The rate at which developments are taking place in electronic publishing, both in actual products and in various types of experiment, is very rapid - approaching the rate at which meetings and conferences on such topics are taking place. No short review can cover all of these, but indications will be given of the main areas in which such developments have occurred or are occurring. Such developments have impacts of different types on the various participants in the information chain (it is really a network but I must avoid a word now used mainly with a special connotation), such as the research scientist as both a creator and user of information, the primary publisher, the librarian, the secondary publisher (often a data bank creator also), the on-line vendor, the document delivery service and the technologist in industry concerned to be aware of new scientific developments. Some of these impacts, and the views of the various participants have been discussed in a report published a few years ago  and a more recent report  has addressed the impact of electronic publishing on library services and resources in the UK, giving useful references to some of the many surveys and reports to which this conference will add. Those impacts on libraries are both organisational and economic and have significant effects on the other roles of academic libraries in maintaining a book stock appropriate to their users. Perhaps the most significant conclusion of this latter report is that there is an urgent need for publishers of scholarly electronic journals to establish a code of practice covering peer review, dating of documents, the availability of an archive independent of the publisher and the introduction of a universal system of identifiers and tagging to facilitate access and establish copyright.
The report mentioned above (1) was novel in seeking the responses to questionnaires of 4000 scientists and engineers and 600 libraries. It also contains references to some other more limited studies. Some of the information obtained was rather was rather surprising; it was also supplemented by personal interviews with a small group of scientists, librarians, funding agencies and publishers. Less than 20% of scientists in academic institutions made frequent use of current awareness services, as opposed to more than 40% in industry, and chemists were more frequent users than any other group of scientists, but even so other studies show that most adopted low technologies for information acquisition. For all users CD-ROM versions of abstracts was the most popular modern route, but the printed journal ranked first for nearly all disciplines. In addition to lack of awareness, major constraints on the use of new technologies were given as habit, time and lack of training resources. Although 74% would like access to electronic journals a smaller number (58%) is prepared to submit work to an electronic journal. Mathematicians and physicists seem to be those best equipped with electronic access to the library from their office or laboratory. As might be expected confidence in using electronic information access decreased sharply and steadily with age above 25, while the satisfaction with information about such methods moved in the opposite direction with age.
The acceptance by journals of papers in machine-readable format is now common, although the flexibility of the recipient to make use of a wide range of different types of disc is not always as great as the authors would wish. For traditional journals it is usual for a print-on-paper version also to be submitted so that refereeing can be carried out in the normal way. A possible development is indicated by the experiment the American Physical Society is carrying out with a selected group of referees for Physical review and Phys Rev Letters who will receive papers electronically and submit their reports in that way. At present refereeing delays are a significant contribution to the total submission-publication period. Machine-readable (CD-ROM) versions of a number of journals both old and new are now available for the material of a given volume or year, sometimes at little extra cost to the subscriber to the print-on-paper version but there seems little enthusiasm for this with librarians with long-established primary journals. There are, however, real hopes of saving library space in due course if at small expense old bound volumes can be replaced with CD-ROMS, the technology for which is well established even if the lifetimes cannot yet be guaranteed to equal those of 18th Century volumes of Phil Trans Roy Soc. Librarians are however slightly apprehensive about the possible expense they may incur if they receive current journals electronically, if they then have to establish their own local electronic warehouse from which their users can draw material. In some fields (e.g. Particle Physics) e-print archives, or for some users Bulletin Boards, have replaced a former widespread circulation of non-refereed reprints but those are unlikely to become accepted reference sources in, say, Condensed Matter Physics where readers will be reluctant to wade through a vast mass of material before it has been peer-reviewed, and even in Particle Physics, where some accept that priority is determined by the pre-print or the electronic archive that incorporates it, the need is felt by many for a traditional paper to appear in due course, and with speeded up processing by referees and publishers that may not be long after the dust has settled after disputes in the electronic archive. Some of these aspects will be discussed in other sessions. There are signs that peer-reviewed material in electronic form will be recognised by funding agencies and promotion panels but this is unlikely to be the case for non-refereed material in Bulletin Boards and electronic "archives". Citation of electronic material in more