The presentation of magazines, books, and journals in digital formats is accelerating as an alternative to conventional print-on-paper media. More and more publishers are embracing the electronic publication concept - often from a fear of being left behind in the technological dust more than from a desire to meet market demand, which for the most part remains nonexistent.
Most scientists continue for now to rely on tried-and-true print media. However, in the future, electronic publishing could have a major impact on the way scientists acquire and disseminate information.
Today, an increasing number of on-line scientific journals and magazines are being introduced, often on an experimental basis. And a growing number of sophisticated electronic library systems, each containing a wide range of text- and graphics-based information resources, are being developed.
But as publishers get swept up in this wave of new technology, they are simultaneously troubled about how to replace diminishing revenues from traditional print publications and protect intellectual property when it is available in an easily copied electronic form.
Established publishers are also facing a growing challenge from scientists who have discovered that scientific information can be disseminated fairly easily and economically using electronic print (e-print) archives. In some fields, particularly physics, the e-print concept has proven to be compelling in its simplicity and immediacy, and growth in use of such systems has been exponential.
It remains to be seen whether conventional scientific publishers will adopt the e-print idea - or possibly get steam-rollered by this new form of grassroots publishing. Publishers face further challenges from document delivery services, which undercut the need to subscribe to journals.
A major advantage of electronic publishing over conventional printing is elimination of the costs of printing, paper, and mailing entailed in the production and distribution of print journals. In addition, electronic publishing makes it "easier to search through masses of documents to find what you want," says John Hearty, director of reference services marketing for OCLC, the Online Computer Library Center. The center, based in Dublin, Ohio, is a nonprofit organization that works with publishers to develop electronic journals.
"It's possible to get regular updates on topics you're interested in," says Hearty. "You can develop your own journal on demand across all the publications a library subscribes to. Information is delivered directly to your workstation, so there's a convenience factor. And data can be pulled out and manipulated in spreadsheets." Such advanced capabilities are not yet available with many existing electronic publications, but they could become realities in the future.
In the chemical sciences and related fields, the American Chemical Society, headquartered in Washington, D.C., is taking a leading role in electronic information delivery. ACS Director of Publications Robert H. Marks believes that demand for the society's electronic publications will increase over the next few years as demand for traditional print products declines.
All 23 ACS research journals are currently available as the CJACS Plus file on the STN International on-line system. Users can view scanned page images from issues of the journals going back to 1992, including all graphics and tables. The full text of articles (without graphics and tables) is also searchable from 1982 to date for all journals. Page images can be downloaded to a user's computer and then printed out.
For the first time this year, ACS journals will not have to be laboriously scanned page by page into the system. Instead, they will be created in fully digitized form and published directly in an electronic format.
Page images of ACS journal articles from 1992 forward are also now accessible with SciFinder, a graphical user interface that facilitates nonexpert searching of databases created by ACS and by the Chemical Abstracts Service (CAS), a division of ACS located in Columbus, Ohio. SciFinder does not yet provide for searching and browsing of the full-text contents of individual articles, but it may be upgraded to do so.
ACS also is investigating the feasibility of making annual subscriptions to ACS journals and magazines available through the SciFinder arrangement. SciFinder-based subscriptions would provide a fixed-price alternative to CJACS Plus, users of which are charged fees based on connect time, the number of answers displayed, and other factors.
CD-ROM (compact disc read-only memory) versions of the Journal of the American Chemical Society and Biochemistry are also now available, and ACS hopes to add more journals to this series, called ACS Publications on Disc.
CD-ROM-based chemistry journals are also available from other publishers. Current CD-ROM offerings in chemistry include the European Journal of Biochemistry (published by Springer-Verlag, Berlin) and the Journal of Biological Chemistry (published by the American Society for Biochemistry & Molecular Biology, Bethesda, Md.).
"So far the response to our CD-ROMs from ACS members has been very good - much better than anticipated," says Marks. "But we're a little disappointed with the library response." Hundreds of individual subscriptions have been sold, but only a handful of the more expensive library subscriptions, he says.
"Librarians, and maybe even publishers, don't quite know what to make of journal CD-ROM products," comments Gary Wiggins, head of the chemistry library at Indiana University, Bloomington.
Electronic scientific journals from several publishers are being distributed by OCLC, including the Online Journal of Current Clinical Trials, launched by the American Association for the Advancement of Science, which is headquartered in Washington, D.C. The journal generated insufficient contributor and subscriber interest, in part because there was no regular print version, prompting its sale to the current publisher, Chapman & Hall, London.
OCLC also distributes on-line versions of Applied Physics Letters (published by the American Institute of Physics, College Park, Md.), Electronic Letters (published by the Institution of Electrical Engineers, London), and Immunology Today (published by Elsevier Science, New York City). OCLC has plans to make available later this year on-line versions of 30 journals from the Current Opinion series (published by Current Science, Philadelphia), Physical Review Letters (published by the American Physical Society, College Park, Md.), and publications in Elsevier's Trends series. Print versions of these publications will continue to exist, with on-line presentation viewed as an alternative delivery mechanism.
People view and search OCLC-distributed on-line journals with Guidon, an interface that runs under Windows or in conjunction with the World Wide Web access program, Mosaic. A Macintosh version of Guidon will be introduced later this year. Guidon supports full-text searching by subject, title, author, keyword, and date, and searches can be structured with a range of Boolean and proximity operators.
Guidon can handle page images that have been digitized by electronic scanning of published pages. However, it also handles digital documents encoded with Standard Generalized Markup Language (SGML). A standardized code for identifying different elements in an electronic document, SGML has become the preferred format for electronic publishing.
From a user's viewpoint, a key advantage of electronic publishing "is going to be getting all your information on a single workstation," says Hearty. One of OCLC's goals for the near future is to integrate different types of information - primary literature (journals), secondary services (bibliographic and abstracting data), and tertiary information (such as directories and encyclopedias) - so a user can skip back and forth between them, all under a graphical user interface.
This type of capability is already available in part to STN users, who can, for example, use SciFinder to search for secondary information (such as abstracts) on the CA Registry File and then view a paper corresponding to a specific abstract on the CJACS Plus file.
Future information integration functions of OCLC on-line services will also include "advertising capabilities - hot links to ads," says Hearty. "If you found a piece of equipment mentioned in an experimental procedure, you could push a button to access documentation, specifications, or advertising on that equipment. You could also request further information from the publisher or advertiser. But the advertising would be unobtrusive - if you didn't want to look at it you wouldn't have to." Advertising links are currently available in the on-line version of Immunology Today.
The electronic publishing concept is also being studied at American Telephone & Telegraph Corp. AT&T's Bell Laboratories unit in Murray Hill, N.J., has developed RightPages, an electronic library interface program that is currently undergoing three major trials.
The first of these is at Bell Labs, where 350 users at 25 sites can use RightPages to access more than 120 titles from 41 publishers, primarily in the computer science, electrical engineering, and business fields. A second experiment, called Red Sage, allows 330 users at the University of California, San Francisco, and other local sites to see on-line versions of 70 publications from 19 publishers, covering radiology, molecular biology, and other topics. And 135 users at several Bristol-Myers Squibb sites can view 13 titles from eight publishers on molecular biology, biochemistry, and other topics.
According to RightPages manager Melia Hoffman of AT&T, the RightPages system "has been using scanned page images for the bulk of its input." This is time-consuming because published journals must be laboriously entered into the system page by page, using scanning equipment that digitizes the page images.
However, RightPages personnel "have recently started testing full electronic production and delivery," says Hoffman, "thus allowing publishers to create a high-quality electronic product with minimal extra work beyond traditional paper processes."
Another electronic library project is TULIP (the University Licensing Program), a cooperative effort by Elsevier and university libraries to study the feasibility of networked distribution of scientific journals.
TULIP is a database of scanned images of 43 Elsevier and Pergamon Press (Tarrytown, N.Y.) journals in materials science and engineering. Scientists at 17 participating universities, including nine University of California campuses, can access these journals over local area networks and make printouts.
TULIP implementations differ at the various sites. In some cases, system interfaces are search-oriented, permitting users to search for authors, titles, keywords, index terms, and (in some cases) to search the full text of articles. In other cases, the interfaces enable users to browse publications - by choosing a journal title from a list, selecting a specific issue, viewing the table of contents, and selecting a specific article.
A further variation is that some users are able to see screen displays of page images, whereas others must print out articles in order to view them. A major goal of the TULIP experiment is to determine which of these format variations work best for users.
Formally speaking, the TULIP experiment is going to conclude at the end of this year. However, Elsevier plans to continue its electronic publishing efforts. According to TULIP project manager Jaco Zijlstra of Elsevier, the company intends to provide scanned page images of over 1,000 Elsevier journals to a selected group of customers on an experimental basis. In the long term, Elsevier will move toward providing electronic journals in a fully electronic (SGML-encoded) format instead of as scanned page images.
