Electronic Journals: A Short History and Recent Developments

Hans Roes
Tilburg University Library
June 1996

Paper presented at the International Summer School on the Digital Library, Tilburg, August 1996.

Introduction

This paper, to be presented at the International Summer School on the Digital Library, August 1996 at Tilburg University, is a sequel to and an extension of an article written in 1994 [Roes, 1995] on the subject of electronic journals. It first sets outs to give a very short account of the evolution of electronic journals, clearing some definitional issues in passing. The first section ends with an overview of what is available on the Internet at the time of writing (Spring 1996) of this paper. Next attention is paid to directory services on the Internet which point to existing electronic journals and alerting services for new electronic journals. The following section deals with library experiences with electronic journals, paying specific attention to a recent initiative at Tilburg University Library aiming at an integration of Internet resources in general into the existing library infrastructure. The next section discusses possible consequences for the scholarly communication process in terms of scenarios and key problems in these scenarios. Finally an assessment is given of the current state of development of electronic journals and some tentative conclusions are drawn.

Evolution of the electronic journal

Lancaster [1995a] distinguishes between four manifestations in the evolution of electronic publishing over the last thirty years:

Use of computers in the preparation of printed publications, starting in the early sixties.
Distribution of text in electronic form of exact equivalents of paper versions. Examples are the ADONIS, Red Sage and TULIP projects.
Distribution in electronic form only with a few 'value added' features like searching and alerting.
Completely new types of publications exploiting hypermedia possibilities.

It is important to note that these manifestations are logical - and not chronological - steps. In practice all of these forms (still) occur. Lancaster chooses a strict definition for an electronic journal as "one created for the electronic medium and available only in this medium." If we add, like Lancaster and many other writers on the subject do, the requirement of peer review, the first scholarly electronic journal can be dated back in 1979. These early experiments did not last long. A review carried out by the author in 1994 [Roes, 1995] dated the oldest then existing electronic journal in 1987. Lancaster [1995a] mentions as the most important reasons for failure:

not enough readers and authors with the necessary equipment
other technological barriers, notably bandwidth and poor graphic abilities of computer screens
potential authors could see no obvious rewards

While the first two factors have diminished in importance, although not equally among the various disciplines in the scholarly community and not equally on a global scale, the third factor is still impeding the development of the electronic journal.

Nevertheless, literature on the subject has mushroomed since the early nineties as illustrated by Charles Bailey's bibliography [1995] on the subject. This growth in the literature can be explained by the advantages electronic journals offer compared to the dominant print medium in the present day information chain. These advantages of electronic journals can be summarized as follows:

electronic journals allow for more speed in the scientific communication process
electronic articles are not limited in size
it is possible to add experimental data, software, and even multimedia extensions like simulations
articles can be linked with the body of scholarly literature
there is a promise of a more open peer review process
electronic journals demand no library space nor shelving costs nor can they be stolen from the library
electronic journals are available 24 hours a day, 7 days a week
electronic journals can easily be merged with alerting services

Most promising is the expectation that electronic journals are a solution to the serials pricing crisis, which is gradually changing and perhaps eroding the role of libraries. The central notion here is that electronic journals can potentially be cheaper produced and distributed than their printed counterparts. But the cost of an electronic journal is widely debated.

One of the most important factors in the development of electronic journals in the past two years is the emergence of the World Wide Web as an enabling technology [Weibel 1995]. Use of the Web is widespread and it offers the possibility to realise many of the theoretical advantages associated with electronic journals, especially hyper- and multimedia and linking with the body of literature. What than, is the true size of the phenomenon at the moment, compared to the number of existing printed journals which can be safely estimated to be well over 100,000 ?

