One of the great advantages of the web is that there's a bunch of free stuff—that's the truism, anyway. But free stuff comes from somewhere, and it is rarely, if ever, free to produce. I am interested in exploring some of the costs of digital work, using as an example the Walt Whitman Hypertext Archive, a project I co-direct with Ed Folsom.¹ Since 1995, many people, myself included, have described our site as free, yet a considerable amount of resources continue to go into its making. I want to explore that conundrum.

First, though, some background: in the mid-1990s, some scholars (especially Ed Folsom) began talking about the need for a hypertext edition of Whitman's works. At the time, I was teaching at the College of William & Mary, and one of my graduate students, Charles Green, was keenly interested in the new developments in textual scholarship and the new digital archives that were only then beginning to appear. Green and I traveled to the University of Virginia to meet John Unsworth, director of the Institute for Advanced Technology in the Humanities (IATH), and Jerome McGann, director of the Rossetti Archive. At our meeting, I became enthusiastic about attempting to produce an electronic edition of Whitman despite the magnitude of the undertaking and the difficulties we inevitably would encounter. Still, I recognized that a fortunate set of circumstances was at hand: the University of Virginia has one of the great collections of Whitman manuscripts; I was then located relatively near Charlottesville; and leading people in humanities computing were offering to lend assistance. When Ed Folsom agreed to serve as co-director of the Whitman archive another crucial element fell into place.

From the start, our aim has been to produce a scholarly edition of Whitman on the web. We are attempting this in part because Whitman's writings defy the constraints of the book. Documents associated with a Whitman poem might well include an initial prose jotting containing a key image or idea; trial lines in a notebook; a published version appearing in a periodical; corrected page proofs; and various printed versions of the poem appearing in books, including (but not limited to) the six distinct editions of Leaves of Grass. The fixed forms of print are cumbersome and inadequate for capturing Whitman's numerous and complex revisions. Moreover, the economics of print publication have led previous editors to privilege one edition or another of Whitman's writings—usually the first or last version of Leaves of Grass. Our goal is to create a dynamic site that will grow and change over the years. We are currently putting online both facsimile and etext versions of all the editions of Leaves of Grass (other titles will go online as time and resources allow). We recently posted an extended biography of Whitman that Ed Folsom and I wrote; eventually, this biography will contain links to photos, maps of Whitman-related locations, and short essays about Whitman' s friends and associates. In addition, the archive provides access to the contemporary reviews of Whitman's work, all known photographs of Whitman (complete with annotations), and introductions to each edition of Leaves. We also offer the only comprehensive current bibliography of work—including books, essays, notes, and reviews—about Whitman.

Nothing appears by magic: we still live in a world of labor, expenses, payments, and a multitude of material objects down to the level of wire and cable that make possible a virtual archive. When users visit a deep scholarly archive on the web they are experiencing the (mostly real) benefit of displaced costs. Instead of money being spent by the user at the point of contact, money is spent elsewhere along the line: by universities in the form of faculty time, equipment, graduate student assistance, and internal grants; by external funding agencies; and, in our case, curiously, by more than one publisher.

The involvement of publishers is paradoxical, counterintuitive, and especially worthy of exploration. When Ed Folsom and I had just started attempting to make Whitman's vast work easily and conveniently accessible to scholars, students, and general readers, Primary Source Media, a commercial publisher, unexpectedly asked us to produce with them a CD-ROM that would overlap with our own plan of work. With great speed—though without editorial introductions and sophisticated tagging—they enabled us to make available an extraordinary amount of Whitman material that had never before been electronically searchable: all twenty-two volumes of the New York University Press edition of the Collected Writings of Walt Whitman, all six editions of Leaves of Grass published in Whitman's lifetime, all 130 extant photographs of Whitman, hundreds of digital images of poetry manuscripts and more. There was a downside, however: The material came to consumers with a hefty price tag. I'm sure the pricing was partly influenced by the large permission fees Primary Source Media had to pay New York University Press. Interestingly it costs only about a dollar, as a process, to burn a CD-ROM, so Primary Source Media could have aimed to recoup its investment plus make a profit by selling thousands of copies at, say, twenty dollars, or far fewer copies at a high price. They chose the latter strategy. Ed Folsom and I undertook the editing as work for hire, receiving a one-time payment. We do not get royalties and had no influence on their marketing and pricing policies. We have been told that the Whitman CD-ROM was a business success, that Primary Source Media did much better than merely recoup its investment.

