Electronic Journals: a survey of the literature and the Net

HTML version of an article published in JoiN, Journal of Information Networking, vol. 2, no. 3 pp. 169 - 186. Some links to electronic sources added. Text may vary slightly from printed version. Content remains intact.

Hans Roes
Tilburg University Library
PO Box 90153
5000 LE Tilburg
The Netherlands

Abstract

Starting from the present day, printed information cycle, the evolution of the electronic journal is explored. A survey is presented of existing peer reviewed electronic journals. Next the literature on the subject is reviewed showing that not all of the perceived advantages of electronic journals have yet materialised. This can partially be explained by the difficulties facing libraries when trying to integrate the minimal but already diverse set of existing electronic journals into their infrastructure. The article concludes with an analysis of the enabling/inhibiting role which the primary actors in the printed information chain could play in the further evolution of the electronic journal. Libraries and scholars have a common stake but face the long standing experience of publishers in organising the refereeing and editorial process.

Introduction

Approximately 200 years after the invention of printing the first scientific journals appeared almost simultaneously in London (Philosophical Transactions of the Royal Society of London) and Paris (Le Journal des Sçavants) in the year 1665. The growth of their numbers has been remarkable. In 1800 there were estimated to be 100 primary journals, in 1850 1000, in 1900 10.000. If growth had sustained in this tempo there would have been 1.000.000 journals around the year 2000 [1]. Growth has however, and fortunately so, dropped of, and in 1991 there were estimated to be approximately 133.000 primary journals [2]. How many of these journals can still be considered to be `alive' is unclear though.

Today, the printed journal is the most important medium in which the progress of science is recorded. It is the centre of an information cycle in which the actors are scholars, both as producers and consumers of information; primary and secondary publishers both commercial and not-for-profit, which organise the editing and refereeing process and offer abstracting and indexing services; and libraries which select titles from the published universe to satisfy their patrons' needs, archive the record and supply the information on demand, both to their own institutions and via Inter Library Loan to remote institutions. The main functions of the journal are: communication and dissemination of information; quality control which is carried out through a peer review process; and archiving [3]. A fourth important function, closely connected to the quality control function, is the assignment of priority. These basic functions are nowadays fulfilled by, and therefore attributed to, printed journals, but they are fundamental to the process of scholarly communication. They should apply to any new medium if it is to serve adequately as a carrier for communicating and recording scholarly work.

There are signs though that the printed information cycle might collapse. The familiar term in the library world to indicate this danger is the serials pricing crisis. Library budgets are simply inadequate to catch up with the growth in both the volume and price of journal literature and the economic basis of what Karen Hunter of Elsevier has called the "publishing ecosystem" [4] is eroding. Also it is obvious that the printed journal is losing its main function of current-awareness communication. Scholars, even before the electronic era, have found faster and more informal ways of communicating their work. But since the inception of the Internet, with its many conferences organised in listservers, newsgroups and bulletin boards, the scientific community has turned into a truly global village. In this setting it was to be expected that an electronic counterpart of the printed journal would come into existence and it did so in the late eighties of this century, more than 300 years after printed journals started. The question is whether the electronic journal can become the core of the virtual library, fulfilling the functions served by the printed journal of today and perhaps innovate or even transform the way in which the information cycle works.

This article addresses the transformational potential of electronic journals by first trying to assess the current magnitude of the phenomenon. How many electronic journals are there and what are their characteristics ? Next, the literature on the subject will be reviewed so as to get an impression of the expectations and perceived problems associated with electronic journals. Then the problems facing libraries in dealing with this new medium will be investigated. The article concludes with an analysis of the potential of electronic journals to fulfil the basic functions of communication, quality control and archiving, and an analysis of how the main actors in the information cycle, can inhibit or enable the transition from printed to electronic journals.

Electronic Journals: definition, scope and characteristics

The basic problem in defining what an electronic journal is, is that if one sticks to much to the present world of the printed journal, Kessler would be right in observing that "electronic journals at first glance appear to be the networks' leading anachronism" [5]. Judy Myers et. al. describe this dilemma as follows: "The term `electronic journal' does as little to describe [the] future as the term `horseless carriage' did to capture the promise of the automobile." [6] Probably the best way of approaching the problem of what an electronic journal is, is to look at some developments in the library and network world and ask if they constitute what might be called electronic journals, serving the basic functions of communication, quality control and archiving.

