The key outcomes of the workshop consist of recommendations based upon the working group reports. Since the deliberations of the three working groups frequently overlapped, the recommendations below are not attributed to any single group.

  1. More systematic research and definition of terms are required to address the attributes and requirements for the scientific and technical information package of the future. Such requirements, in addition to the "package," include packaging that enables identification, indexing, searching, and retrieval of the information.

  2. It is essential to address the critical issue of the integrity of the scientific intellectual enterprise in the electronic-publishing world. The dynamic nature of the electronic product requires new processes be established for sustainability, certification, and ownership of scientific and technical information. A process of authentication of scholarly merit, equivalent in thoroughness and rigor to conventional peer review, must be established and maintained.

  3. Financial support for archival infrastructure development and continued maintenance must be secured. New archival arrangements or institutions must be considered, as the traditional libraries, private-sector publishers, and the national archive may not be equipped to handle this new environment. For example, archiving responsibilities, including effective location systems for all document versions, might be delegated by publishers and libraries to trusted third parties.

  4. Whatever arrangements are made, the new archival system, like traditional libraries, must support archival redundancy in order to avoid the loss of records that could otherwise be caused by "single-point" failure, and respect national and international standards. Short hardware and software "half-life" could be overcome with the establishment of a "meta-archive" in which stable archives of software, hardware, and applications programs could be preserved.

  5. Members of the scientific and technical communities who publish should work together to send a coherent message to the software developers regarding requirements and standards for authoring tools that specifically support scientific and technical electronic publications. Authoring tools should be intuitive, easy to use, and affordable, if not freely provided by the publishers. Such tools must support graphics, equations, and other specialized technical formats.

  6. The focus in development of electronic documents should be on adding value to traditional publishing, for example, by enabling links to material both internal and external to a document, taking full advantage of the expanded technical possibilities offered by electronic publishing, and full account of user wants and needs. Such value-added functions should include facilitating greater access to existing and future print publications, including "search and seizure" features such as dynamic citations and dynamic abstracts.

  7. Requirements and standards for output-file formats must take into consideration the wide variety of workstation capabilities that are available globally. User tools should be affordable and user-friendly. Otherwise, access and participation will not be available to those less technologically advantaged.

  8. New pricing and revenue models are needed to ensure that costs of production, dissemination, and utilization of scientific and technical information are reduced during and following the transition from print to electronic publishing. Pricing per article and custom journal and article packages, for example, could prove to be cost-effective in an electronic-publishing environment. Ultimately, however, traditional publishing practices that transfer most of the costs to the end-users may need to be reconsidered.

  9. Care should be taken in recasting copyright standards for electronic media in order to preserve a balance between the interests of the copyright holder and the public good. Appropriate principles of fair use should be maintained in an electronic-publishing environment, and technical innovations aimed at upholding these principles, such as "electronic reserves," should be implemented when possible.

  10. There is a need for close and ongoing cooperation among professional scientific and technical societies, commercial publishers, authors, libraries, and end-users in identifying and overcoming obstacles to the creation and utilization of scientific and engineering documents in a globally networked environment.

The first recommendation is the most vague in that it calls for more research on the information package of the future. The point of that recommendation is that if one thinks only of electronic publishing as rendering electronically what one does in print publishing, then we're missing many of the possibilities of electronic publishing. The working group wasn't sure what the information package of the future would look like, hence the call for more research, but they were sure that it should look completely different from what is traditionally conceptualized as the scientific and technical information package. For example, papers of the future might consist of information modules, thus allowing a reader to search the literature in a particular area for all the paper modules that deal with research methods. And there's a whole new literature that librarians call "gray literature" springing up — for example, research notes or artifacts that can be portrayed electronically that in the past haven't made it into the formal scientific literature.

The second recommendation notes that it's essential to address and maintain the intellectual integrity of the scholarly enterprise. Both professional-society and commercial publishers certify that material has gone through a review process and certify that it's authentic; they must continue to do so in an electronic environment. Such issues as ownership of intellectual property also need to be reexamined in this new publishing environment.

Recommendations three and four speak to the importance of archiving. Many people at the workshop said that the job of archiving shouldn't fall to publishers or libraries, but there should be new institutional arrangements created for the archiving function. Efficient document location would be a necessary feature of such archives. Another key point is the need to archive hardware as well as data, information, and documents. For example, are CD-ROMs going to be readable five years from now or ten years from now?

"We found the people to be knowledgeable not just on the technology of electronic publishing, but also more knowledgeable in scholarly-content areas than publishers often give them credit for"

Recommendation five deals with the need for authoring tools that make sense from the authors' perspective.

Recommendations six and seven speak to users and prospective users of electronic publications, including the use of electronic media to enhance the value of existing print materials through dynamic abstracting and indexing. Concern was also expressed about pushing the technology so far that it leaves behind not just people in the developing world, but people in the industrialized nations as well who might not have the technological capabilities that others have.

Recommendation eight focuses on new pricing and revenue models. Electronic media present possibilities for new products, such as pricing by the article and customized journals. The working groups at the workshop also said that we may ultimately need to rethink the current model where the majority of costs are transferred to the end-users. The groups suggested transferring more of those costs to authors, research funding agencies, or perhaps even outside foundations.

The ninth recommendation focuses on intellectual property. While much concern has been expressed about intellectual property from the point of view of the owner of the intellectual property, there is also concern about intellectual property from the point of view of users. Much of the recent dialogue on rewriting copyright standards, in the opinion of our workshop participants, has paid too little attention to principles of fair use that have been developed over the years.