The 1994 review of the author revealed a number of 39 peer reviewed electronic journals [Roes, 1995]. A recent study by Hitchcock et al. [1996], concentrating on STM journals, counted 83 journals of which 44 first appeared in 1995 (the perfect match in numbers is merely coincidental since the two samples differ somewhat in focus). A sample from Harter [1996] counts 131 electronic journals of which 77 are peer reviewed. These numbers are insignificant compared to the number of existing printed journals, but 1995 has shown an accelerating trend. If for instance the number of electronic journals doubles every year - not unlikely given the above numbers, then starting with a number of 100 would give a number of 100*(2^10) = 102,400 electronic journals in 10 years time. Apart from illustrating the astonishing character of exponential growth rates, this points to the fact that electronic journals have the potential to significantly alter the information chain in the near future. Librarians should pay attention, and, better even, try to influence developments.

The Hitchcock [1996] sample shows that a majority of the electronic journals are 'electronic editions' and that a significant number of them require paid subscriptions or will do so in the near future. This is correlated with the fact that in the Hitchcock sample most electronic journals come from commercial publishers. Also, most of the potential advantages, mentioned in the previous section, are hardly realised, which corroborates the findings of the author in his 1994 study.

Another way to illustrate the importance of the development is to look at the activities of commercial publishers. A short overview of current projects by Taubes [1996] mentions Elsevier's Electronic Subscriptions (EES) aiming at making all the 1100 Elsevier journals available electronically, while John Wiley (326 journals), Blackwell (125 out of 200 journals by September 1996) and Academic Press (175 journals) are also trying to get a head-start. At the same time, secondary publishers like ISI [Kimberley, 1995] and Swets are also trying to get into the electronic publishing business. These are all manifestations of Lancaster's second kind, implying electronic versions of printed papers. But they are probably also necessary steps towards commercial journals which more fully exploit the possibilities of the electronic medium. Elsevier for instance is planning some fully electronic journals (Gene-COMBIS and New Astronomy) to experiment with the new medium.

Another related development which should be mentioned here are electronic preprints (or e-prints). The most spectacular example is of course Paul Ginsparg's server which started in high energy physics in 1991 but has gradually broadened its scope and now serves a community of over 35.000 researchers, processing over 70.000 transactions per day [Ginsparg, 1996]. Similar examples are gradually evolving in other disciplines with a preprint tradition, one such an example is the Tilburg / Maastricht Grey Files project also in the program of this Summer School. The Grey Files project will be scaled up to a national distributed service for e-prints in economics.

All these developments illustrate that there are some profound and, most likely, irreversible changes taking place in the information chain. The first major group of players in these developments are scholars like Paul Ginsparg - founder of the Los Alamos preprint archive - and Stevan Harnad - founder of one of the first electronic journals, Psycoloquy. This points the way to greater control for scholars. The second major class of players are commercial publishers which try to defend their interests. Libraries are mostly standing by and watching, which is not a wise attitude.

Electronic Journals: where to find them and how to keep track

There are several directory services on the Internet / World Wide Web which try to keep up with new developments in the field of electronic journals. On of the oldest is the Directory of Scholarly Electronic Conferences maintained by the Association of Research Libraries (ARL). The ARL directory is connected with the NewJour-L list, a listserver devoted to alerting subscribers to new electronic journals. Hints for subscribing to this list, which at the moment points to about ten new journals per day, as well as past contributions to the list, can be found at the NewJour Archive. A problem with both the ARL directory and the NewJour archive and list is that it is hard to distinguish between newsletter and magazine type of publications and scholarly, peer reviewed journals.

Another worthwhile source for learning about electronic journals is the WWW Virtual Library: Electronic Journals. Yet another possibility is to look at the efforts some libraries have undertaken to chart the territory. One such an example is the library of the University of Houston which maintains the page Scholarly Journals Distributed Via the Web.

For those who are more interested in journals and other Internet resources pertaining to a particular subjectfield, FAQs, Frequently Asked Questions (and answers, of course) can be rewarding. One such an example in the field of economics is the FAQ Resources for Economists on the Internet (look for online journals).

For those who are interested in discussions about electronic journals, the HyperJournal Forum offers excellent possibilities. It also points to general resources on electronic journals, as well as archives of past debates.