The data produced by Primary Source Media was tagged in Borland database format, a proprietary coding system. In my view, Primary Source Media would have been much better off to use SGML, a recognized international standard that would ensure cross-platform usability, address the need for long-term preservation, and facilitate the exchange of data. Initially, it appeared that Ed Folsom and I would have a long-term working relationship with Primary Source Media because, after issuing the CD-ROM, the publisher proceeded to move Whitman material online, and we were well on our way to coaxing Primary Source Media toward the SGML world. In addition, Ed Folsom and I, attracted by the idea of providing easy access to the works of the self-styled poet of democracy, persuaded Primary Source Media to donate the out-of-copyright etext of Whitman's writings to the Electronic Text Center at the University of Virginia, where the texts would be available to the world without charge.² This was a significant amount of material—all six editions of Leaves of Grass and Whitman's prose works. Yet the request was not totally outlandish because we realized that the sales potential of the Primary Source Media CD-ROM stemmed from their success in making the modern copyrighted New York University Press edition of the Collected Writings available in electronically searchable form (for those able to afford it). We argued that donating some nineteenth-century texts to a "free" site would be a good public service and that this would support an educational endeavor (we had recently received a FIPSE grant from the U.S. Department of Education to develop pedagogical material in conjunction with the Dickinson Electronic Archives, edited by Martha Nell Smith, et al.).³ Gradually, as processing allows, the material requested from Primary Source Media is being made publicly accessible. Perhaps what mattered is that Primary Source Media saw an opportunity to exchange data for know-how. That is, their staff saw a chance to benefit from this arrangement because they were interested in launching SGML publishing initiatives and felt they could learn some of the techniques David Seaman and his team developed for automating the conversion of the text from Borland database form to the Text Encoding Initiative (TEI) standard. (Primary Source Media had used Borland on a number of big projects.) I don't know whether the good deed argument or the hard-headed argument worked better. Incidentally, I might say that many of our plans with Primary Source Media went for naught since the firm was bought out by the Gale Group, which seems to have scuttled all plans to develop SGML publishing in conjunction with deep archives of single authors. But the ongoing cooperation of Frank Menchaca, senior editor at Primary Source Media, in continuing to provide etext at no charge represents a commitment to public access (this despite the lack of any compelling benefits to the publisher, given their change of priorities).

Three other publishers have assisted us: the University of Iowa, which allows us to reprint and reformat in annualized form the quarterly bibliographies appearing in the Walt Whitman Quarterly Review; Cambridge University Press, which allowed us to use the etext of all of the contemporary reviews I had earlier published with them; and Garland Publishing which granted us the right to reproduce approximately 10 percent of the entries in The Walt Whitman Encyclopedia. Iowa cooperated because my co-director, Ed Folsom, edits the journal and controls copyright. Cambridge obliged us, I suppose, because they didn't actually own the material they had printed in book form: that is, all the reviews were already in the public domain. Having priced the volume I did for them at $95 dollars in 1995, Cambridge realized full well that their sales were primarily to libraries and they had pretty much already exhausted that vein. Garland's situation was similar: their sales had been made, and they probably concluded that giving away some of their product would not hurt any potential future sales but might actually help by raising the visibility of the Walt Whitman Encyclopedia.

Our good luck with publishers has extended to libraries as well (though there have been some exceptions, as described below). Currently the Whitman team, with funding from a Collaborative Research grant from the National Endowment for the Humanities, is concentrating on editing the poetry manuscripts, fundamentally important documents that never before have been gathered, transcribed, encoded, and made available. We are providing both digital images of the individual manuscript pages and transcriptions. One publicity person said that we are, in effect, unlocking the doors of locked-up rare book rooms. However, the task is not as easy as turning a key. Currently the end user experiences no difference whether she encounters a donated set of images—like the wonderful scans we received from Special Collections at the University of Virginia—or images that we have had to pay for. All of the manuscripts are experienced in a uniform way, at no cost, whatever the expense of an individual item to the project. Ideally of course, in the ambitious way of recent electronic archives, we would like to provide images of every single poetry manuscript that Whitman left. That probably won't be possible because chasing down every last manuscript is a never-ending task: new Whitman material keeps turning up, as seen recently in a significant sale of material at Christie's.

