Ruth Ludwin, of the University of Washington's Seismology and Earthquake Information Center, she talked about the "loose cannon management style" and the numerous creative partnerships it seems to inspire. The Seismology site hosts the Volcano Research Center and the Tsunami Research center, for instance. I was left with an impression of UW as a vast intellectual fermentation vat, always yielding unexpected and interesting results. 

A Fast Learning Curve

It seems clear that success on one project creates partnerships, trust, and a skills platform that lead to new initiatives. Joe Janes, the director of the University of Michigan's School of Library and Information Science at the time he founded the Internet Public Library, had picked up the necessary skills from his involvement in the UM Library's Digital Library Project, which itself benefited from early collaboration on the JSTOR project. The University of Iowa's Scholarly Digital Resources Center was a logical outgrowth of the university's Information Arcade, and its Center for Electronic Resources in African Studies also grew out of the faculty's previous partnerships with the University Library on research proposals. 

Standards and Continuity

Some projects, especially in the early days, began on a kind of ad hoc basis as the project creators were figuring out the Web, and eventually had to conform to evolving institutional "best practice" standards. But at many institutions, the standards and best practices came first. As early as 1991, three University of Michigan departments — the School of Library and Information Studies, the University Library, and the Information Technology Division — sponsored an ongoing faculty symposium to discuss issues presented by electronic information, including collections, user support, and funding. 

The University's Digital Library Program was an outgrowth of that symposium. Originally, it worked in a project mode, says Wendy Lougee, but increasingly over time it has become more focused on providing functionalities that help faculty to create and preserve Web projects — durable formats, metadata standards, collaboration with subject-specialist librarians, and a usability testing process [Wendy P. Lougee, "The University of Michigan Digital Library Program: a Retrospective on Collaboration within the Academy," Library Hi Tech, vol. 16, no. 1, 1998, http://www.umich.edu/~gateway/policies/]. Sustainability is assumed and planned for all projects; continuous evaluation is built in, so that projects can be improved, preserved, and migrated to other platforms as technology changes, and so that lessons learned on one project will incorporate into others. 

A similar evolution occurred at the University of Iowa, which created the Scholarly Digital Resources Center "in order to go beyond collecting electronic texts to provide a platform for creative projects in electronic text and multimedia format" [Barbara I. Dewey and Carol Ann Hughes, "Sharing Minds: Creating the Iowa Scholarly Digital Resources Center," Information Technology and Libraries, vol. 18, no. 2, June 1999]. Another goal at Iowa, and increasingly at other universities, is interoperability, enabling the user "to easily cross the boundary between the texts on the Web and bibliographic records in the library's online catalog without barriers...." 

Most of these universities also post guidelines for officially sponsored pages, to insure that the pages are appropriate to the university's mission and reflect the highest ethical and legal standards of conduct. The University of Michigan's Publishing Policies, Guidelines, and Instructions offer an unusually thorough and specific example [http://www.umich.edu/~gateway/policies/]. 

Educational Mission

Many of the Web sites serve multiple educational purposes. The sites: 

• allow researchers to pass on knowledge to both their own students and to the community whose taxes paid for the creation. 

• offer students an opportunity to learn by contributing to the development of the Web site. 

• steer students to quality Web resources and attempt to show them what genuine scholarship looks like. 

Similarly, the University of Iowa's Scholarly Digital Resource Center uses Information Arcade graduate assistants. Hired for 3 years at a time, the assistants develop increasing expertise with both technology and information management, while they help build projects like the CERAS collection [see above] and the Chautauqua project, which serve a wide public. Cornell's Legal Information Institute employs Cornell law students, who learn on the job. Tom Bruce, co-director of the Institute, was quoted as saying, "We use students extensively as editors. They are a tremendous source of creativity and of editorial and substantive expertise" [Linda Myers, "CU Law Institute Web Site Has Latest Legal Information, from Miranda to Elian," Cornell Chronicle, April 27, 2000 http://www.news.cornell.edu/Chronicle/00/4.27.00/Legal_Info_Inst.html]. 

The University of Maryland Library provides a particularly nice example of libraries that use their Web pages to steer students to the best resources in their fields [http://www.lib.umd.edu/ETC/SUBR/resources.html]. Each subject page points students to databases, electronic journals, subject gateways, specialized tools, and — best of all — human resources, e.g., the related University departments, as well as the librarians who serve as subject specialists for specific topics. 

Grace York, of the University of Michigan Library, says she created her impressive Internet resources — Government Information on the Web, Statistical Resources on the Web, and Documents in the News — "to serve as a reference tool for my staff in answering the questions it receives, and to serve as a teaching tool for classes" [my emphasis, "Webmastery: A Guru Named Grace," Searcher, February 2000]. 