The Institute for Scientific Information (ISI), Philadelphia, and IBM's Almaden Research Center, San Jose, Calif., are collaborating to develop a prototype electronic library system for initial implementation by August 1995. The system will provide users with desktop access over local area networks to bibliographic data, abstracts, and table-of-contents data from 1,350 scientific journals contained in the Life Sciences edition of Current Contents, an ISI publication. In addition, page images will be accessible for journals to which the pilot site subscribes (and for which publishers have agreed to allow electronic delivery).
The system will have a client-server architecture - a setup that allows users to access it from personal computers running most operating systems, such as DOS (and Windows), Macintosh, and UNIX - and a user interface based on Lotus Notes. Pilot sites for the project include Brookhaven National Laboratory, Upton, N.Y.; Lehigh University, Bethlehem, Pa.; the New York City Public Library; Purdue University, West Lafayette, Ind.; University College, London; and Glaxo Research & Development Ltd., Hertfordshire, England. The sites operate independently - that is, people at one site cannot access journals at other sites.
The objective of the ISI-IBM project is to test the many variables relating to electronic distribution of information and evaluate new technologies for use in future digital libraries. Areas to be explored include protection of intellectual property rights; billing, accounting, and business management of electronic libraries; pricing scenarios; and patterns of use of electronic journals.
In addition to such nongovernment efforts, a joint initiative of the National Science Foundation, the Department of Defense's Advanced Research Projects Agency, and the National Aeronautics & Space Administration is funding six digital library research projects. The projects are designed to advance the technology for collecting, storing, and organizing digital information and making it available over networks.
One of these projects, centered at the University of Illinois, Urbana-Champaign, will focus on journals and magazines in the science and engineering literature. Mosaic will provide users at several universities with access to the publications.
Many other scientific electronic publishing ventures exist. For example, an electronic version of Protein Science, published by the Protein Society, Seattle, will soon be available on World Wide Web. According to Protein Science electronic publishing coordinator Stephen H. White and Editor-in-Chief Hans Neurath, "The electronic library will, if done right, greatly simplify our lives as working scientists. This is not to say that the electronic journal will entirely replace the printed journal which, as far as we know today, will always have a place in the distribution and storage of scientific information."
Electronic publishing can be so inexpensive that some scientists and librarians are already beginning to view traditional publishers as obsolete middlemen. For example, physicists can now transmit scientific papers directly to e-print systems - storage and distribution services that make the papers accessible to other scientists - thus bypassing publishers and the traditional peer review process.
The father of the e-print concept is theoretical physicist Paul H. Ginsparg of Los Alamos National Laboratory (LANL), founder and operator of electronic archives of nonreviewed preprints of physics, nonlinear dynamics, mathematics, and economics papers. The archives are electronic versions of an earlier system in which printed copies of physics preprints (papers that had already been submitted to journals) were mailed to a list of subscribers.
The first e-print system went on-line in August 1991, covering theoretical high-energy physics and serving 160 users. From that modest start, e-print archives have experienced explosive growth and are now some of the largest and most active databases on Internet - with more than 25,000 active users worldwide and more than 45,000 electronic transactions per day.
Archives are relatively inexpensive to set up, and no fees are charged for use. However, publishers point out that federal government support of the people and computer equipment needed to operate e-print archives helps make the systems inexpensive.
E-print archives currently cover 20 topical areas in economics, 15 physics subdisciplines, seven areas of mathematics, five nonlinear dynamics topics, the field of atmospheric and oceanic sciences, and the area of computation and linguistics. For those with access to World Wide Web, the home page address for LANL's e-print archives is http://xxx.lanl.gov/.
The most chemistry-oriented file in the physics group is the chemical physics archive, which is co-organized by the theoretical chemistry and molecular physics group at LANL and the department of chemistry at Brown University.
In a recent article in Los Alamos Science, Ginsparg explained that the e-print concept "began as an experimental means of circumventing recognized inadequacies of research journals, but unexpectedly became within a very short period the primary means of communicating ongoing research information."
According to Ginsparg, e-print systems are "entirely automated - including the submission process and indexing of titles, authors, and abstracts - and allow access via e-mail, anonymous FTP [file transfer protocol], and World Wide Web." The advantages are that "the communication of research results occurs on a dramatically accelerated time scale and much of the waste of the hard-copy distribution scheme is eliminated. In addition, researchers who might not ordinarily communicate with one another can quickly set up a virtual meeting ground and ultimately disband if things do not pan out - all with infinitely greater ease and flexibility than is provided by current publication media."
Ginsparg writes that "members of our community have ... learned to determine from the title and abstract (and occasionally the authors) whether we wish to read a paper as well as to verify results ourselves, rather than rely on the alleged verification of overworked or otherwise careless referees. The small amount of filtering provided by refereed journals plays no effective role in our research."
The point is, if electronic distribution is so simple and peer review superfluous, who needs the traditional scientific publisher?
"What role will be played by publishing companies [in electronic research communication], and how large will their profits be?" asks Ginsparg rhetorically. "Can publishing companies provide more value than an unmanned automated system whose primary virtue is instant retransmission?" In the long term, he says, "it is difficult to imagine how the current model of funding publishing companies through research libraries (in turn funded by overhead on research grants) can possibly persist."
Not so fast, says ACS's Marks: "Research isn't completed until the results have been peer reviewed and published. An author might have a good reputation, but many authors with good reputations have submitted papers of questionable quality."
Ginsparg tells C&EN, "It is clear that electronic systems can easily accommodate peer review, and undoubtedly will. I am not opposed in the least to the principle of peer review - I simply object to the inept way it is currently implemented in my discipline and [have] argued, and moreover demonstrated, that it is so badly handled that we lose nothing by abandoning it. This undoubtedly differs from field to field."
In physics, he says, "we have a reader community that essentially coincides with the author community, [whereas] in chemistry I'm told that there is a readership - in industrial labs, et cetera - that vastly outnumbers the authoring community. Hence, some of the economic lessons in my community may not carry over so directly."
But at least in physics, Ginsparg says, he would prefer that "the current onetime all-or-nothing [system], as implemented in the print format" be replaced by a much more flexible system of review. He has been considering establishing an alternative form of peer review on e-print archives, but he can't yet say exactly what will emerge.
Asked about prospects for a broader extension of e-print archives in chemistry, Jim Doll of the department of chemistry at Brown University, who helped develop the archive in chemical physics, says e-print system use "is most natural at the moment in areas where there are relatively common standards for document preparation." In physics (and chemical physics), the commercial scientific word processing program TeX is sufficiently commonplace that it offers a convenient standard, he says. "This tends not to be the case at the moment in all areas of chemistry."
According to Doll, "It is likely that emerging 'portable document format' software will soon bring these electronic publication activities into much broader areas of chemistry. Such software will make documents portable across a broad range of platforms. ... When this software becomes a reality, the use of electronic archives will likely expand significantly."
Libraries have traditionally provided the largest share of subscription revenue for scientific books and journals. But all bets are off if electronic publishing continues to gain in popularity.
For example, document delivery services now permit librarians and library users to quickly get single copies of articles by computer downloading, fax, or mail. Hence, people can now scan contents pages of journals and buy scientific information by the article, eliminating the need to subscribe to the journals.
Document delivery services are deeply troubling to publishers. "All we get is a copyright clearance fee, and journal articles that aren't accessed don't generate any revenue," says Marks. In effect, the existence of single-article services means that "there are a lot of people republishing our material. One of the things I've thought about is withdrawing permission to copy for resale from the Copyright Clearance Center [Danvers, Mass.] and saying that these services can't reprint and sell our material."
On the other hand, Ginsparg contends that "many researchers in the future will refuse to sign copyright agreements that assign sole distribution rights to the journal's publisher. In fact, as a national lab employee, I have never been permitted to sign any copyright agreement without attaching a disclaimer that the U.S. government retains rights to distribute in any way it pleases."
Ginsparg adds, "Certainly researchers and their funding institutions should insist on sharing copyright and retaining distribution rights to material in the format produced by authors. ... The person who has really been articulating this for years and deserves the credit is Ann Okerson." Okerson is director of the Office of Scientific & Academic Publishing at the Association of Research Libraries, Washington, D.C.
From Marks's point of view, the problem with single-article ordering is that "a journal publishes everything submitted that is deemed top-quality science, whether or not anyone ever even reads it. In effect, an archive of peer-reviewed research is created. Subscribers to the printed journal pay for everything that's there, and the cost is fairly reasonable because it's distributed among a large number of subscribers. If you have everything available electronically and people only pay for what they print out, we won't be able to support this system."
Marks adds, "We have to figure out how to make the transition from the printed product to the electronic product. During this period, we have to generate enough revenue from both sources to keep the program going. That's really the challenge of migrating to the electronic age."
[ACS Home Page] [ACS Publications Division Page]