Library experiences with electronic journals

Whether or not libraries should collect electronic journals is not the question. If libraries want to play a role in the information chain, they have to collect and disclose information independent of the type of medium [Woodward, 1994]. The problem is that the new medium poses a lot of new questions, and, given the lack of standards [Weibel, 1995], there is no simple answer to any of these questions. Important factors to consider are storage, access, selection, acquisition, bibliographical control and training of library staff [Woodward, 1994]. Bailey's [1995] bibliography shows that the literature on the subject of library experiences with electronic journals is scant, while the majority of the articles mentioned deals with technical problems of acquisitions and processing. The main problem though is one of integration [Roes, 1995; Weibel, 1995]. One of the most adopted solutions nowadays seems to be to build an electronic equivalent of journals shelves: Web pages with journal titles ordered alphabetically and / or by subject, offering users merely access by browsing. This is way below the capabilities of what a library can and should do, although it must be admitted Tilburg University Library hasn't come up with a better solution yet. But there is some interesting work in progress, which is also partly addressed in the lectures by Joost Dijkstra (acquisitions), Thomas Place (cataloguing of electronic documents) and Corry Stuyts (Grey Files).

The key word in this effort is integration. This integration has been achieved as far as the Elsevier journals - which are electronically available - are concerned and as far as the electronic preprints produced at Tilburg University are concerned. From the Online Contents database (journals), or the Attent database (preprints), a user can click a button and go transparently to the full text of the article or paper to which she found a reference. This same method should apply to electronic journals available over the Internet. If such an electronic journal qualifies in the selection process as being of importance to our user population, then such a journal should be treated in the same way as if it were a paper journal. The acquisitions department should keep track of new issues / articles appearing on the Internet connected to the selected journals. The title of the journal itself should be catalogued in our Online Public Access Catalogue (OPAC) and the articles should be described in our Online Contents database or in any other dedicated database like Excerpta Informatica for computer and information science. The philosophy behind it all is that users should not be bothered with new forms of information, but rather to make new information available regardless of the form it takes. The infrastructure with which the user is familiar should remain the same, or, better stated, should follow it's own course of development. Since most electronic journals are available through the World Wide Web, this implies a search and retrieval interface implemented in a Web browser like Netscape.

Of course there are problems with this approach. A first problem has to do with the question what type of document we are dealing with. If it is an electronic dissertation, a catalogue description in the OPAC is called for. An electronic article is described in Online Contents, an electronic preprint in Attent. These are all types of electronic documents which bear close resemblance with their printed predecessors. But what about an interesting World Wide Web site, for instance the homepage of the United Nations Organization ? Could we call this a document ? That's doubtful, but suppose we would call it a document, what would be the most appropriate place to catalogue it ? Reference databases are designed for the description of stable and finite documents, and a homepage hardly qualifies on any of these characteristics.

Probably this problem will aggravate in the future as the notion of a document will be undermined by exploiting the hypermedia possibilities of the World Wide Web. Again, our catalogues and reference databases are based on the notion of stable and finite documents. Hypermedia documents will not be stable but rather subject to change, and the question of where the boundaries of a hypermedia document lie, which can be linked to other hypermedia documents which are in turn linked to other documents etcetera, is an arcane one. A more down to earth problem will be payment for copyrighted documents. How are we going to budget that ?

A rather fundamental problem for libraries has to do with archiving, which is one of the present central functions of libraries in the scholarly communication system. Electronic publications suffer inherently from fast developments in information technology. This problem seems to become more manageable now that most electronic journals are appearing on the World Wide Web. The Web's language, HTML - an SGML document type definition (DTD) - has the advantage that as long as the document type definition is known, any SGML parser can make sense of the documents. Also there is a upward compatibility in different HTML versions, although private extensions to HTML, like Netscapes, might create problems. Another problem not yet solved is the stability of URLs. Web documents can change place without notice which in turn can make an electronic document practically irretrievable. Solutions for this problem, Uniform Resource Identifiers, are underway [Weibel, 1995] but the key problem here is that there is no clear vision as to which party in the information chain is responsible for archiving. This is yet another reason for libraries to be involved in the development of the electronic scholarly communication system, since libraries already perform this task in the present day printed system.

