REPORT

PREAMBLE
Electronic journals create added value in publication that has great appeal to scientists and publishers, and the number of electronic journals in science, engineering and medicine - refereed and unrefereed - has increased dramatically in recent years. But electronic publishing, with its greater flexibility and variety of presentation, challenges conventional norms and practices. How will the existing culture and practices associated with publication in science be affected? What standards should apply in the electronic environment?

To explore these and related questions, an international workshop was organized from 12-14 October 1998, under the auspices of the American Association for the Advancement of Science (AAAS), the International Council for Science (ICSU), and the United Nations Educational, Scientific and Cultural Organization (UNESCO) to examine the application of electronic methods to the publication of scientific journals with a view to encouraging the development of internationally recognized practices and standards. The Workshop was intended to build on the work of the international conference of "Experts on Electronic Publishing in Science" that was convened by ICSU Press and UNESCO in February 1996. A major recommendation emerging from that meeting urged the convening of a forum involving scientists and their organizations "to formulate codes of ethics and of conduct for electronic publication which would spell out the reciprocal obligations of the scientist and the community on such matters as peer review, citation, integrity and authentication of material and archiving."

The Workshop enjoyed broad international and scientific representation, as reflected by the presence of participants from universities in Africa, Europe, India, Latin America, and the USA. The Learned Societies represented those with extensive publishing programmes in Astronomy, Chemistry, Physics, Psychology and Medicine, with other disciplines being covered in a less formal fashion. In addition, there were representatives of the National Academies of Russia, Sweden and the United States. Participants also came from National and Institutional Libraries, the International Council for Scientific and Technical Information (ICSTI), the International Network for the Availability of Scientific Publications (INASP), the Open Society Institute, and the Pan American Health Organization/WHO, as well as from the main sponsoring organizations. In addition, scientific, technical and medical publishing was represented through the Association of Learned and Professional Society Publishers (ALPSP), STM Publishers, several commercial companies and consultants.

RECOMMENDATIONS
Wide variations in the practices and traditions of scholarly publishing across disciplines suggests caution in attempting to construct generalizations intended to apply broadly. Nevertheless, the Workshop was able to identify a number of issues that merit attention by the scientific community at large as a basis for the series of recommendations that follows.

1. Defining a Publication. Digital processing facilitates the production and preservation of several public versions of a document or scientific paper and the Workshop identified three significant iterations. These were the initial posting to an electronic database, which would in some circumstances determine priority for scientific and patent purposes; the definitive refereed version carrying the imprimatur of a journal; and a current version possibly incorporating corrections, modifications, references to subsequent citations and links to related material. In some subject areas different versions are common and even encouraged, in others they are strongly discouraged. Whatever the normal practice may be, the existence of multiple versions brings the possibility of confusion in citation and referencing, and the Workshop

RECOMMENDED that each publicly available version of a document carry a full specification of its status laid out in a visible and readily understandable manner.

2. Citation. Because of the possible existence of multiple versions of a document, there is need for a convention on the citation of electronic material. Here, precision is more important than the compression favoured in print publication, so that full specification alongside appropriate metadata becomes practicable. The citation should clearly identify the version being referred to or read. While use of the URL and URN is convenient, more precise standards for identifying digital objects must be developed. The Workshop

RECOMMENDED that the scientific community become involved in the development of standardised citation practices that are friendly to science, include appropriate metadata, capable of automatic assignation and easy to use.The International Council for Science (ICSU) should also be involved in some way with the development of standards for identifying digital objects.

3. Peer Review. The ease of publication in electronic media reinforces the case for adequate quality control in terms of both the scientific content and the presentation. Peer review, commonly involving the use of targeted and anonymous referees chosen by knowledgeable editors, is widely accepted within the scientific community at large, although its imperfections are also recognised. In the electronic publishing world, the value of alternative systems involving exposure to the unfiltered comments of readers is being explored, but it was considered that more evaluation was necessary and that it was essential for readers to be aware of the refereeing policy adopted by each journal. Formal peer review was regarded as essential in arriving at the final version of a scientific publication, although greater openness in the mechanisms adopted was desirable. It was pointed out that more research is needed concerning the views of users and the Workshop believed that the findings of a current ALPSP project involving study of the aims and requirements of contributors to journals will be of great interest to those concerned with scientific communication. The Workshop also considered that peer review and refereeing form an essential part of the training and duties of scientists, that they should be carried out conscientiously, and

RECOMMENDED that scientific societies and/or journals establish and distribute guidelines in order to maintain the quality and integrity of the review process. Concern was expressed that any reduction in the controls applicable to electronic publication increases the opportunities for scientific misconduct, such as the faking of results and plagiarism, although it was noted that the technical features of electronic publication (such as the ease of automatic scanning and searching for similarities) raises the chance of detection. The Workshop

RECOMMENDED that research into the application of electronic methods for the detection of scientific misconduct be encouraged.

