Goals and Missions

We are delighted to have tens of thousands of users a day from all over the world, whose feedback helps to improve our service.

Introduction

The Etext Center was founded on a commonsense vision that combines emerging network and digitizing technologies with our subject expertise and library skills.

Our twin mission is as follows:

to create an on-line archive of standards-based texts and images in the humanities

to build and support user communities adept at the creation and use of online resources

The assumptions that lie behind this mission statement are presented in brief outline below; they have worked well for us, and in large measure they are generalizable to other digital publishing endeavours, and to the general notion of a digital library itself.

These issues can be found treated at greater length in various publications that have come out of the Etext Center, some of which are online.

Standards are your friend

XML, SGML, EAD, TEI, TIFF -- our life is a flurry of those acronyms that allow us to create data in as standardized a fashion as possible. Some, like SGML, are de jure standards; others like tiff, are ubiquitous de facto ones, and we cross our fingers.

Data that owns its own structures -- tagged data that needs only ASCII to survive is central to our digital library production; it networks well, is not beholden to any piece of software for its existence, and it allows products from multiple sources to be accessed together through a common web-based interface.

Our consistent use of SGML and XML has meant that our data is as usable and as malleable now as it was in 1992. The paucity of SGML software continues to be a burden, but our texts have adapted quickly to the arrival of the web, to XML, and are now being rethought at the metadata level for full-scale cross-database searching. In an online library, data needs to be nimble to survive -- it needs to be able to re-shape itself to take advantage of new display and layout possibilities, new and larger-scale searching and text analysis possibilities.

Software comes and goes, but the standardized data remains. With our tight library budgets, this is all we can afford to invest in on any real scale.

It is important to know digital ephemera when you see it: there's nothing wrong with buying or creating proprietary data -- multimedia CDs, pdf files, or elaborate PowerPoint classroom presentations, for example -- as long as you know that those objects are not going to be part of a long-term library collection, and will either leave the collection or present costly conversion needs.

Online delivery is the dominant delivery

More and more, our users expect the digital library to come to them, wherever they happen to be. This has been a fundamental aspect of our service since 1992:

"When Thomas Jefferson built an 'academical village' in the rolling greenery of his native Virginia, he put a white-porticoed pantheon at its centre and filled it with books. The library's prominence proclaimed his revolutionary faith in learning and truth. In today's University of Virginia, a new revolution can be glimpsed....a shared commons in cyberspace, available to anyone prepared to go to a library." The Economist, August 27, 1994

Data that is delivered to us only on CD-ROM in proprietary formats is too often orphaned in our collections.

Core traditional library skills are also core digital library skills

The digital library is a library more than it is a computer center. Work in the Etext Center is more textual than technical -- we are not a lab in which books are fed into computers.

User training, collection development, and cataloging skills are central to a vibrant electronic collection, and much of our success is a direct result of our ability to integrate the electronic data into the library as a whole, and to draw on the skills that make the print-based library work.

Collections

Content Reigns

Nothing brings people back to an online service like a well-selected, reliable, integrated, and cataloged set of data, especially if it is searchable and can be delivered quickly through a common interface across collections.

Collections, like gardens, need design, shaping, and building. Our subject librarians allow us to build collections that have coherency; judicious selection serves both our local University users and internet visitors equally.

The Electronic Text Center's holdings include approximately 70,000 on- and off-line humanities texts in many languages (including online Chinese and Japanese literature) and hundreds of thousands of related images (book illustrations, covers, manuscripts, newspaper pages, page images of Special Collections books, museum objects, etc.)

Whenever legally possible, we provide public access to our browseable and searchable texts, and have thousands available in this form, including the following highlights:

African American, including Letters from Liberia; Native American; American Civil War; including the letters and diaries in the Valley of the Shadow, alongside the Booker and Nettleton collections; Thomas Jefferson; William Shakespeare; Samuel Taylor Coleridge; Women Writers; Young Readers.

We do not have the legal ability to provide general access to materials that are commercial publications -- materials for which we are simply a customer. Examples include the Chadwyck-Healey databases (including the Patrologia Latina, all of English poetry up to 1900, and thousands of other poetry, prose, and drama titles) and the Oxford English Dictionary. All such commercially-available publications are available to University of Virginia users; many are also licensed by VIVA, the Virtual Library of Virginia.