The first development which comes to mind are projects experimenting with electronic equivalents of printed journals. One of the oldest examples is ADONIS where images of articles published in printed journals are distributed on CD ROM [7]. More recent examples are Elseviers TULIP project in the United States and the RightPagesTM Service at AT&T Bell Labs [8]. A similar project with Elsevier at Tilburg University library involving about 100 journals which will be made available as images on end users' workstations, can also be mentioned here [9]. Still older examples are full text databases run by the major host organisations. All of these projects involve journals and all of them are by definition electronic, but one would not want to call them electronic journals since they are more adequately described as electronic versions of printed journals. The notion electronic journal suggests something new, something synergetic, something which has sprung from the Net itself.

On the Net itself there are numerous candidates which fulfil the main function of communication: computer conferences organised in listservers and/or through Usenet; electronic pre-print services; electronic newsletters and, indeed, phenomena which label themselves as electronic journals. These are not derived from print serials [10] and therefore come closer to what one would expect. The question is whether they can be regarded as scholarly, peer reviewed journals.

The problem is of course the anarchic condition of the Internet which makes every attempt of describing the state of a development look like shooting at a moving target. Three major starting points for finding out about the state of electronic journals are, first of all The World Wide Web (WWW) Virtual Library, second the CICnet electronic journals project and the ARL Directory of Electronic Journals, Newsletters and Academic Discussion Lists

The WWW entrance to electronic journals on the Internet shows a lists of (August 1994) about 40 journals which can be directly accessed, as well as pointers to "Various Archives and Indexes." Most of these point to magazines. One of them points to the CICNet archive.

The CICNet archive can be accessed via gopher (URL gopher://gopher.cic.net) and there are also many gophers in the world which point to this huge archive. Problem with the CICNet journals project is that it seems to archive anything which describes itself as an electronic journal, or newsletter, or magazine. The archive contains currently (August 1994) about 700 journals which are accessible via an alphabetic gopher directory, as well as via a subject tree. Among these 700, quality varies enormously from occasional magazines to high quality peer reviewed journals. Also it is absolutely not clear how many of the journals and magazines archived can be considered to be still `alive'. It is quite amazing how this problem, which is also known from studies of the market of printed journals, reproduces itself so quickly in the world of electronic journals. Nevertheless, the CICNet archive is a useful starting point for exploratory purposes.

A third approach is offered by the Association of Research Libraries (ARL), which produces the Directory of Electronic Journals, Newsletters and Academic Discussion Lists, in combination with the listserver NEWJOUR-L. The Directory can be retrieved at URL gopher://arl.cni.org. The ARL directory is separated in two files. Academic discussion lists, containing almost 1800 entries, and Journals and Newsletters with over 440 entries according to Ann Okerson's Introduction to the 1994 Directory, dated april 1994 [11]. It is beyond any doubt the most comprehensive overview. The difference between the CICNet number of 700 and the 440 journals and newsletters in the ARL directory is probably in large part due to the fact that the CICNet archive also contains journals which are no longer active, although there could be other reasons. The NEWJOUR- L listserver is intended for the announcement of new electronic journals and newsletters. Publishers of electronic journals are encouraged to send messages about their ventures to this list which also serves to maintain the ARL Directory. It is possible to subscribe to this list by sending an e-mail message to listserv@e- math.ams.org. The body of the message should read: SUBSCRIBE NEWJOUR-L Firstname Lastname.

According to Okerson [11] the file Journals and Newsletters contains 74 refereed journals, but these are not as such recognisable from the descriptions in the file. A first selection by the author showed a number of 57 potential candidates for refereed scholarly journals. These 57 were all checked at the given network address to look for closer data on year of first issue, subject field, formats and for possible editorial statements. In this process 18 journals dropped out of the sample. The most common reason for this was that articles or issues could not be obtained at the given address, two journals were abstract journals communicating abstracts of articles appearing in print, one journal had been announced but not actually started, another journal published only book reviews, a couple of journals appeared to be discussion lists and one journal turned out to be a pure database, at least no articles could be detected at the given address.

The remaining 39 candidates are, without any claim for being exhaustive, listed in an appendix. The following data on these electronic journals could be collected:

Year of first issue Number

1987                1
1988                -
1989                2
1990                5
1991                4
1992                6
1993                15
1994                6

Although the number of (apparently) peer reviewed electronic journals is still very small, especially when compared to their printed counterparts. The numbers indicate though an accelerating trend. Coverage of the ARL directory is up to April 1994.

Mode of access      Number

mail/usenet         21
gopher              26
ftp                 22
wais                3
www                 14
proprietary         3

Modes of access vary considerably while many journals offer a combination of modes of access. A strong newcomer is World Wide Web. Some of these journals offer also the hypertext features made possible by the Web's format HTML. Apparently these hypertext features are mainly used for linking inside an article (from contents to text, from footnotes to references and back) and not so much for linking with other articles on the Web, yet, since the body of literature to link with is, because of its recent development, of course, fairly limited. The three journals with proprietary access are also the only three journals which are commercial ventures, the well known Online Journal of Current Clinical Trials and its recent companion the Online Journal of Knowledge Synthesis for Nursing, and Electronic Letters Online.