Public Service Mission

One of the questions I asked the people I interviewed was, "Does your institution have a fundamental commitment to making taxpayer-funded research widely available to citizens?" Most of the university representatives said yes, though some of them pointed out potential or actual conflicts with intellectual property rights, as in the ownership of distance-education course materials prepared by individual professors. But Jennifer Conway, Associate Director of the University of Pennsylvania's Humanities Forum [http://humanities.sas.upenn.edu/], said: 

Absolutely. Universities = knowledge. The University of Pennsylvania, consistently ranked among the nation's 10 universities, has a number of world-class programs for which public outreach is fundamental: e.g., Center for Community Partnerships, Center for Advanced Judaic Studies, Center for Bioethics, Fox Leadership Program, Penn Humanities Forum, numerous initiatives within The Wharton School (e.g., Leonard Davis Institute of Health Economics, Lauder-Fisher Center), the University of Pennsylvania Medical Center (e.g., Oncolink, Institute on Aging, among many), Penn's School of Nursing, Grad. School of Education, Law School, ... and on and on.... 

Certainly many other university Web projects have to be regarded as pure acts of public service, making unique university resources — documents, databases, images, and information — available to far more people, such as the Perry-Castaneda Library Map Collection at the University of Texas Library [http://www.lib.utexas.edu/maps/historical/], and the Rutgers Alcohol Studies Database [http://www.scc.rutgers.edu/alcohol_studies/alcohol/], formerly available only to those who visited the Center for Alcohol Studies at Rutgers. 

Continuity

The Web resources that have the best chance of long-term survival clearly are those projects integrated with the digital library initiatives run by academic libraries. Visit the Digital Library Federation [http://www.diglib.org/], "a consortium of libraries and related agencies that are pioneering in the use of electronic-information technologies to extend their collections and services" by: 

• identifying standards and "best practices" for digital collections and network access. 

• coordinating leading-edge research and development in libraries' use of electronic information technology. 

• helping start projects and services that libraries need but cannot develop individually. 

The University of Michigan and Cornell are so certain of the continued joint maintenance of the Making of America collection that the schools are creating cataloging records for the documents in it, which any library can add to its own catalog [Walker, Kizer, "Integrating a Free Digital Resource: The Status of Making of America in Academic Library Collections," RLG DigiNews, February 15, 2002 [http://www.rlg.org/preserv/diginews/diginews6-1.html]. 

The good news is that on many campuses, digital library initiatives are not solely the preserve of the library but are campus-wide partnerships, funded and supported at the highest levels of the university administration, which bodes well for the stability of individual faculty projects operating under their auspices. 

Some prime resources, however, such as the Legal Information Institute, rely on a patchwork of funding sources — the university, grants, partnerships, and individual donors — which makes it more difficult to assume the long-term viability of the Web sites in their existing, noncommercial form. Others that may be purely personal or course-related faculty projects have been known to vanish, like the excellent guide to SciEd resources formerly hosted on the astronomy server at the University of Washington. 

Inadequate Reward Structure

It was obvious that everyone I talked to got great psychological rewards from their work on the Web. That's just as well, because it's equally clear that while librarians may advance their careers by creating digital resources, faculty generally are not rewarded for their Web work with pay increases, promotion, sabbaticals, or tenure. 

The problem seems to be that traditional peer-review mechanisms for research have not yet expanded to consider resources created for or published on the Web, though there are signs that some scholarly associations are beginning to consider their responsibilities in this area. A new resource called MERLOT [Multimedia Educational Resource for Learning and Online Teaching] has been set up specifically to provide peer review of discipline-specific Web resources, applying the same measures of scholarship as those applied to standard scholarly publication [http://www.merlot.org/Home.po]. 

As Eric Shulenberger told me, academic inertia is a powerful force, and until these measures of scholarly achievement gain wide credibility, untenured scholars who devote enormous amounts of time to the creation of Web resources, at the expense of formal publishing, may place their careers in jeopardy [Jeffrey S. Young, "Ever So Slowly, Colleges Start to Count Work with Technology in Tenure Decisions," Chronicle of Higher Education, February 22, 2002, http://chronicle.com/free/v48/i24/24a02501.htm]. 

Where Librarians Shine

I found it remarkable that at so many research universities, librarians were there first, in many cases as early as 1991, leading the charge to develop major digital projects. That's because the librarians were among the first to realize the challenge presented to the knowledge enterprise by a glut of easily available but unvetted information, seeing the need to do for the Net what librarians have always done for traditional resources: select, arrange, annotate, and educate. 

But perhaps the greatest contribution of librarians to the digital enterprise has been long-term thinking and a broad-based view of information. After all, librarians knowwhat happened to documents on microcards, music on 8-track tapes, and videos in Betamax format. We also know that valuable information is distributed in multiple formats that should not be mutually exclusive. From the first, librarians have insisted on standards, on interoperability of databases, on accessibility, on scalability, and on the maintenance and preservation of digital records. 

The great research institutions have provided a supportive environment for the creation of educational internet resources. But their librarians told the institutions what needed to be done and how to do it. Let's hear it for them.