A final concern which is often mentioned by librarians is access. Scholars need to be connected to the Internet and need to have the right software and skills to make use of the electronic body of literature. Quite often the fear is expressed that the Third World will not have adequate access to the world's information resources. Odlyzko [1995] and Ginsparg [1996] very properly remark that the cost of building and maintaining a large library collection far exceeds the cost of installing networked computers and that in fact scholars in the Third World will be better of with electronic journals. This argument though depends quite strongly on the type of economic model which evolves. Which brings us to the question of the future of electronic journals and the information chain.

Impact on the information chain

This section deals with possible scenarios first. Next the key problems in these scenarios are dealt with, while finally an assessment is made of the current state.

Scenarios

In an overview of possible scenarios for the future of the journals system, Bailey [1994] observes that many scenarios can be better characterised as large scale blueprints for alternative scholarly communication systems. His own view is that electronic journals will start out as a complementary development which will gradually replace the role of printed journals. This is probably the safest assumption, but a lot of discussion remains on wether the change will come about abruptly or rather more smoothly.

Odlyzko [1995], in a paper which caused a lot of discussion among librarians and publishers, sees the development of the electronic journal as the inevitable outcome of two forces. One is a technology pull, more and more tools are becoming available for scholars to run the publishing business themselves. Ginsparg's preprint server supports that notion, although it also shows that the development will be uneven among the various scholarly disciplines. The second force is an economic push caused by the exponential growth in scholarly literature. This is, in Odlyzko's view also the basic factor behind the serials pricing crisis. In mathematics alone, Odlyzko's discipline, the production of new papers is estimated to be 50.000 per year. No library in the world will be able to cope with such growth rates. The second factor here is that the average article will be read by a constant number of scholars, and this number is not likely to rise. On the other hand it is important for scholars to have potential access to the whole body of literature produced in mathematics, which poses an impossible task to libraries, at least an impossible one within the paradigm of the printed information chain.

Odlyzko sees a central role for preprint servers to cope with the pressures mentioned above. Through these preprint servers, papers will be freely available which implies that libraries will be less needed by scholars. Consequently there will be less pressure on libraries if they plan for yet another round of cancellations in subscriptions. These cancellations drive up the price of journals since the fixed costs of these journals have to be covered over a smaller number of copies, which will again lead to new cancellations etcetera. Basically, Odlyzko scenario is the same as the scenario librarians are familiar with since the phrase serials pricing crisis was coined. The only difference is that the vicious loop is started from the outside because of the existence of preprint servers.

An interesting extension to this scenario is offered by Harnad's subversive proposal (summarised by Brent [1995]). In this proposal scholars themselves are the driving force in the change as they are called upon by Harnad to make their printed papers available electronically instantly after they have been accepted for publication in printed journals. Thereby the need for printed journals would collapse almost immediately. In that way publishers would be forced to radically alter their business away from print and into an electronic model which, in Harnad's view, is a lot cheaper.

Both in Harnad's and Odlyzko's view there is a possible role for publishers if they are prepared to fundamentally change there mode of work. As to the role of libraries, Odlyzko [1995] is very clear: "Technology will solve the librarians' problem, but will also eliminate most of their jobs." In his view there is no need for a library if the whole body of scholarly literature is available from scholars' desktops. The central notion here is that libraries exist because scholars need them, which is a truism indeed. It should be noted that in a more recent publication, Odlyzko [1996] sees a diminishing role for libraries as collectors of books and journals, but an increasing role for libraries as gateways to outside information sources. He now recognizes the skill of libraries in the organising information spaces.