4. Full and Open Access Versus Intellectual Property Rights. The Workshop considered the conflict between the needs of scientists for ready access to large databases and collections of scientific observations and the requirements of the aggregators for commercial reward for their endeavours. It was felt that the results of publicly funded research should be clearly recognised as a ''public good,'' and that full and open access to the data collected was essential for scientific advancement. This does not necessarily mean ''free'' access by research workers, since the part played by publishers, aggregators, librarians and other facilitators merits adequate recompense if they are to continue in their valuable roles. However, such intermediaries must not be seen as gatekeepers preventing access to material expensively acquired by the public purse. The long-established balance of copyright protection is being changed by new legislation in Europe, the USA and other countries, as well as by a potential WIPO sui generis right. Because new technology makes copying and illegal use, such as piracy, easier, there is a trend toward the tightening of controls that appear to favour the producer as against the user, but provision for the use of databases for non-commercial purposes is essential for the healthy development of scientific research. The Workshop

RECOMMENDED that the attention of the scientific community, funding agencies and legislators be drawn to the fact that the scientific enterprise is crucially dependent upon the ability of research workers to make use of collections of facts and observations and that measures that limit access to such material are contrary to the public good. Legislators are urged to provide for a mechanism permitting 'fair use' of large databases in order to promote full and open access to critical data for scientific research and education with little adverse effect on the commercial interests of the owner. It was also

RECOMMENDED that different approaches to the assignment of copyright be encouraged and that they be evaluated for their effectiveness in making information available to scientists.

5. Privacy. It is now possible for journal editors and publishers to collect detailed information on the nature of the material accessed and the usage, and to compile author/user profiles. The kind of information being collected should be clearly stated by journal editors and/or publishers, as well as the use to which it is being put. Whilst statistical data of this type can be of great value to both users and providers, and their use legitimate, disclosure of the findings should be subject to the agreement of both parties. In particular, specific information relating to individuals should not be divulged to anyone without the permission of the subject. The Workshop

RECOMMENDED that guidelines be developed to cover the use of statistics and information collected in the course of database server operation. The Workshop also

RECOMMENDED that, as part of the suggested guidelines, journals adopt and visibly post policies that disclose to users the nature and destination of the information collected. Readers should have the opportunity to opt out of the collection of their data.

6. Archiving. The Workshop recognized that a major factor restraining the adoption of electronic methods of publication was the lack of archiving facilities, for there is currently little assurance for authors and publishers that, with the advancement of technology, material issued electronically will remain available and readable in the decades and centuries to come. The deposition of copies of printed works in national archives has served society well, but there is no counterpart as yet for electronic works. The Workshop

RECOMMENDED that moves toward the establishment of national archives, perhaps on an internationally collaborative basis, for the long-term preservation of electronic publications be encouraged as fundamental to scientific and cultural development. Such efforts should be mindful of changing technologies, so that collections can meet new standards or be converted to new formats.

7. Developing Countries. The Workshop considered that electronic publishing represents a unique opportunity for developing countries to promote the advancement of their scientific communication. It has the potential for improving access to the world literature, for making up gaps in local collections, and for improving the visibility of their own scientific contributions. Electronic publishing also represents a challenge both to the scientific community and to those responsible for its organization. It was accepted that improvements in basic infrastructure and telecommunications facilities are required and

RECOMMENDED that the attention of Governments and Funding Agencies be drawn to the needs of science and education when local telecommunication and computer facilities are being developed. Regional differences are significant and hence local initiatives are very important. Examples from Latin America, Eastern Europe, and Africa were reported. As one means of enhancing the quality and visibility of journals published in developing countries it was

RECOMMENDED that ICSU Press and UNESCO (with INASP) should investigate proposals for the ''twinning'' of journals or groups of journals produced in the developed world with those published by learned society and other publishers in developing countries. The goal should be the development of true partnerships while avoiding relationships which might heighten the dependency of one party on another.

8. ICSU. The members of the Workshop urged that ICSU Press consider the production of 'Guidelines to Best Practices in Electronic Publication'.