Moreover, there are complexities because the economic, preservation, and permission policies of individual libraries differ from one another significantly. Certain libraries could be described, kindly, as aggressively hostile. I quote from one letter: "our standard permission fee for non-profit sites is $65 / per image for the first 20 images -and $40 / per image thereafter. Permission is granted for one-time, non-exclusive use, for a period of up to seven (7) years. We also ask that resolution for the internet be limited to 72 dpi, that a watermark be embedded into the image(s) that can withstand compression, and that the standard permissions statement appear...." The Whitman Archive is not trying to build something ephemeral but a developing product and an ongoing editorial process that can be passed on, reused, and improved by future scholars. Images that must be taken down after a few years are of little use. Like the recently issued Handbook for Digital Projects, we start with the "premise that investing in digital conversion only makes sense if institutions are prepared to provide long-term access to digital collections."⁴ Our standard for digital scans is 600 dpi; an image at 72 dpi is of such poor quality as to be of little value to scholars. After being presented with such a combination of barriers—high price, low quality, and limited time of use—I wonder why this library did not just forthrightly refuse to cooperate.⁵

Despite this and one other case, libraries in general have been remarkably supportive and forward-thinking. I have been especially pleased with the cooperation we are getting from libraries as we explore the feasibility of creating a virtual finding aid for Whitman manuscripts, an online guide intended to pull together information about holdings now dispersed in over sixty libraries. This project should provide an opportunity to experiment with methods for virtually reintegrating dispersed collections of Whitman manuscript materials using the standard for archival description, Encoded Archival Description (EAD); this project should also offer an unusual opportunity to experiment with a deeper engagement between scholars and archivists, in which scholars might enrich the item-level descriptions of archival materials.⁶ We are currently seeking grant funding to support this complex technical, social, and intellectual undertaking.

Grants help finance expensive tasks, but they have a less obvious economic importance in providing validation for projects. Recent developments in higher education—an extraordinary concern with rankings and a shift away from state support—have intensified the pursuit of grants at many institutions and, accordingly, have increased the standing of those with a track record of getting grants. The validation received from a grant can offset the questions that are sometimes raised about electronic work. We live in a time, still, when some departments refuse to credit properly scholarly editing, and an editor who chooses to work on the web—given that some departments resist crediting internet publications—is taking a double risk. Some colleagues may ask: How do we know whether electronic work is any good? Should it really count? Isn't it ephemeral? Others may assert (ignoring many exceptions) that web publication is not refereed and thus should not count.

No doubt Ed Folsom and I found it easier to work on the Whitman project because we had already been promoted through the ranks and thus were insulated from concerns about job security and the next promotion (though we remain subject to annual merit evaluations). Electronic scholarship is a trickier business for graduate assistants and assistant professors. It can pay off in significant ways, but the reception such work will receive is more uncertain than for comparable print publications.

Yet even while academic departments are often ambivalent, at best, about crediting electronic scholarship, they frequently provide financial support for these projects. The reasons departments are willing to do so are complex and varied—just as they are when departments give release time or student assistance for anything—for example, a traditional monograph. Interestingly, graduate students work on web projects, by and large, when departments not only approve of these undertakings but are willing to underwrite them at least to some degree. Within an academic reward system noteworthy for its paradoxes, graduate students operate economically in ways that are mainly straightforward. For the Whitman project, graduate students work a set number of hours and are recompensed for it by salary, tuition waiver, and benefits. First at William & Mary and now at Nebraska, I have had one or two students helping me (working a combined total of anywhere from seven to twenty-seven hours per week). Nationwide, graduate student wages, benefits, and working conditions are receiving increased and needed attention. I wish I could say that students working for the Whitman project fare better than their peers, but in terms of direct compensation for their effort they receive an amount neither better nor worse than is typical for graduate students with other types of assistantships. However, students working on humanities computing projects often develop distinctive—and highly marketable-skills. While enriching and diversifying their record as they prepare, most often, for work as professors, they also provide themselves with skills and knowledge of information architecture that leave them open to other types of academic employment, employment that frequently pays better and has better job security than a tenure-line position in the humanities. The first three students who worked for me on the Whitman project—Charles Green, Robert K. Nelson, and Matt Cohen—were hired into full-time staff positions at William & Mary in Information Technology. In the face of a difficult academic job market, gaining specialized knowledge and marketable skills are not the worst things in the world, especially when students can demonstrate that the experience enhances their academic profile.