The final recommendation was that a continuing dialogue is needed among the scientific and technical societies, the publishers, authors, libraries, and end-users. We were particularly impressed at our workshop by the energy, knowledge, and enthusiasm of the members of the library community. We found the people to be knowledgeable not just on the technology of electronic publishing, but also more knowledgeable in scholarly-content areas than publishers often give them credit for.

The papers included in this special issue are expanded and revised versions of papers presented at the workshop and are representative of the workshop's paper sessions.

"Designing Electronic Journals With 30 Years of Lessons From Print" provides an in-depth review of scientific publishing trends since 1960, through surveys of more than 12,000 scientists and other professionals. Survey results presented by Carol Tenopir and Donald W. King of the School of Information Science, University of Tennessee, indicate that while the usefulness of reading scientific journals remains high, publishers' pricing strategies have caused the number of personal subscriptions to drop. Instead, readers are relying more heavily on libraries and other shared collections. However, due to high fixed costs of publishing and a shrinking number of personal subscriptions, both commercial and professional society publishers have compensated for their losses in personal subscriptions by dramatically increasing their prices for library subscriptions. The lesson for electronic publishing is clear: We must find ways to avoid the vicious pricing cycle that is evident in the world of print. While evidence suggests that costs of electronic publishing may be only slightly less than paper, electronic media makes possible disaggregation of various "levels of information," such as titles and abstracts, reviews, and individual articles. Tenopir and King conclude that success will depend in part on inexpensive access to separate articles and in part on the ability to easily locate and access them. Pricing related to site licenses needs to be reexamined to account for organization size and readership. They predict that print and electronic journals are likely to coexist, but electronic journals may be best able to service smaller, niche markets.

Lloyd A. Davidson and Kimberley Douglas, librarians at Northwestern University and the California Institute of Technology, focus on "Digital Object Identifiers: Promise and Problems for Scholarly Publishing." Recognizing the critical need for a flexible, robust system of permanently identifying digital information, the authors nonetheless are critical of the approach taken by the major scientific and technical publishers in establishing a system of Digital Object Identifiers (DOIs). In particular, they argue, DOIs as presently conceived are not responsive to the needs of smaller publishers and the library and user communities.

"If electronic media can contribute to a decreased time period between article submission and final publication, it would overcome a serious problem with conventional publications"

The paper by Marion O. Hagler, Janet C. Rutledge, William M. Marcy and Ted E. Batchman, "Archival Journals: Perspectives Gained by E-Publishing IEEE Transactions on Education," describes a series of successful experimental publications on CD-ROMs for the IEEE Transactions on Education journal, published by the largest professional society of engineers, with more than 300,000 members worldwide. The authors demonstrate that electronic publication for engineering education can offer advantages to authors, editors and reviewers, as well as educators, students, and archivists. This technology expands the "media palettes" for authors, at the same time it increases the efficiency of the review process while providing a means of archiving engineering software, and sharing complex, file-based innovations in teaching materials for educators and students.

Another excellent example of electronic publication can be found in Living Reviews in Relativity. Jennifer Wheary, Lee Wild, Bernard Schutz, and Christina Weyher, all of the Max Planck Institute for Gravitational Physics, describe this physics journal as a research tool offering an integrated information system in their article, "Thinking and Developing Electronically." The authors point out that the rapidity of developments in relativity theory, along with its interdisciplinary nature, have made access to up-to-date review articles essential to the work of scientists in this field. Like other authors, they stress the importance of high-quality content. They ensure it through invitations to prospective authors by an editorial board of qualified experts, and a peer-review process prior to publication. The plan behind the journal's development was to complement the roles of traditional print journals and existing electronic resources by facilitating access to them, along with a review system to assist researchers in organizing and evaluating these ever-increasing storehouses of information. References cited in Living Reviews are stored in the Living Reviews Reference Archive, which is searchable online. The Max Planck Institute underwrites the journal, and encourages collaboration with other institutions interested in developing effective noncommercial scholarly publishing outlets.

In "Into a Glass Darkly: One Scientist's View," R. Keith Raney from the Applied Physics Laboratory, Johns Hopkins University, offers a reserved perspective on the potential usefulness of electronic publishing. To achieve widespread acceptance, the technology must: 1) meet the needs of the scientific community; 2) overcome major obstacles related to traditional print publication; and 3) provide new capabilities not possible through conventional means. In terms of the scientific community's needs, Raney expresses concerns about maintaining standards for the quality of published information, and the durability of archival records. He suggests that if electronic media can contribute to a decreased time period between article submission and final publication, it would overcome a serious problem with conventional publications. Finally, he highlights new capabilities of the electronic environment, referred to as "search and linkage," that permit the construction of dynamic-citation sites, such as the one at the Institute of Physics, and dynamic abstract sites such as IEEE Bibliographies On-Line. It is these two innovations that Raney claims provide users with unprecedented power.

Ann Peterson Bishop of the Graduate School of Library and Information Science, University of Illinois at Urbana-Champaign, discusses problems in accessing electronic documents in her article "Logins and Bailouts: Measuring Access, Use, and Success in Digital Libraries." Drawing on her research for the Digital Library Initiative project at the University of Illinois, Bishop discusses the problems involved in monitoring and overcoming barriers to user access to the project's testbed, taking as an example problems encountered in logging into an electronic-document system. Following efforts to streamline the authentication and registration forms required for access, the research team still found that a high percentage of potential users withdrew before completing the registration process.

Taken as a whole, the papers included here, the other papers in the SeDEP Workshop proceedings, and the reports of the SeDEP working groups [1] make a strong case for the importance of social and economic factors in shaping the eventual form and importance of electronic publishing in the scientific and technical communities. E-publishing advocates and skeptics alike would be wise to pay closer attention to socioeconomic factors.

— Joseph R. Herkert and Christine S. Nielsen