Central players in these scenarios are always scholars, since they are always the beginning and the end of every possible form the scholarly information chain takes [Tenopir 1995]. The role of the other players in the information chain is dependent on wether they will be able to offer value added. In some analyses it is noted that universities and libraries can play other potential roles than nowadays. Schauder [1994] for instance sees a role for librarians as literary agents for scholars seeking publishing outlets. Lancaster [1995a] stresses the fact that universities and university presses have lost control over the publication process and now have the opportunity to regain some of that control. Similar, but a bit more bitter remarks about university presses and scholarly societies are voiced by Lynch [1994] and Gasaway [1994]. Central in these notions is that crises are also implicit opportunities. But libraries and academic administrators must be willing to recognise these opportunities and turn them into strategic actions.

Key problems

In the meantime, developments, although they can gain an enormous momentum in the next few years, are still slow. This is because there are some key problems for which a clear and simple solution has yet to arrive.

One of those key problems is the question of the economic model. Graham [1996] gives a quick summary of six possible economic models but doesn't arrive at a clear solution. Printed journals are based on subscription models which could be transferred to the electronic realm in the form of site licenses which effectively kill the need for personal subscriptions, since scholars can access the electronic journals from any place in the world as long as their institution has a site license. Pay per view systems are also proposed but could carry prohibitive prices. On the other hand, Taubes [1996] remarks that site licenses could inhibit the interlinking of articles which is one of the key advantages of an electronic journals system, and a pay per view system would be needed to complement the site license model to make it possible to follow links to journals not captured in a particular site license.

The whole question of the economic model is linked with the question what the true costs of electronic publishing will be. Publishers insist that the cost of electronic publishing will not be lower than in a printed world because print costs are simply not the most important costs occurred. Odlyzko [1995], Harnad [Brent 1995] and Ginsparg [1996] disagree with this view as being papyro-centric and stress the need for a re-engineering of the publishing process in which scholars can play a far more important role. In their view this could lead to a model in which electronic journals could be, and should be, for free. Ginsparg adds that the low costs of maintaining servers which remain, could be covered by page charges to be paid by authors.

This leads to the conclusion that a subsidised model is the most likely one to evolve if scholars, backed by libraries and academic administrators, indeed take over the role of publishers. It is important to stress that the printed information chain nowadays also floats almost entirely on subsidies. But an important part of the subsidy is channelled into the market via subscriptions to journals published by commercial enterprises.

Another main question is therefore if publishers are prepared to cope with the problems intrinsic to the present day system and if they have enough incentives to offer to scholars to stay with them. Harrison and Stephen [1995] point out that new electronic journals have to cope with the problem that any new journal has - attracting authors - but also have to encourage scholars as readers and writers to adopt new styles of working. This point is reinforced by Schaffner [1994] who stresses the importance of the fact that authors need to have confidence in electronic journals. In her view enabling technologies are no sufficient condition to bring about a quick change in a system which has developed over more than 300 years. Peek [1994] also stresses the importance of psychological and sociological factors. Tenopir [1995] points to the difference in interests between scholars as authors and scholars as readers. She questions the willingness of authors to be subject to a system of open peer review and disputes the openness of the invisible colleges. In the present day system there is very clear distinction between formal and informal communication, and between authors and readers. These distinctions might be blurred in an interactive system and the question is valid wether authors would like such a change. Publishers as de facto owners of the present day journals have a competitive advantage because of these psychological and sociological factors.

Current state

Over the past two years some issues have become more clear and less academic. We have already mentioned the studies of Hitchcock [1996] and Harter [1996] which show that there is an acceleration in the growth of the number of electronic journals. Both studies also show that the role of commercial publishers in the development of electronic journals is growing. The Hitchcock sample shows that 33 of the 83 electronic journals in the sample are commercial ones. The Harter sample gives a number of 9 fee based journals on a total of 77 peer reviewed journals. Also both samples show a considerable share of electronic versions. We have also already mentioned the initiatives of publishers in launching electronic versions on a large scale. This points to a scenario were a dual system will coexist for some years. The problem with this development is that it is a costly one. As the TULIP Final Report [TULIP 1996] states: "At the moment, managing large digital collections locally, is harder and more expensive than managing a comparable print collection." At the same time it can be mentioned that libraries have spent more energy in developing this dual model than in solving the question how to integrate electronic journals only. This is illustrated in the TULIP project and in the Tilburg project with Elsevier (EASE).