ASCII               29
TeX                 11
Postscript          10
Other               7

Multiple formats are possible, but ASCII is still the most popular, which implies that the possibilities for graphics, equations and non latin character languages are severely limited. Because of these limited possibilities of ASCII, and perhaps because many mathematicians are familiar with the format, most math journals choose TeX.

Looking at subject fields it appears that Mathematics is the area having the largest number of peer reviewed journals: 8. Math seems also the fastest growing category with three journals started in 1993 and two started already in 1994 and another one announced to start late summer 1994. The second largest field is education, followed by the social sciences. Library science is represented by three journals and there are two electronic journals on the subject of electronic journals. Remarkable is that in the field of computer science and information technology there are only three electronic journals.

Finally, 5 of the electronic journals in the sample are derived from printed counterparts which have an older history. In table 1, the year in which the electronic sibling started was chosen.

Unfortunately, no data could be found indicating the use of existing electronic journals. Very little is known about user acceptance of this relatively young medium.

Compared to the over 130.000 printed journals the phenomenon of the electronic journal seems to be insignificant. Also it seems that the number of electronic journals is not growing as fast as many expected. Okerson [12] for instance, predicted in 1991 that there would be about 100 refereed electronic journals in 1995. That number will probably not be achieved, but it certainly will not take 135 years to reach the 100 as it did with printed journals, since electronic journals are the main candidates for information carriers in the emerging virtual library.

Having hopefully cleared the ground a bit it is interesting to look at the expectations and perceived problems which are discussed in the literature on electronic journals.

Review of the literature

The first advantage of electronic journals lies in their speed of communication and is mentioned by nearly everyone writing on the subject. This advantage is, of course, dependent on the speed of the peer review process, but this is usually also carried out via the Net. Well linked with this observation is the notion that the article as a more natural unit of information [2, 13] can be sent out immediately after the review process has been completed and does not have to wait until an issue can be completed. This raises the important question as to whether there will remain a need for an entity like the journal. It probably will, since printed journals have a tacitly understood hierarchy which will most likely be reproduced in the electronic world as well. A journal title functions as an important measure of quality of an article because of the editors and their editorial policy associated with it. But the need for issues as bundles of articles will disappear, contributing to the speed of communication.

Another advantage is that there are no limitations to the size of an article [14] and the electronic format can be used to add experimental data or to create even multimedia documents [15, 16]. Moreover the principal possibility of linking an article with the body of literature on a given subject is sometimes mentioned [14] or at least a link with discussion on the article itself, creating interactivity in a system which Harnad calls "open peer review" [17, 18]. Whether this will replace the pre-publication peer review is doubtful since it would undermine the standing of a journal which is to a large extent based on editorial policy. Harnad, as a founder of Psycoloquy, one of the leading pioneers in this field, goes even further and thinks that this might lead to a fourth cognitive revolution (after speech, writing and printing) because the speed of communication between scholars will approach more the speed of thinking, thereby boosting creativity and productivity. In this sense, Harnad expects a true transformation of the process of scholarly communication. The journals in the sample hardly exploit these opportunities. Two journals have related graphics material, one via separate JPEG/MPEG (still and moving video) files, two link with comments and replies, one adds source code which is merely another form of ASCII, and three use limited hypertext features.

Compared to Harnad the expectations of librarians are more down to earth. Electronic articles have the advantage of not demanding shelf space nor binding costs. They cannot be mutilated nor stolen and are always available [2, 19]. Much expressed is the belief that they are a solution to the serials pricing crisis [15,16,18,20] and give universities the possibility of regaining control over the publication process which is now largely in the hands of commercial publishers [2]. Finally, electronic articles can be merged into an alerting system based on user profiles [14] while Manoff et. al. observe that the difference between catalogues, abstracts, indexes and full text can gradually disappear [15]. This would indeed be one of the main advantages compared with library systems of today where the retrieval of references is a piece of cake compared to the sometimes cumbersome retrieval of the documents referred to [21]. Whether it implies if catalogues and indexes could disappear remains to be seen, since these have important functions in retrieval. But through integration of systems, retrieval of pointers to information and retrieval of the actual documents pointed to, can be combined in one act.

One of the basic problems of electronic journals, at least partly due to their relative youth, is that they fail as yet to attract prestigious authors. This observation is often associated with a perceived lack of quality control. According to Bennion the electronic journal lacks quality control [3], but Harnad points out rightly that this is independent of the medium although he ascertains that prestige is important [17]. Acceptance of the medium by authors is also considered to be a major problem by Line [22], and Metz and Gherman [2].

