Library Issues

Editor: Richard M. Dougherty, University of Michigan.
Contributing Editors: Steve Marquardt, South Dakota State University; William Miller, Florida Atlantic University; Sarah Pritchard, Smith College; Mark Sandler, University of Michigan

Vol. 17, No. 6 July 1997

Scientific Journals Online: Five Emerging Models

by Mignon Adams

While the scientific journal has been the primary means of communicating the results of scientific research for over 300 years, the subscription prices of these journals began increasing precipitously in the late 1970s. At the time academic libraries started their almost-annual practice of cutting—sometimes slashing—journal lists. Prices continue to rise—50 percent just since 1993,1 making it more and more difficult for college and university libraries to maintain the journal collections their users demand.

Campus researchers often blame the library for smaller journal collections, but over the past five years they have expressed hope that "when all this gets electronic," or "when everything's on the Internet," prices will be lower or even disappear. Now everything is electronic and is migrating to the Internet—but most is not free.

As publishers of scientific journals are moving uneasily through a transitional period, the models currently emerging indicate that costs will remain the same. Difficult decisions lie ahead for academic librarians, as they determine how much of their collection will be electronic, how much will be print, and how they will maintain retrospective access.

Five Emerging Models
In the early nineties publishers began their first forays into fulltext online, using CD-ROMs and e-mail. Elsevier's Adonis service offered access to the fulltext of over 1,000 journals, using a single search engine and a number of CD-ROMs. Adonis was never widely adopted, probably because of the need to juggle over a dozen CD-ROMs and because users were limited to one workstation.

Another early effort, by OCLC, was to provide two journals by e-mail. Editors of these journals completed the peer-review process within 24 hours and the articles were instantly available to subscribers. Technical problems and confusion over pricing did these in. Other e-journals exist, but none have gained the prestige of the peer-reviewed print journals. Clearly, other models were needed. The adoption of Worldwide Web technology provided the means.

Table-of-Contents/Document Delivery. Developed pre-Web, but now using its technology, the CARL company's Uncover and Reveal programs provide researchers with a table-of -contents service coupled with faxed documents delivery.

Via the Web, a user enters journal titles of his or her own choosing. As CARL enters those journals into its database of 17,000 journals, the tables of contents are automatically sent via e-mail to the user. He or she may then request any of the articles to be faxed. CARL hopes by the end of 1997 to have publishers' agreements for electronic transmission.

A number of colleges and universities have chosen to subsidize an electronic table-of-contents service linked with document delivery for their users. At first, this may not seem to make economic sense. Articles obtained through the service run from about $12-$20 a piece (a copyright fee to the publisher is made). However, when comparing the costs of a $5,000 subscription for a journal that may be of interest to only three or four researchers with the cost of articles ordered by those researchers, the proposition begins to look better. Institutions report such scenarios as cancelling $100,000 worth of subscriptions and replacing them with an annual costs of $20,000 for delivered articles.

However, the library does not know at the beginning of the year what the total costs will be. And faxed charts and illustrations are often of inferior quality or even indecipherable.

This model concentrates on the article as the unit of purchase. The article still comes bundled with the journal issue (in the table of contents), but after that, the journal issue itself becomes inconsequential, and is no longer an economic unit.

Single Journals on the Web. A number of scientific journals have developed their own websites. A few are available at no cost, while others have a subscription fee or charge for viewing or downloading an article. For now, at least, journals that move on to the Web are keeping their identity: individual issues with a table-of-contents and articles which correspond to the printed issue. In the future, journals may well print articles as they are received, or add sound and video clips.

Stanford University's HighWire Press has been in the forefront of this model, mounting so far 14 scientific journals, including Science, published by scholarly societies.2 Each journal is a stand-alone product, with no common search engine. (One is planned soon.) Most charge, plan to charge, or charge for sections of the journal, but subscriptions must be negotiated with each individual journal. For the researcher for whom the Journal of Biological Chemistry is an essential publication, then direct access to the latest and archived issues at his or her workstation may meet many needs for fast and current information.

HighWire Press states as its mission to return scholarly publishing to its roots, where the reason for publication was to advance scientific knowledge, not to make a profit. However, within the past year, commercial publishers have begun to develop still yet another profit-making model.

Publishers' Collections on the Web. Mellon Foundation funding made it possible for the Johns Hopkins Press in 1995 to mount the 40 journals it publishes in fulltext through the Web, with one search engine and full graphics. Called Project Muse, the collection includes only one scientific journal, but it served as a prototype for other publishers.