These projects give important insights though in user behaviour with this new medium. As the TULIP [1996] report shows, desktop access to journal literature is well received by faculty and students, although graduate students are the most frequent users. Also the notion that paper will be here to stay for a while is corroborated in this study. Main problem in the TULIP project, and confirmed by experiences at Tilburg university is the lack of critical mass inherent in this period of transition.

The fact that paper is important to scholars is also confirmed by Olsen [1994] in one of the few in-depth studies of the use of electronic journals. Paper has certain advantages: people find it easier to read and concentrate on paper, they can highlight and annotate on paper, like to browse quickly through an article, like to work with a number of different articles at the same time. Olsen quite rightly remarks that any new system should be designed to support the creative process which is central to any scholarly work.

The impact of electronic journals at present is the subject of the already mentioned study by Harter [1996]. This study aims to measure the impact of electronic journals on the scholarly communication process by investigating the extent to which articles in electronic journals are cited and which type of articles in electronic journals are referenced. Unfortunately, at the time writing of this contribution, only the reference study part of the Harter paper is available. This reference study shows that, out of 4317 references counted in 279 articles in 74 peer-reviewed electronic journals, only 1.9 percent were references to electronic resources. Moreover, 80 percent of these references to electronic resources come from only three journals. The Harter study further examined these references and determined that only 52 percent were accessible, that is, they could be retrieved given the information contained in the references. The other half of the references could not be followed ! Given the age of the articles from which the references were taken, one to two years, this is a rather disturbing result. Harter concludes that at present electronic journals play almost no role in scholarly communication.

This observation leads us, finally, to articles about the potential of electronic journals as measured by a study of attitudes of librarians and academic administrators towards electronic journals by Lancaster [1995b], and a study by Schauder [1994] of ideas among academics about electronic journals.

Lancaster sent out 309 questionnaires to US academic librarians and administrators in November 1993. Over 70 percent of the librarians returned the questionnaire, while less than 40 percent of the administrators responded, yielding a total of 150 usable questionnaires. Main conclusions of Lancaster are that:

on the whole respondents were not optimistic that many of the possible advantages of networked publishing would actually be realised
respondents were not to confident about academics' ability to implement and manage a publishing network
establishment of a scholarly publishing network was the lowest priority for administrators and close to the lowest for library directors.

Lancaster is left with the impression that the whole idea is completely new to many of the administrators although there is a small group of enthusiasts. The impression that library directors are also quite unaware about the potential of electronic journals is rather more disturbing. Unfortunately no other studies of this kind could be found. And there is a possibility that attitudes could have changed over the last two years given the acceleration in growth.

Schauder received 582 questionnaires from academics in Australia, the US and the UK on the subject of (electronic) publishing. His main conclusion is that electronic publishing from the viewpoint of academic authors will give greater diversity and choice. Many academics use information technology to prepare manuscripts. Schauder notices a paradoxical motivation among scholars. On one hand their prime motivation is to publish in established and prestigious journals, on the other hand there is the tradition of free sharing of information. Prestigious journals, wether published by for- profit or not-for-profit organisations are generally more expensive than the less prestigious ones which are frequently subsidised. Schauder also found that there is little awareness among scholars about the consequences of ceding copyright to publishers.

Trends

The last two years have shown three remarkable trends. The first is that publishers are preparing to go electronic, or have already done so and are now scaling up their experiments to full blown services. These experiments mainly concern electronic versions but there are some interesting experiments with electronic only. Current projects add up to a couple of thousand journals. The next few years should show if the large number of small publishers which still operate in the market for scholarly journals will be able to join the leading publishers. In their effort, these smaller publishers will probably be aided by abstracting and indexing services and subscription agents which see new business opportunities.