A related factor seems to be that the electronic journal is considered to have low visibility [15]. This problem seems not only inherent to relative youth but is also due to the fact that electronic journals have hardly made their way into the information infrastructure called libraries. Few libraries have as yet incorporated electronic journals in their catalogues, and the reports of libraries that are struggling with the problem give the impression of very diverse solutions [16, 23, 24, 25, 26]. Also, electronic journals are hardly covered in conventional abstracting and indexing services [10]. Part of these problems can in turn be attributed to a lack of standards in the area, which means that there is no single solution which applies to even the few electronic journals that exist today. Modes of access for instance, vary from automatic e-mail, e-mail of table of contents with the possibility of requesting separate articles by e-mail, to ftp, and the networked information retrieval tools Gopher, WWW and WAIS. This in turn brings up the question whether articles should be downloaded (and sometimes printed to be shelved like their print counterparts !), and if so to which platform, or if should be relied on networked access. In the latter case, userfriendliness of different solutions is questioned. Even something as simple as citing an electronic article is sometimes considered a problem [13, 27]. Rooks also observes that library personnel sometimes lacks training in coping with networked information [19].

A more fundamental problem raised by Metz and Gherman is that electronic journals could become a parallel to the printed journal system and thereby an extra burden, not alleviating but aggravating the serials price crisis. Only if universities develop policies to regain control of the information cycle can electronic journals be a solution [2]. This requires the cooperation of scholars which closes the circle.

Two related and very basic problems touch upon the quality control and archival function of journals. Printed articles have a big advantage from a library point of view: they are stable bibliographical entities which librarians know how to describe formally. In other words, the integrity and origin of a printed document is beyond any doubt. The integrity of a machine readable (and editable) computer file is not. A copy of an article can not be changed as easily as a computer file, although occasionally there arise problems in tracking bibliographic data of mutilated (hard)copies. Whether this is a real problem remains to be seen because there are technical solutions for this [28]. Because the problem mainly arises when files are copied from one machine to another it is important to have at least one site on the Net (which can be mirrored) responsible for maintaining the original, and the files themselves to contain a pointer to this originating site (and possible mirrors). This is all the more important if and when authors like to revise earlier versions of their papers, a natural thing to do with files, although not yet implemented in the now existing electronic journals. This implies a responsibility for version control and archiving on the part of the authors and publishers, which leads to an even more fundamental problem.

Although paper is not at all a perfect archival means (there are serious problems in conserving many publications from the nineteenth century) its archival properties are well known. Electronic publications though will inherently suffer from the fast developments in information technology meaning that future conversions of e- publications will be inevitable [28]. Standards could be helpful here, but it should be observed that if some of the advantages mentioned above, notably multimedia and hyperlinks, materialise, the very way in which they materialise could aggravate the archival problem if certain e-publications are too software dependent. This problem is also connected with the access problems mentioned above, and it becomes also more complex if a variety of formats is in use. A solution could lie in standardising on an SGML approach, also in use on the Web, since SGML, based on simple ASCII as it is, guarantees maximum portability of documents. But the conditions remain anarchic and it remains to be seen in which way SGML develops. If the problem is not solved, there is a genuine threat of destruction of information if porting of documents over generations of systems becomes to costly an affair.

Library problems with electronic journals

Libraries are allocated three functions in the present day, mainly printed information chain. They select, acquire and store information items from the published universe in order to develop a collection to satisfy there current and future patrons needs. They describe these information items formally (title, author, publisher, year etc.), by content in order to make retrieving by subject (the `aboutness' of an information item) possible, and add a pointer to the thus obtained catalogue records so as to enable retrieval of the physical information object. Finally the information items are stored and made available on request of their patrons or other libraries through Inter Library Loan. In this way, libraries play an important role in fulfilling the functions of the journal. Via their catalogues the communication function is supported although in the case of periodical literature abstracting and indexing services, online or on CD ROM, produced by secondary publishers play a far more important role. By selecting information items to add to their collections the quality control function is in a certain way enhanced. The most important contribution of libraries lies though in their archival function which in a sense supports again the communication function, but now between generations of scholars. How do electronic journals affect these three library functions ?

The first function is least affected, since the process of selecting information depends on the content of information rather than the medium. Libraries have incorporated new media like microfilm, music on different media, and videotapes in their collections with little problems. The selection process is steered by the information needs of their patrons, subject to collection development policies, and is often not dependent on the type of medium. Most electronic journals have the advantage of being free of subscription, but it should be remembered that they still cost money in terms of equipment and, more important, of library personnel. Whether electronic journals remain free of subscription if they would replace or substitute printed journals on a large scale is unlikely though. Also it is important to keep in mind that the decision to `collect' an electronic journal could become dependent on the technical state of the art of the library involved. Another danger could develop in the case of commercial journals with proprietary interfaces like the two Online Journals mentioned earlier. Now that the rise of IR protocols like Z39.50 and ISO Search and Retrieve promises the possibility of uniform interfaces to different database engines, a proliferation of interfaces to electronic journals would only confirm the rule that what people learn from history is that they don't learn from history. Of course this point is closely connected to the question which economic model will evolve.

The difference arises with the other two functions of libraries because the type of medium changes fundamentally from physical to non-physical. A library catalogue is usually limited to a description of information items which are physically present in that library. Electronic journals and articles are not physically present in a library, not even if they are stored on a local library computer since in a networked world there is hardly a difference between disks mounted on a local machine and disks mounted on a remote machine (although there is an important administrative difference in case the location of files changes, but a solution can be found if a scheme of stable URI's connected to changing URL's is followed). Besides, the problem is not just one of having a local or remote network address for a given electronic journal, it is also one of modes of access, which already now vary, and with every innovation introduced will vary even more. The basic problem here is that innovations seldomly substitute but more often add to the world of existing possibilities: radio did not disappear when television was introduced, fax did not replace telex, and electronic mail will probably not eliminate fax. Existing cataloguing formats and rules offer little help in describing the network-pointers and methods of access to electronic files [15], so libraries seek a solution in extensive annotations [26]. Library organisations are dealing with this problem [29] but the question is whether the library community will be able to keep up with new developments popping up at a seemingly ever increasing rate on the Internet. Migration of an electronic journal archive to a different format or mode of access implies a mutation in a catalogue record. Apparently, libraries are not the only organisations struggling with these problems, abstracting and indexing services also hardly cover electronic journals, yet.

As to the third library function, making information items available on request, the problem is one of integration, and here again the varying modes of access and formats are a crucial problem. Navigating tools like Mosaic offer an integrated solution with an easy point and click interface for different formats, but surely Mosaic will be succeeded by another tool, let alone the interesting question whether Mosaic should be considered a navigation tool or a browser, which is certainly a less pretentious label. And again, technical limitations might dictate collection development policies.

The conclusion that "there is no clear sense of how libraries should handle electronic journals." [16] is quite obvious. Also, the impression seems justified that most users of electronic journals find out about them directly and bypass the library. Libraries and abstracting and indexing services are clearly not yet ready for this new medium and are in fact also not needed by scholars, yet. This might change if electronic journals take over the load of printed journals on a large scale. For 8 math e-journals, a mathematician doesn't need a library or Mathematical Reviews, in fact she can easily build her own virtual library using for instance the Mosaic "Hotlist" feature. This will change if the number of electronic journals grows considerably. The final section tries to analyze this potential.

The transformational potential of electronic journals

Manoff might be right in observing that "it is electronic journals that will become the focal point of the battle between universities, as they attempt to recapture their intellectual property, and publishers, as they attempt to maintain their economic dominance." [15]. Turning back to the three functions of journals mentioned above this section tries to evaluate if electronic journals stand a chance as a networked successor of their printed counterparts and what the role of the major players in the information chain could be in terms of enabling or inhibiting the transition.

As to the function of communication, the electronic journal seems to offer the greatest advantage. Whether this is a real advantage remains to be seen because speed is not everything. Communication also asks for visibility and retrievability. The first observation is that more critical mass is needed. As long as the phenomenon is quite insignificant there is little pressure for integrating electronic journals into the changing information chain. Integration is needed because printed journals will not disappear all of a sudden because of the emergence of an electronic alternative. More critical mass could be generated by upgrading the existing electronic preprint services to electronic journals, simply by adding more peer review. Most preprint series already have a moderate form of peer review and in some disciplines printed articles hardly seem to have a communicative function any longer, they are `just for the record', an expression which reminds of the archival function. An alternative would be to encourage the establishment of new electronic journals. The short history of electronic journals learns that this is dependent on enthusiastic, reputed and visionary scholars like Harnad. Scholarly societies and university presses could also play an enabling role, while tenure policies could be changed to give electronic publications as much value as print publications. All these factors are dependent on scholars and their organisations. Libraries and library organisations should try to integrate electronic journals into the existing information infrastructure designed for printed journals. This implies cataloguing electronic journals, incorporating them into reference databases for periodical literature where they exist and try to provide either direct access from those reference databases or indirect via a link to the full text of the articles. Such integrated systems can be designed to incorporate different formats [21], and can further be enhanced with alerting services. This goes way beyond setting up a Gopher or Mosaic page, which seems to be the current solution and which lacks retrieval possibilities and does little to support generic literature searches. In this sense a Gopher reminds very much of a library shelf, where users can come and browse and find something worthwhile if they are lucky. An alerting service based on knowledge of what patrons need in their day to day work, drawing from a vast and ever changing pool of (networked) resources, coupled with easy and fast document delivery is perhaps the only way libraries can survive. Rosenfeld and Janes recently pointed out that in order to navigate the Net there is a need for travel agents [30]. Libraries are natural travel agents given their expertise in selecting information, i.e. separating the chaff from the wheat and thereby assisting users to determine the usability of information which is also an important task on today's network with its low signal/noise ratio; map and disclose the available information in an integrated way, staying comprehensible for their patrons; and offer access to scholarly communication, whatever the medium. These are traditional functions of a library and they will remain important in a virtual library world. Libraries could go further and play a supportive role in the publishing and editorial process [20], perhaps in close cooperation with university presses and learned societies where they exist, offering their expertise in database building and, more important, database maintenance, and networking.

Publishers have an enormous advantage because of their long standing experience in the printed information chain and because of the existing editorial infrastructure they control. Publishers are slowly moving into the direction of electronic publications and are experimenting on a small scale. They face two basic problems. They first of all fear a loss of income because electronic publications are not easily copyright protected, whereas scholars hardly care about, nor receive, copyright on periodical publications, they care about being published [31]. Libraries are by nature also more interested in communication than copyright. There are technical and legal solutions to the copyright problem but these inherently inhibit the communication function of journals by their very nature. The second problem publishers face has a cultural and infrastructural nature: their whole production process is oriented towards printed products of high quality, and although they are reorganising towards a more electronic environment, this is a complex task for which many small publishers lack expertise and the few great publishers suffer from their size. This problem is aggravated because publishers will try to stay as long as possible in the printing business implying that they too, like libraries, suffer from a parallel system. The basic point is though copyright and publishers are the only ones who have a stake. The only justification for copyright is added value.

Interesting roles could be played by intermediary organisations like subscription agents and national library automation organisations like OCLC and PICA. Subscription agents are moving towards supplying reference databases and individual article supply (e.g. Swets/Swetscan and UnCover). They are in a good position to enhance these services with electronic articles. The only problem is of course that they can hardly charge for electronic articles if these are also available for free on the Net, like the majority of electronic journals right now is. Whether this will remain so is a matter which outcome depends on the question if a viable economic model for electronic publications can be designed. Again the question is where the added value is. This seems to imply that library automation organisations are in a slightly better position with services like OCLC's FirstSearch and PICA's national Online Contents database which can easily be enhanced to incorporate electronic articles. If they will be is probably more a matter of overcoming the organisational slack which seems inherent to these institutions, and the cost they have to charge their member libraries for such a service.

As to the function of quality control, the initiative is with the scholars themselves as they already perform this function in the present day information cycle. If the awareness among scholars grows that the serials pricing crisis is not just a problem for librarians and that the Net offers the potential for transforming the information cycle, then a basic condition for transition will be fulfilled. Libraries today can stimulate this awareness by showing the potential of electronic journals and incorporate them into their information infrastructure along the lines indicated above. Quinn suggests that libraries should take on the role of publishers and take responsibility for the editorial process [20]. This will depend much on the particular situation at an institution. Stimulation by university boards, national organisations of universities and learned societies could be of great importance.

In the area of quality control publishers have a real advantage not only because of their long standing knowledge of organising the refereeing process and attracting reputed scholars as editors, but also because they control the prestigious (printed) journals of today which have a competitive advantage over the electronic newcomers in attracting the best articles. Harnad, editor of the electronic journal Psycoloquy, has shown though that an electronic journal can gain prestige, but this achievement seems not to have been reproduced yet.

The archival/integrity problem is a difficult one but by no means a problem which cannot be solved. Here libraries can offer contributions, whether or not on in the role of publisher. Building on a long standing expertise with print publications they understand the issues at stake, and can work together with IT professionals to offer solutions. In this area it seems best to start with low level technology, implying that interesting experiments with for instance hypertext links should be regarded with reservation. Here libraries are in an advantage since none of the other parties in the present day information chain have any expertise in the area.

Conclusion

Electronic journals offer certainly possibilities for a transformation of the information chain. In a sense, this transformation process is already underway and electronic journals are but one instance of this process, other examples being electronic preprint services and listservers and (mainly sci.*) newsgroups on Usenet. The important difference is that electronic journals should not just be a medium for scholarly communication but should also serve the functions of quality control and archiving. The question seems not to be whether a transition from printed to electronic journals will take place, but rather how fast this process will be and if it can be managed. Quinn argues strongly for a managed transition because evolution is a slow and expensive process and the serials pricing crisis is diminishing quickly the role of libraries which should in his opinion be managers of the process. Furthermore he is afraid the present system might collapse from pressures of preprint services [20]. Harnad on the other hand seems to intend to quicken the process by which the printed information chain could collapse. In a very lively discussion he started early in the summer of 1994 on several listservers he proposes a subversive action in which every `esoteric' (i.e. scholarly, not for the market) author establishes a globally accessible ftp site for her preprints, and refuses to withdraw these preprints from the ftp archive after the article has been accepted for publishing, after peer review [31].

Which ever course will be taken, librarians are advised to follow and participate in the discussion if they want to stay in business. This article was intended as a contribution to that discussion and if it were not for a promise the author made to the editor it would have been submitted to an electronic journal (insert smiley).

References

[1] Derek J. de Solla Price, "Little Science, Big Science", Columbia University Press, New York, 1965.

[2] Paul Metz and Paul. M. Gherman, "Serials Pricing and the Role of the Electronic Journal", College & Research Libraries, vol. 52, no. 4, pp. 315 - 327, 1991.

[3] Bruce C. Bennion, "Why the Science Journal Crisis?", Bulletin of the American Society for Information Science, pp. 25 - 26, March 1994.

[4] Karen Hunter in: Ann Marie Cunningham and Wendy Wicks (eds.), "Three Views of the Internet", NFAIS Report Series no. 3, NFAIS, Philadelphia, PA, 1993.

[5] Jack Kessler, "Directory to Fulltext Online Resources 1992", Meckler, Westport, 1992.

[6] Judy E. Myers et. al., "Surfing the Sea of Stories: Riding the Information Revolution", Mechanical Engineering, vol. 114, no. 10, pp. 60 - 65, 1992.

[7] B.T. Stern and R.M. Campbell, "ADONIS, Publishing Journal Articles on CD-ROM", Advances in Serials Management, vol. 3, pp. 1 - 60, 1989.

[8] Melia M. Hoffman et. al., "The RightPagesTM Service: An Image Based Electronic Library", Journal of the American Society for Information Science, vol. 44, no. 8, pp. 446 - 452, 1993.

[9] Joost Dijkstra, "A Digital Library in the mid-Nineties: Ahead or on Schedule, Information Services and Use, forthcoming.

[10] Charles W. Bailey Jr., "Networked-based Electronic Serials", Information Technology and Libraries, pp. 29 - 35, March 1992.

[11] Okerson, Ann, "Introduction to 1994 directory", gopher://arl.cni.org.

[12] Ann Okerson, "The Electronic Journal: What, Whence, and When?" The Public-Access Computer Systems Review vol. 2, no. 1, pp. 5-24, 1991. To retrieve this article send e-mail to LISTSERV@UHUPVM1.UH.EDU, body of the message should read GET OKERSON PRV2N1 F=MAIL.

[13] Alan Singleton, "Electronic Journals for Everyone?", Physics World, pp. 27 - 31, November 1993.

[14] Andrea Keyhani, "The Online Journal of Current Clinical Trials: An Innovation in Journal Publishing", Database, pp. 14 - 23, February 1993.

[15] Marlene Manoff et. al., "Report of the Electronic Journals Task Force MIT Libraries", Serials Review, vol. 18, no. 1, pp. 113 - 129, 1992.

[16] Thomas E. Nisonger, "Electronic Journals: Post-Modern Dream or Nightmare: Report of the ALCTS CMDS Collection Development Librarians of Academic Librarians Discussion Group", Library Acquisitions: Practice and Theory, vol. 17, no. 3, pp. 378 - 380, 1993.

[17] Stevan Harnad, "Implementing Peer Review on the Net: Scientific Quality Control in Scholarly Electronic Journals", 1993 in: "International Conference on Refereed Electronic Journals: Towards a Consortium for Networked Publications." University of Manitoba, Winnipeg 1-2 October 1993". ftp://ftp.cc.umanitoba.ca/e-journal. Or follow this link.

[18] Stevan Harnad, "Post-Gutenberg Galaxy: The Fourth Revolution in the Means of Production of Knowledge", PACS Review, vol 2, no. 1, pp. 39 - 53, 1991. To retrieve this article send e-mail to LISTSERV@UHUPVM1.UH.EDU, body of the message should read GET HARNAD PRV2N1 F=MAIL

[19] Dana Rooks, "Electronic Serials, Administrative Angst or Answer", Library Acquisitions: Practice & Theory, vol. 17, pp. 449 - 454, 1993 (Paper presented at the Texas Library Association Pre-Conference, "Electronic Access to Serials", San Antonio TX. March 9 1993).