conventional journals also presents some problems; some journals have announced that since pagination may differ in the electronic and print-on-paper versions the electronic version will indicate that reference to it should be as for the print-on-paper if that has not yet reached the libraries. On the other hand those dedicated primarily to the electronic archive will use the assigned archive reference number as the means of reference. Presumably by-no-means-private Bulletin Board material can be referred to in conventional journals as "private communication". There is a steady growth in the number of rapid communication journals that have both traditional and electronic forms even when the efforts of referees and editors to maintain them as high-prestige journals and the efforts of authors therefore to appear in them make them by no means rapid. Electronic only journals (peer-reviewed) seem to be developing much more slowly in spite of the conviction of some that they are the pattern of the future, perhaps because authors fear invisibility in part of the world and perhaps because of the continuing pleasure young researchers still feel at seeing their work literally in print. What is still very uncertain is the economic basis for this type of journal, since the intellectual and organisational work that was traditionally paid for by the subscriptions to a printed form will still have to be carried out and other costs will be involved, even if, as the optimists think, the network remains free for ever. This is a matter of great concern to smaller learned society publishers and will be discussed in other sessions. Some of the technical aspects, both those already available and those being developed, will also be discussed in other sessions, and include hypertext links from points in the text to figures, tables and even papers referred to if they are available from the same server.
A number of studies and investigations of possibilities now on the horizon are being carried out and one of particular interest is the SuperJournal project that has been reported by David Pullinger in a report with that name (British Library R& D report 6126) published by the Institute of Physics. It is clear that with adequate band width for the network and standardisation on conversions of text to SGML there are no significant technological barriers to the creation of electronic journals with all the features that the enthusiasts would desire. The Institute of Physics has itself a number of developments under way and gives details on its world-wide web home page; other publishers also give details of their electronic products and projects on home pages.
Online searching of bibliographic data-bases has been available for a long time in many libraries, but some librarians are concerned about the strain on their budgets, especially if inexperienced users are allowed to couple directly to quite expensive vendors. A popular alternative for abstract journal searching is for the library to provide CD-ROM versions which also help with the ever increasing size of print-on-paper abstract journals. This is also a suitable form for citation indices and data compilations, being very easy for alphanumeric data and currently being extended to a number of other categories of data, such as chemical structural formula, crystallographic structures and maps. A detailed discussion of electronic access to materials data is given in a recent issue of Materials Research Bulletin .
Document delivery, which has been with us in some form from the information stone age, is also capable of total transformation in the electronic age, but the potential costs inhibit many librarians. (This is even true of the British Library express fax service.) Very few users in the survey discussed above (1) were aware of the ADONIS project, and a popular and efficient electronic document delivery service will probably have to await solutions to the problems of the electronic archive.
One area of great uncertainty to be discussed at this meeting concerns the nature and management of the ultimate electronic archive, a matter of great importance, if not always recognised by those who assume that the function of the information system is to tell them what their colleagues and rivals have been doing in the last few weeks. In fact, a quick glance at recent papers in even rapidly-moving fields will show how dependent today's scientist is on the work of his forbears and his access to what they published. Consideration of the nature of the electronic archive raises the question of whether, if we search it by keywords or authors' names and refer to our findings only by the entry in that archive, the concept of the "journal", the "volume" and the "issue" will still be needed. The new electronic journals in effect maintain their own mini-archive in the full text database of the journal, but there has been little discussion of the nature, management location and financial responsibility for an electronic archive for the full text of all established journals. The British Library report emphasises both the need for accessible archives independent of the publisher, the considerable cost of creating them and the uncertainties about the lifetimes of their contents.
 The Royal Society (1993) The STM Information System in the UK A study on behalf of the Royal Society, the British Library and the Association of Learned and Professional Society Publishers.
 The British Library (1994) The impact of electronic publishing on library services and resources in the UK.
 Materials Research Bulletin (1995, August, p40) Electronic Access to Factual Materials Information: The State of the Art
Listing by Author's Name, in alphabetical order
Return to the ICSU Press/UNESCO Conference Programme Homepage
University of Illinois at Urbana-Champaign
The Library of the University of Illinois at Urbana-Champaign
Comments to: Tim Cole