For a graduate student, working on a large electronic project may provide other indirect benefits with some economic implications. With the Whitman project, graduate students encounter a somewhat unusual form of scholarship and a different sense of the academy and its possibilities. The scholarship that they see modeled is no longer inevitably and only the solitary professor working on a monograph. The activity is more social and, I think, frequently more rewarding for that very reason. Students on a project often work far more closely with faculty than did students with other types of assistantships. Humanities professors seldom make much money on sales of their publications, but significant publications are the route toward promotions, merit increases, and mobility in the field. These students enjoy unusual access to archival material, make and share in new discoveries, and consequently have greater publication opportunities than most of their peers.

In various ways both subtle and profound, the web environment is contributing to altered social and economic circumstances that directly affect how professors and students work, how that work is valued, and what work is in fact possible to contemplate. Earlier, I mentioned the "mostly real" benefits of so-called free sites. We might ask: Is what the Whitman archive has done a sustainable model for the production of other full-scale scholarly editions on the web? We have been fortunate with grants, publishers, libraries, and generous universities. But if it requires such a constellation of good fortune to produce an electronic scholarly edition, do we have a sound economic model in place? As the questions imply, I don't think we do.

These days projects can hardly rely on ample grant money. The NEH is considering new restrictions that would limit the number of times that scholarly editions can be funded. Meanwhile publishers have not adequately negotiated their roles in this new environment. Commercial publishers have been too ready to compromise long-term viability and scholarly thoroughness for short-term goals, and university presses have been slow to engage with electronic publishing perhaps because of concerns about the learning curve and the costs of web publishing. I am heartened, however, about the new developments involving IATH and the University Press of Virginia. With help from a million-dollar grant from the Andrew W. Mellon Foundation, the Press is in the process of developing a digital imprint. Part of the grant money is explicitly earmarked to support and document experimentation with various business models for web publishing. Ideally, the Press, working with scholarly editors, will find a way to continue to deliver material free of charge while at the same time charging enough for particular kinds of services—use of robust search engines, for example—to sustain projects. Collectively, we need to find a way to succeed. Otherwise, the model of no-cost consumption could have the hidden cost of severely restricting what it is possible to accomplish in terms of large-scale electronic scholarly editing.

Notes:

1. The Walt Whitman Hypertext Archive, ed. Ed Folsom and Kenneth M. Price, http://www.whitmanarchive.org/archive1/index.html. [back]

2. A notable feature of digital editing is the possibility for the exchange, migration, and improvement of texts. Swaps are possible even across the commercial/noncommercial line; our experience shows that scholars can even request direct donations with some hope of success as long as the cost to the commercial entity is low. [back]

3. The Classroom Electric: Dickinson, Whitman, and American Culture, ed. Kenneth M. Price, Martha Nell Smith, et al. http://www.classroomelectric.org/. [back]

4. Handbook for Digital Projects: A Management Tool for Preservation and Access, ed. Maxine K. Sitts (Andover, Massachusetts: Northeast Document Conservation Center, 2000), p. 8. Also available online at http://www.nedcc.org/assets/media/documents/dman.pdf. [back]

5. The price of digital images varies markedly depending on the nature of the institution, the special care that may be required for rare materials, the cost—if needed—of a professional photographer, etc. Typically we encounter prices ranging from $15 to $50 per high-resolution scan. [back]

6. John Unsworth of IATH first formulated the idea for this undertaking. Katherine Walter, Chair of Archives & Special Collections, University of Nebraska-Lincoln Libraries and Daniel Pitti of IATH have helped refine how this idea might be implemented. [back]

Comments?

Published Works | In Whitman's Hand | Life & Letters | Commentary | Resources | Pictures & Sound

Support the Archive | About the Archive

Distributed under a Creative Commons License. Matt Cohen, Ed Folsom, & Kenneth M. Price, editors.