The second trend is a modest acceleration in fully electronic journals which are also fully scholarly ventures. Whether the acceleration will continue and gain momentum in the next few years is unclear though. Awareness among scholars about the possibilities is still insignificant and motivation to go into their own publishing ventures is, in the present system, small. Academic incentives are still more geared towards publication in settled printed journals and there is little reward for setting up a journal.

The third, and perhaps most important trend is the steady growth in e-print servers as a complement to publishing in traditional journals. These servers satisfy the need for quick communication in relatively homogenous groups of scholars. But these servers are still mostly for informal publication, final results are still recorded in traditional journals. The next few years should learn whether it is possible to add a system of peer review to these servers which would further erode the role of the printed journals, with or without electronic versions.

If we look at the involvement of libraries in these three trends the conclusion is that they are mainly active in publishers' experiments with electronic versions, hardly involved in the development of truly electronic journals and very modestly involved in e-print projects. This can be explained by the lack of awareness of the possibilities with library directors and academic administrators as indicated by Lancaster's survey. The disappointing lesson is that libraries are missing an important chance to be involved in the ongoing restructuring of the information chain. The outcome of the process could be a cheaper and more accessible system which could restore the role of libraries in organising information spaces, a role which has diminished significantly over the past two decades. If libraries want to survive as active participants in the process of creating knowledge than they have to get involved and they better do it soon.

Hands on part

General Resources and Starting Places

Directory of Scholarly Electronic Conferences: ARL directory.
NewJour Archive: Archive of contributions to NewJour-L alerting service.
WWW Virtual Library: Electronic Journals: World Wide Web virtual library. Collection of electronic journals with special section for peer reviewed journals.
HyperJournal Contents Page: Discussions and resources on electronic journals.

Examples of electronic journals

Postmodern Culture: One of the "older" journals.
Gene-COMBIS: Elsevier's fully electronic journal on "Computing for Molecular Biology Information Service".
New Astronomy: Elsevier's fully electronic journal in astronomy, not yet started in June 1996.
The Astrophysical Journal: Electronic edition with lots of features. Articles available in HTML and PDF formats. Fully searchable.
The Public-Access Computer Systems Review: Peer reviewed journal featuring articles about electronic journals and other issues in developing digital libraries.
LIBRES: Library and Information Science Research Electronic Journal: Peer reviewed journal in library and information science. The journal has problems attracting publications.
D-Lib: Digital Libraries Magazine. Although of the magazine type very worthwhile with monthly issues.

Preprint Servers

Paul Ginsparg's e-print archive in physics: The first and largest public e-print archive.
Grey Files: E-print archive Tilburg University.
The International Philosophical Preprint Exchange: E-print archive in philosophy. Link might be slow.
NetEc homepage: Refers to WoPEc, electronic working papers in economics.
Economics Working Paper Archive: E-print archive, Washington University, St. Louis.

Library examples of incorporation

University of Houston.: A directory providing links to established Web-based scholarly journals that offer access to English language article files without requiring user registration or fees.
University of Sydney Library: Access by subject and alphabetically.
University of Pennsylvania Library: Access by subject and alphabetically.