In 1996 Academic Press announced a similar effort, called IDEAL, which makes its 175 journals available in fulltext through its Website, with a single search engine. The pricing scheme is complicated, based on a percentage of the current price of journals subscribed to by a library in a previous year (thus negating the effects of subsequent cancellations). Most participating libraries have chosen to join through consortia, thus able to pool their subscriptions and allow mutual access to all or nearly all the 175 journals. An advantage is that members of the consortia with no subscriptions to Academic Press journals can participate at a minimal fee (usually $1,000).

Academic Press publishes journals in many fields and many scientific disciplines. However, while their journals are highly respected, with a few exceptions they are not the stars of their fields, not the ones that faculty members are clamoring for.

Following these leads, other publishers are announcing plans to mount their fulltext journals with a single search engine. Blackwell and Elsevier plan to have theirs ready this summer, and several other university presses (MIT and Berkeley) have projects in progress. Different pricing stratagems are being discussed, ranging from a flat fee in advance for all, selected publications, or pay-per-article.

A major advantage of this model for a library is that it may be able to greatly increase the number of journals available to its users with a small additional investment, particularly if it participates with a consortium. However, searching by publisher is an artificial construct at best. Some librarians argue that accepting an entire publishers' output means receiving the worst as well as the best. Marginal publications may continue that might not otherwise survive in the marketplace.

In this model, the publishers' collection becomes the economic unit. Not only does the journal issue become irrelevant, but the journal itself may do so.

For more information you may access these sites:

Academic Press
CARL Uncover
HighWire Press
Project Muse

Index/Abstract with Links to Fulltext. Other vendors are developing products which include journals from many publishers. Ebscohost, Infotrac, and Proquest (with Wison soon to come) all offer web-based indexes with links to fulltext periodicals. While each vendor offers hundreds of fulltext publications, the percentage of those which are scientific journals is small. These products do an admirable job of supporting undergraduate needs, but are unlikely to be of much interest to the scientific researcher. Of much more use would be a scientific indexing or abstracting service with links, but only a few have been developed.

Ovid offers fulltext links from Medline. Only 60 of Medline's 3,800 journals are available, but they are among the most highly-regarded clinical journals. The product has been well received by medical libraries, but the cost is quite high; there is a considerable timelag between publication and the appearance of the fulltext in the database; and Ovid has struggled with a number of technical problems.

Another Medline effort, currently free in its prototype, is the National Center for Biotechnology Information's PubMed, which asks publishers to link their web-based journals to their Medline search engine. So far 17 mostly biology journals are participating, 13 of them among HighWire Press's collection.

Search Engine for Library Subscriptions. The most recent model is closer to the traditional model of receiving indexes and journals separately. In this model, a vendor provides a search engine with links to fulltext journals, and the library separately subscribes to the electronic journals it chooses. In some cases, the costs of an electronic version of a journal is a bit less than the print, or an electronic subscription can be added to a print one for a modest cost.

Both OCLC and Ovid expect to have such services in effect during the summer of 1997. OCLC also promises to archive the journals it handles.

Which will Prevail? The five above models indicate a variety of ways to approach the provision of electronic fulltext journal articles. Surely not all five will continue to be available five years from now—but it's certainly not clear which model will prevail. As software developers have demonstrated, it's not necessarily the best product which wins the marketplace. In the meantime, both users and libraries will need to carefully wend their way through the transitional period.

The Users
From 1991 to 1995, Elsevier conducted an experiment, nicknamed TULIP, with nine American universities. The fulltext of 43 science and engineering journals was mounted on campus computers and made available through campus networks. Usage and reactions were less favorable than had been expected. There were unforeseen infrastructure problems, such as the large amount of storage required and the demands placed on the campus infrastructure.

In its final report,3 Elsevier identified the following qualities as essential to researchers' acceptance of electronic access:

According to these criteria, none of the products now available meet the user requirements that Elsevier determined are necessary for acceptance of fulltext access. Uncover's table-of-contents/document delivery perhaps comes closest. It is timely. It covers many journals for a period of up to seven years. Yet its faxed images are often of poor quality and of course there are not links to other articles since the final product is not in machine-readable form. Seeing an article almost immediately on a computer screen is far better than waiting 24 hours for a faxed article.

The publisher's collections presently available meet perhaps the fewest of the criteria. They cover many disciplines, none in depth. Most library users are barely aware of a journal's publisher, and certainly they do no ever conduct a search based upon one publishers' journals. A scientist whose research area is, say, "sea anemones," does not want to be told that she can look in Academic's IDEAL for the three journals that might have an article, and Elsevier's ScienceDirect for five others, and OCLC's for two more. While Elsevier's journals might will link to other Elsevier journals they are unlikely to link to Academic's.