[20] Frank Quinn, "A Role for Libraries in Electronic Publication", EJournal, vol. 4, no. 2. To retrieve this article send e-mail to LISTSERV@ALBANY.BITNET, body of the message should read GET EJRNL V4N2.

[21] Hans Roes and Joost Dijkstra, "Ariadne: the Next Generation of Electronic Document Delivery Systems", The Electronic Library, vol. 12, no. 1, pp. 13 - 20, 1994.

[22] Maurice B. Line, "The Publication and Availability of Scientific and Technical Papers: an Analysis of Requirements and the Suitability of Different Means of Meeting Them", Journal of Documentation, vol. 48, no. 2, pp. 210-219, 1992.

[23] Marlene Manoff et. al., 'The MIT Libraries Electronic Journals Project: Reports on Patron Access and Technical Processing", Serials Review, vol. 19, no. 3, pp. 15 - 40, 1993.

[24] Gail McMillan, "Technical Processing of Electronic Journals", Library Resources and Technical Services, vol. 36, no. 4, pp. 470 - 477, 1992.

[25] Colleen Thorburn, "Cataloging Remote Electronic Journals and Databases", Serials Librarian, vol. 23, no. 1/2, pp. 11 - 23, 1992.

[26] Lawrence R. Keating et. al., "Electronic Journal Subscriptions", Library Acquisitions: Practice & Theory, vol. 17, pp. 455-463, 1993.

[27] Michael E. Stoller, "Electronic Journals in the Humanities: a Survey and Critique", Library Trends, vol. 40, no. 4, pp 647 - 666, 1992.

[28] Peter S. Graham, "Intellectual Preservation in the Electronic Environment" in: Arnold Hirshon (ed.), "After the Electronic Revolution, Will You Be the First to Go?", American Library Association, Chicago and London, 1993.

[29] In the Netherlands the Royal Library has set up some guidelines for the cataloguing of resources on the Internet, see "Basisrichtlijnen catalogiseren Online Resources in de GGC", gopher://gopher.konbib.nl/M/kbarchief/gedocarchief. This work seems to draw on an OCLC project. See for a summary of this "OCLC Internet Cataloging Experiment", gopher://gopher.bubl.bath.uk, Main Menu, section BH1C.

[30] Joseph W. Janes and Louis B. Rosenfeld, "And Magellan Thought He Had Problems: `Navigation' in a Network Environment", LIBRES: Library and Information Science Research Electronic Journal, vol. 4, no. 1, 1994. To retrieve this article send e-mail to LISTSERV@KENTVM.KENT.EDU, body of the message should read GET LIBRE4N1 JANES.

[31] Harnad has archived this discussion, which is taking place on several listservers and via private e-mail in a publicly accessible file

Author's note

Hans Roes is deputy librarian for collection development and information services at Tilburg University Library, the Netherlands. This article is a revised and expanded version of a paper presented at the INET'94/JENC5 conference, Prague, June 1994, proceedings edited by B.R. Plattner and J.P.A. Kiers. The author wishes to thank Marijke van der Ploeg for assistance in a literature search on the subject, and Jola van Luyt-Prinsen, Hans Geleijnse and two anonymous referees for their valuable comments on earlier drafts of this paper. Responsibility for the content rests though solely with the author. The views expressed are those of the author alone.

Appendix. Sample of 39 refereed scholarly electronic journals

American Mathematical Society Bulletin
Arachnet Electronic Journal on Virtual Culture
Architronic; The Electronic Journal of Architecture
Catalyst; The Community Services Catalyst
CLIONET: an Electronic Journal of History
Digital Technical Journal
Education Policy Analysis Archives
EDUCOM Review
EJOURNAL
Electronic Journal of Analytic Philosophy
Electronic Journal of Combinatorics
Electronic Journal of Differential Equations (EJDE)
Electronic Journal of Sociology
Electronic Letters Online
Electronic Transactions on Numerical Analysis
Federal Communications Law Journal
Indiana Journal of Global Legal Studies
Information Technology and Disabilities
Interpersonal Computing and Technology
Issues In Science and Technology Librarianship
Journal of Artificial Intelligence Research (JAIR)
Journal of Political Ecology
Journal of Statistics Education
Journal of Technology Education
Journal of the International Academy of Hospitality Research
LIBRES
Murdoch University Electronic Journal of Law
New Horizons in Adult Education
New York Journal of Mathematics
Online Journal of Current Clinical Trials
Online Journal of Knowledge Synthesis for Nursing
Public-Access Computer Systems Review
Postmodern Culture
PSYCHE; An Interdisciplinary Journal of Research on Consciousness
PSYCOLOQUY
SOLSTICE
Surfaces
TESL-EJ; Teaching English as a Second or Foreign Language
Ulam Quarterly