One interface/many collections

Online SGML data allows us to provide common navigational metaphors across all our collections, irrespective if they are locally-created or from any of a number of different vendors. Whether you are searching the Japanese language literature along with its English translations, or the virtual Poe archive, biblical materials or the Salem Witch Trials -- or indeed searching them all together in Modern English (excluding the Japanese materials) -- you are using the same search engine and the same web form structure. Users can choose whether to see results summarized by work or by result, and move from a brief context (KWIC concordance view) to larger amounts of context: page, chapter, act, section, etc.

The heirarchical nature of SGML files not only makes it easy to "step up" to a larger amount of surrounding context to a search result, but also to build "on-the-fly" tables of contents for easier browsing.

Users

User communities need active creation

The value of a library service that provides training in and orientation to a new medium was evident from the beginning -- this from early 1994:

"The E-Text Center is both a real place [with] a helpful staff as well as a cornucopia of 'on-line' texts. I've dreamed of owning a library of great books, so the mere listing of just some of the e-texts available through the center makes me giddy." Anne A. Oplinger. "The Pleasures and Perils of Setting Up House in the Electronic Academical Village." UVA Alumni News. January/February 1994.

Our frequent series of short courses in TEI, HTML, SGML, text-analysis software, scanning techniques, and now XML have contributed to the general vitality of our humanities computing endeavors at Virginia.

Well-crafted training sessions, coupled with the ad hoc user support given to walk-in clients in our digital Centers, provides the "primordial soup" from which rises both a general literacy in the use of on-line resources, and also a growing suite of individual research and teaching projects.

Such training is a fundamental element in the building of a user community for electronic resources. Such a community is in turn a fundamental necessity in a vibrant digital library. The provision of electronic materials is not enough in itself, now any more than it was when we began our efforts. It is not simply that supporting a user community gets the most value from the data one buys, and garners support for a new undertaking within an institution. More important, and more exciting than that, the community you create quickly becomes a vital part of the resources on which you draw: users create texts, re-purpose existing product, and provide a high degree of critical observation on the various aspects of the digital library in which they invest themselves [see Users as Responsibility and Resource.].

Strength through partnerships

The Etext Center runs on partnerships. Even before we talked to initial users in 1992 we were "selling" the Center to those in the library who had skills or other resources that we needed, and also to de-mystify what we do -- to help incorporate ourselves into the fabric of the library. Since then, we have worked closely with Special Collections, with local museums, with faculty and student users from near and far, and with other libraries and archives.

Current user projects that demonstrate the range of our collaborations include:

The Plymouth Colony Archive Project at the University of Virginia
Witchcraft in Salem Village: the Witchcraft Trials of 1692. Ben Ray, University of Virginia Religious Studies Department.
The Abraham Cowley Text and Image Archive Dan Kinney, University of Virginia English Department
Lienu zhuan Anne Behnke Kinney, Associate Professor of Chinese, University of Virginia
Mark Twain in His Times: Stephen Railton. University of Virginia English Department.
Thomas Jefferson on Politics & Government: Quotations from the Writings of Thomas Jefferson. Eyler Robert Coates, Sr., editor.
Studies in Bibliography: 1948-1997: 50 years of articles on literary, historical and bibliographic matters, from the Bibliographical Society of Virginia.
Journal of Southern Religion : the first scholarly journal devoted to the study of religion in the American South.
Arboretum Committee Minutes (1978-1999)

Many more can be found on our projects, publications, and partnerships pages.

Increasingly, we are using SGML and metadata standards to help users build "data partnerships" -- to bring otherwise separated pieces of content together into personalized data collections. Much of our current efforts are focussed on extending the scope and ease of this facility.

We are delighted to have thousands of users a day from all over the world coming to use our resources, and their feedback helps to inform our efforts.

Conclusion

So, a sustainable digital library for us is comprised of standardized data delivered online through a common interface (as opposed to a CD-ROM collection with multiple interfaces); common metadata across collections are already allowing us to achieve cross-database access, and we expect to see this in more and more sophisticated forms.

Amidst all the frenzy, excitement and anxiety caused by the arrival of a major new medium, it is heartening to see that parts of the library world have understood and incorporated these skills, and are firmly on the cutting edge of the world's use of digital publications.

The digital library succeeds to the extent that it can re-articulate traditional library skills in a new medium -- cataloging, collection development, acquisitions, preservation, reference, users services, special collections. Activity in the digital medium allows us to re-articulate the social, pedagogical, and intellectual roles of the library in an academic institution, and to serve as a rich content provider for many thousands of other users worldwide.

David Seaman
University of Virginia Library 1993/95/99/02