Products that link fulltext to a specialized index or abstracting services are closest to what Elsevier determined users need. They provide access to a large portion of the information from one source. They can link from one article to another. They can have sufficient journal and time coverage. There's only one search engine to learn. Such products available now however have too small a proportion of the journals, and the linking lags too far behind publication.

Given the high interest in electronic fulltext, and the limitations of the products now on the market, academic librarians have difficult decisions to make over the next several years.

Implications for Libraries and Institutions
It is clear that if present models of electronic fulltext prevail, library costs of providing scientists access to their scholarly journal literature will not decrease. The cost of a single electronic subscription is now very close to a print one. Elsevier is offering those who make a three-year commitment to its locally-mounted Electronic Subscriptions an increase of "only" 9.5 percent for each of the last two years. In some cases, where a vendor has put publications from several different publishers together, the costs are significantly more than print versions.

Some publishers, however, are offering special consortial pricing, which will allow individual institutions to have access to a larger number of electronic publications for the same price as their print subscriptions cost. While their costs would not decrease, institutions could received more for their outlay.

Library will save money in other ways. An electronic journal does not need to be checked in, carried to a shelf, returned after use, or be bound. Missing pages do not need to be replaced. These savings, though, may well be offset by the costs of acquiring and maintaining equipment to view the journals.

In order to take full advantage of Web-based electronic fulltext databases, campuses must have a suitable infrastructure. All researchers should be directly connected to the Internet, with computer equipment capable of viewing and downloading image files. A standard printer may take as long as thirty minutes to print an article with extensive graphics.

In addition to costs, there are other considerations in adopting electronic fulltext. Scientists by and large are a conservative group, slow to change their ways. Some still cling to their printed abstracts, sure that electronic searching cannot yield better results than their well-honed manual methods. Libraries and institutions will need to promoted the products they buy and be prepared for a great deal of one-on-one training.

Archiving is a large concern. Scientists do need more than just access to current materials. Chemists have a particular need for older materials. Traditionally, libraries have been responsible for the preservation and storage of books and journals, as publishers merge, dissolve, or simply clear their warehouses. If materials are purchased only electronically, who will take on the responsibility of seeing that they are available to future researchers? Publishers say they will, but publishers come and go. Archiving is an important part of OCLC's proposed service, but this is a service not even in operation yet. Libraries cannot make a total commitment to electronic formats until the question of how older materials are accessed is resolved.

For the past decade, scientists and librarians have been awaiting the advent of online journals. Scientific journals are now moving rapidly to electronic access, primarily through the Web. Earlier dreams of researchers' taking back the dissemination of their findings will probably not be realized. Instead, commercial publishers are creating models that will maintain costs at previous levels. Even as librarians realize that electronic will no mean cheaper, they also know that sooner or later they will need to provide electronic fulltext.

Academic librarians across the country are considering options such as these:

Some librarians are adding electronic formats while at the same time keeping all print ones. Unsure of the future, they want to keep all future options open. While this is the most expensive approach, it's also the safest.

The marketplace and available choices today are very different than they were even a year ago. They are likely to be very different a year from now. The right decision might be easier to make a little later on.

Some librarians are trying to differentiated between those materials they feel are essential to their mission and those which can always be obtained elsewhere or in another form. Popular magazines, for example, might fall in this group, or an esoteric scientific journal.

A few librarians, feeling that the future is undoubtedly electronic, or perhaps extremely pinched for space, are moving very quickly to all-electronic formats. Supporting this view is the contention that if something better is available next year, then a different choice can be make at that time. All the emerging models use the same technology. So investment in equipment will not be lost—Mignon Adams is Director of Library Services and Information Technology at the Philadelphia College of Pharmacy and Science


1 Lee Ketcham and Kathleen Born, "Unsettled Times, Unsettled Prices: 37th Annual Report Periodical Price Survey 1997." Library Journal, April 15, 1997, pp. 42-47.

2Jeffrey R. Young, "HighWire Press Transforms Publication of Scientific Journals." Chronicle of Higher Education, May 16, 1997, page A21.

3The final report of the TULIP project is available on Elsevier's home page:

Mountainside Publishing, Inc.

Library Issues: Briefings for Faculty and Administrators (ISSN 0734-3035) is published bimonthly beginning September 1980 by Mountainside Publishing Co., Inc., 321 S. Main St., #300, Ann Arbor, MI 48104; (313) 662-3925. Library Issues, Vol. 17, no. 6. ©1997 by Mountainside Publishing Co., Inc. Subscriptions: $44/one year; $74/two years. Additional subscriptions to same address $18 each/year. Address all correspondence to Library Issues, P.O. Box 8330, Ann Arbor, MI 48107. (Fax: 313-662-4450; E-mail: Subscribers have permission to photocopy articles free of charge for distribution on their own campus.

last modified: