Copyright (c) 1999 by The Chronicle of Higher Education, Inc. Republished with permission in The Journal of Electronic Publishing. This article may not be published, reposted, or redistributed without express permission from The Chronicle. To obtain such permission, please send a message to permission@chronicle.com. For subscription information, send a message to more-today@chronicle.com.

The Chicago Journal of Theoretical Computer Science sounded like a sure-fire winner when it was first published by M.I.T. Press four years ago. It was the first peer-reviewed journal in theoretical computer science to be published only in an electronic format. It is aimed at a group of scholars who are highly wired, and it is backed by hefty names in academe: In addition to being published by MIT, the journal is edited by a prominent computer scientist at the University of Chicago.

But MIT's computer-science periodical, like many other journals that are published only electronically, has turned out to be an also-ran. Few scholars have been willing to submit articles for publication, and circulation is far below expectations, with only a "minuscule number" of scholars willing to pay $30 for an annual subscription, says Janet Fisher, associate director for journals publishing at MIT Press.

The press has four other electronic-only journals, in fields as disparate as neurology and economics. None has done well at attracting papers from scholars. "Maybe they really need to be something different, and we haven't found the formula yet," says Ms. Fisher.

MIT Press isn't the only publisher of electronic journals facing such woes. Earth Interactions, an earth-sciences journal published by three scientific societies, has received only about twelve papers annually since it began in 1997, compared with the thirty or so per year that had been expected. After subjecting the submissions to peer review, the journal ended up publishing only nine papers in 1997 and 1998 combined. "We are finding it harder than we expected to get submissions," says Keith Seitter, deputy executive director of the American Meteorological Society, which publishes the journal together with the American Geophysical Union and the Association of American Geographers.

No one knows for sure why some electronic journals are having a rough time getting papers and subscriptions from scholars, but several theories are being bandied about by editors, publishers, librarians, and scholars. Scholars are worried, according to the speculation, that electronic publication will not carry much credit toward tenure, or that the electronic journals might fail, carrying prized papers with them into oblivion. Many electronic-only journals publish papers individually, as they are accepted, rather than on a regular schedule.

Such explanations are challenged, however, by the success of a few electronic-only journals, such as the Journal of High Energy Physics and the New Journal of Physics.

"It's clear that some electronic journals are truly struggling to get submissions," says Kenneth L. Frazier, director of the general library system at the University of Wisconsin at Madison. But a full explanation may be elusive, he says, because of the unpredictable character of scholarly publishing as a whole. "No one knows what makes a book sell, either," he observes.

A few editors note that launching any journal — electronic or paper — is a difficult task, because an ever-increasing number of journals are competing for submissions. "I don't believe that being electronic makes a lot of difference in the success of a journal," says Daniele Amati, project chairman for the Journal of High Energy Physics and director of high-energy physics at the International School for Advanced Studies, in Trieste, Italy. Researchers have submitted 592 papers to that journal since it was started in July 1997. Of those, 368 were accepted.

Some scholars worry that electronic journals are likely to be less permanent than printed journals. "There's a lot more sense of experimentation with the Internet. Things come and go," says Mr. Seitter, of the meteorological society. Consequently, scholars may believe that electronic journals are a riskier venue for their work.

"We simply need time to demonstrate that we're permanent," says Cliff McKnight, who is editor in chief of the Journal of Digital Information [formerly http://jodi.ecs.soton.ac.uk/] and a professor of information studies at Loughborough University, in Britain. "It's hard to demonstrate permanence overnight."

Publishing a printed version of an electronic journal seems to be the easiest way to assuage scholars' fears. Ms. Fisher says MIT Press commissioned a survey of computer scientists last year to find out why they were not submitting articles to the Chicago Journal of Theoretical Computer Science. Two-thirds of those who answered said they would be more likely to submit papers to an electronic journal if it also had a print version.

Bruce Mizrach, an associate professor of economics at Rutgers University and editor of MIT Press's Studies in Nonlinear Dynamics and Econometrics, says his electronic journal may get a printed version for precisely that reason. On paper, it would be an "inferior product," he says, because it would lack such online features as hyperlinks and multimedia.

He initially believed that an electronic journal could exist on its own. But time has shown, he says, that a printed version may be essential to widespread acceptance. "At this juncture," Mr. Mizrach says, "I'm willing to not be so stridently ideological about it."

Mr. Frazier, the Madison librarian, finds that a wise attitude. "If I were starting a new electronic journal, I would make sure that there is some sort of print product," he says.

Although younger researchers probably are more adept in multimedia and other technologies that make online journals useful, they do not seem to be flocking to electronic journals the way publishers had expected.

In fact, some senior scholars in geophysics have actively discouraged younger faculty members from submitting their papers to Earth Interactions, on the grounds that their applications for tenure would be stronger on the basis of publication in print journals that were better known, says Judy A. Holoviak, director of publications at the American Geophysical Union.

Similarly, MIT Press's survey of computer scientists revealed that many would not publish their own papers in the Chicago Journal, nor would they encourage their students to do so. "They thought it would not count for tenure in their departments," says Ms. Fisher.

One roadblock to developing the academic reputation needed to attract papers is the limited circulation of most of the electronic journals, according to publishers and editors. The most successful of MIT Press's five electronic-only publications is Mr. Mizrach's economics journal. It has almost 200 individual subscribers, but few university libraries subscribe, says Ms. Fisher. A printed journal published by MIT would have between 300 and 500 subscribers after two to three years of publishing, she estimates.

In contrast, the Journal of Digital Information has "several thousand" registered users. Anyone with World-Wide Web access can register to read it free; its operating costs are underwritten, for now, by the British Computer Society and Oxford University Press. But the journal soon will approach its readers to explore the feasibility of charging for subscriptions, says Mr. McKnight, the editor.

Readers already are on notice: "JoDI will not always be free," warns its Web site. "Development costs money, and it is our aim to provide you with the best journal in its field."

The New Journal of Physics is taking a different tack. Rather than charging readers, it requires authors to pay $500 per article. The charge is the same regardless of the paper's length or use of color illustrations or multimedia tools. The journal's contents are freely available to anyone.

Started last December by the Institute of Physics, in Britain, and the German Physical Society, the journal receives ten to fifteen submissions each month, says Robert Brown, journals director at the British institute. So far, seven papers have been published, and users have downloaded more than 7,000 copies, which he says represents a greater readership than is typical of printed journals. "Do we ever get 1,000 people going to a paper journal and reading it?"

The editors of some electronic journals say one way of encouraging submissions is to forge ties with on-line archives of papers that have not yet been peer-reviewed. Such archives have long been popular in physics and are becoming common in other disciplines. The National Institutes of Health, for example, is discussing the creation of an online archive for papers in biomedical science.

The Journal of High Energy Physics is one publication that works closely with an archive. Its Web site includes a "mirror," or complete copy, of a popular physics electronic archive located at Los Alamos National Laboratory, in New Mexico. The author of a paper in that e-print archive can submit it for publication in the journal by filling out a form on the Web and supplying the paper's identification number in the archive. The "overwhelming majority" of the papers published by the journal have come from the archive, says Mr. Amati, the journal's project chairman.

Marketing also plays a role in attracting both submissions and readers. "You've got to get out there and advertise it and sell it," says Mr. Brown, of the British physics institute. The institute has bought advertisements in physics periodicals and has sponsored receptions at conferences of physicists to promote the New Journal of Physics.

Journals that are not backed by a publisher may not have the resources to beat the bushes for submissions. For example, the Journal of Interactive Media in Education, an electronic-only publication based at the Open University, in Britain, received about a dozen submissions last year, says Simon Buckingham Shum, one of the journal's editors and a researcher at the university's Knowledge Media Institute. But he and the rest of the journal's editors, all of whom are volunteers, haven't had the time this year to solicit manuscripts aggressively. Without a publisher to maintain the journal's visibility, submissions to it have dropped off, he says. "When we get busy, the journal suffers."

Another crucial factor is whether an electronic journal is included in the indexes that are popular with academics in a given field. If the journal is not indexed, scholars may never find papers it has published that are relevant to their research, and other researchers will not be interested in submitting work to it.

Particularly important is coverage by the Institute for Scientific Information, a Philadelphia-based company that indexes about 8,000 journals it deems to be the most important. It produces electronic databases and printed volumes indicating which papers have been cited by other authors, and it compiles statistics that are meant to reflect the impact of various journals in a field. The databases also are used to produce statistics that estimate the influence of academic departments at universities.

"I.S.I. is the one that we feel we really need to capture," says Mr. Seitter, of the meteorological society. "Until Earth Interactions shows up there, it's going to be hard for authors to get the kind of credit they need for their publications."

The company monitors fifteen journals that are disseminated only in electronic format, says a spokeswoman, Jacqueline H. Trolley. They include two from MIT Press — the Chicago Journal of Theoretical Computer Science and Studies in Nonlinear Dynamics and Econometrics — and journals from a range of other fields, including New Astronomy, Postmodern Culture, and Sociological Research Online.

As it does when judging whether to include a printed journal in its database, the company looks for evidence of high academic quality in electronic journals that it is considering, Ms. Trolley says. However, it has adapted some of its criteria: Rather than insisting that an online journal publish regular issues, the company will cover an electronic publication if it disseminates new material at least once every six months.

Other indexes also cover electronic journals. For example, the Chemical Abstracts Service, operated by the American Chemical Society, follows thirty electronic-only journals.

Ms. Fisher, of MIT Press, says it would be very helpful for its electronic-only Journal of Contemporary Neurology to be covered by Index Medicus, an index of biomedical journals that is produced by the National Library of Medicine, in Bethesda, Md., and distributed as an electronic database called Medline. But neither the index nor the database will cover an electronic journal until the journal has published at least twenty papers.

Without being indexed, says Ms. Fisher, attracting twenty papers "is like pulling teeth."


LINKS FROM THIS ARTICLE

Chemical Abstracts Service http://info.cas.org/

Chicago Journal of Theoretical Computer Science http://www.cs.uchicago.edu/research/publications/cjtcs/

The Chronicle of Higher Education http://www.chronicle.com

Earth Interactions http://earthinteractions.org

Journal of Contemporary Neurology http://www.pslgroup.com/dg/1f8e2.htm

Journal of Digital Information [formerly http://jodi.ecs.soton.ac.uk/]

Journal of High Energy Physics http://jhep.mse.jhu.edu

Journal of Interactive Media in Education http://www-jime.open.ac.uk

National Institutes of Health discussion of an online archive for papers in biomedical science http://www.nih.gov/about/director/pubmedcentral/pubmedcentral.htm

New Astronomy http://www.elsevier.com/wps/find/journaldescription.cws_home/601274/description#description

New Journal of Physics http://njp.org

Postmodern Culture http://jefferson.village.virginia.edu/pmc/

Sociological Research Online http://www.socresonline.org.uk/socresonline/

Studies in Nonlinear Dynamics and Econometrics http://www.bepress.com/snde/