References

Bailey, Charles W., Jr.,Scholarly Electronic Publishing on the Internet, the NREN, and the NII: Charting Possible Futures, Serials Review, 20(3), 1994, pp. 7 - 16
Bailey, Charles W., Jr., Network-Based Electronic Publishing of Scholarly Works: A Selective Bibliography, The Public-Access Computer Systems Review, 6(1), 1995, pp. 5 - 21
URL:http://info.lib.uh.edu/pr/v6/n1/bail6n1.html
Brent, Doug, Stevan Harnad's "Subversive Proposal": Kick-Starting Electronic Scholarship - A Summary and Analysis, The Information Society, 11, 1995, pp. 275 - 283
Gasaway, Laura N., Scholarly Publication and Copyright in Networked Electronic Publishing, Library Trends, 43(4), 1995, pp. 679 - 700
Ginsparg, Paul, Winners and Losers in the Global Research Village, Contribution to a Conference held at UNESCO HQ, Paris 19 - 23 February 1996
URL:http://xxx.lanl.gov/blurb/pg96unesco.html
Graham, Tom, Electronic Journals: Pricing Models, Newsletter on Serials Pricing Issues, 161, 1996.
URL:http://sunsite.unc.edu/reference/prices/1996/PRIC161.HTML
Harrison, Teresa and Stephen, Timothy D., The Electronic Journal as the Heart of an Online Scholarly Community, Library Trends, 43(4), 1995, pp. 592 - 608
Harter, Stephen P. and Kim, Hak Joon, Electronic Journals and Scholarly Communication: A Citation and Reference Study, Paper delivered at ASIS Midyear Meeting, San Diego, CA, May 20 - 22, 1996
URL:http://www-slis.lib.indiana.edu/PrePrints/harter-asis96midyear.html
Hitchcock, Steve, Carr, Leslie and Hall, Wendy, A Survey of STM Online Journals 1990 -95: the Calm before the Storm
URL:http://journals.ecs.soton.ac.uk/survey/survey.html
Kimberley, Robert, Electronic Journal Distribution: a Prototype Study, The Electronic Library, 13(4), 1995, pp. 313 - 316
Lancaster, F.W., The Evolution of Electronic Publishing, Library Trends, 43(4), 1995a, pp. 518 - 527
Lancaster, F.W., Attitudes in Academia Toward Feasibility and Desirability of Networked Scholarly Publishing, Library Trends, 43(4), 1995b, pp. 741 - 752
Lynch, Clifford A., Scholarly Communication in the Networked Environment: Reconsidering Economics and Organizational Missions, Serials Review, 20(3), 1994, pp. 23 - 30
Odlyzko, Andrew M., Tragic Loss or Good Riddance ? The Impending Demise of Traditional Scholarly Journals, International Journal of Human - Computer Studies, 42, 1995, pp. 71 - 122
Odlyzko, Andrew M., On the Road to Electronic Publishing, preliminary version distributed by the author, April 15, 1996. To be published in Euromath Bulletin
Olsen, Jan, Electronic Journals: Implications for Scholars, Learned Publishing, 7(3), 1994, pp. 167 - 176
Peek, Robin, Where is Publishing Going ? A perspective on Change, Journal of the American Society for Information Science, 45(10, 1994, pp/ 730 - 736
Roes, Hans, Electronic Journals: a survey of the literature and the net, Journal of Information Networking, 2(3), 1995, pp. 169 - 186
URL:http://pi1103.kub.nl:2080/hanshome/articles/ej_join.htm
Schaffner, Ann C., The Future of Scientific Journals: Lessons from the Past, Information Technology and Libraries, December 1994, pp. 239 - 247
Schauder, Don, Electronic Publishing of Profesional Articles: Attitudes of Academics and Implications for the Scholarly Communication Industry, Journal of the American Society for Information Science, 45(2), 1994, pp. 73 - 100
Tenopir, Carol, Authors and Readers: The Keys to Succes or Failure for Electronic Publishing, Library Trends, 43(4), 1995, pp. 571 - 591
Taubes, Gary, Science Journals Go Wired, Science Magazine, 271(5250), 1996, p. 764
URL:http://science-mag.aaas.org/science/scripts/display/full/271/5250/764.html
TULIP Final Report, Executive Summary, Elsevier Science, 1996
URL:http://www.elsevier.nl/info/projects/trmenu.htm#ExecSummary
Weibel, Stuart, The World Wide Web and Emerging Internet Resource Discovery Standards for Scholarly Literature, Library Trends, 43(4), 1995, pp. 627 - 644
Woodward, Hazel, The Impact of Electronic Information on Serials Collection Management, IFLA Journal, 20(1), 1994, pp. 35 - 45