As of 22 April 2009 this website is 'frozen' in time — see the current IFLA websites
This old website and all of its content will stay on as archive – http://archive.ifla.org
Usage was monitored at each site and statistics analysed to develop factual usage profiles. We are now conducting follow-up studies (questionnaires and interviews) with users and non-users to develop behavioural models which explain their usage and identify the features they value most. Finally we will develop these models into scenarios for future electronic journal services and explore the implications with users, publishers, libraries, and other stakeholders.
To create an electronic version of a journal, you need the article in electronic form. So we also asked authors about if they currently supply manuscripts on disk, and what the advantages would be. Though authors in the sciences were more accustomed to disk submission, authors in all areas were generally willing and perceived the advantages in the same way: their articles would be more accurate, and the editorial process would be easier and more efficient.
The clear implication for publishers is that they must be able to generate electronic files for each article. At the beginning of the SuperJournal project (1996), only half the publishers were in a position to do so. Now SuperJournal regularly delivers electronic versions of 50 journals based on files supplied by the publishers, and none have problems generating the files.
What's taking place in most publishing houses is a re-engineering of the production process: the tasks performed, processes involved, process flow, and staff roles. In many cases there's a redefinition of what work is done inhouse, what is done by their suppliers. Where the production process used to be manual, with production editors marking up hardcopy manuscripts and passing them on to typesetters, systems are now computer-based. Publishers accept, and indeed ask for, author manuscripts on disk, edit them online, and pass files on to their suppliers. At the end of the process there are files for each article standard formats for multiple use. Systems vary from publisher to publisher, but investment can be in millions for hardware, software, new process development, and staff training.
Electronic production systems give the publisher the potential to develop electronic versions of printed journals or indeed totally different electronic journals. What each publisher decides to do with the files depends on their view of what the market wants and their long term publishing strategy. Some make electronic versions of their journals available on their Web site (perhaps on an experimental basis), some offer their journals through one or more aggregators, and the largest offer their own electronic journal services. All are testing the market to see if users really want electronic versions, if libraries are prepared to provide them, and how to position and price them in relation to print.
What libraries can expect is an increasing number of journals to be available in electronic form from a variety of sources. Typically these will correspond directly to the printed version, with the full text article in PDF format. Some will offer the full text in HTML format as well, but this involves more investment (and more work). In most cases the backfile will start from say 1995 or 1996, as few publishers will generate them retrospectively. On the whole this will give researchers what they have asked for: easy access to the full text article, with some backfile.
Publication times vary considerably per journal. There are two distinct parts of the process: peer review of the submitted manuscript by referees, and the production work done by the publishers and their suppliers. For a typical journal in the Life Sciences, publication time might be 6 months, with 2 months for the peer review process. For a journal in the Humanities or Social Sciences, publication times are typically longer, eg 9-18 months, with 4-6 months for peer review.
Most publishers are focusing on the production part of the processes. By implementing electronic processes, they can reduce the production time by 1-2 months. A Life Sciences journal that used to take 6 months to publish can now be published in 4-5 months. Allowing for the longer peer review in the Humanities and Social Sciences, in principle similar electronic production could reduce the overall publication time to 6-9 months. This does pose the publisher with a problem in the short term: what to do with a backlog of 3-9 months of unpublished manuscripts! Typically publishers are phasing in electronic production gradually, publication times will reduce gradually rather than sharply, and they will shorten first in subject areas where there is the greatest time pressure for publication.
In contrast, there has been little change for the peer review process. Individual publishers are experimenting with individual journals, but much depends on the age of their editors and referees. Peer review is largely intellectual: referees need to read and critically evaluate the article, authors need to revise it, and there may be several iterations. Arguably the time saved will be that of transporting the manuscript, and this will not affect publication times greatly.
There is one other way in which publication times can be reduced: releasing articles electronically once peer review, editing, and author corrections have been done. For example, the American Chemical Society releases articles on their Web site on an "As Soon As Publishable" (ASAP) basis, which they say makes them available 2-11 weeks before the printed issue. However, Nature, Science, and the high-profile medical journals have a press embargo policy: print and electronic versions are issued simultaneously on an agreed press date, so that no one use advance notice of important discoveries to unfair advantage. Practice is likely to continue to vary, but publishers are also making the most of journal-based current awareness services to make sure that readers know about articles as soon as they are published.
We have offered users four clusters of journals in different subject areas, each with an average of 12 journals. Users indicate that the "cluster" approach is right, but they would like a "comprehensive" set of journals in their discipline for searching, and the ability to make their own choice, perhaps very selectively, for browsing. What they really want is a seamless approach to using journals: to do a comprehensive search, view the abstracts and full text of particular articles, and link to related articles listed in the references. There is also interest in alerts based on personal profiles to keep them up to date with what's published, but comparatively little interest in multimedia and other bells and whistles.
For true integrated services to emerge, two things are necessary: aggregation (critical mass of titles) and seamless linking among the various titles. In terms of aggregation, the larger publishers are developing their own services, eg Academic Press's IDEAL, Elsevier's Science Direct, and Springer's LINK. The more journals they publish, the more critical mass they feel they can offer to the market. The smaller publishers are experimenting offering their journals through third-party aggregators of different types.
The first integrated systems involving both aggregation and linking are now emerging. High Wire Press based at Stanford University enables linking among journals on a reciprocal basis. However, they have set a clear agenda to focus on learned society journals and to exclude commercial publishers. This is one way to set up a service on a small scale, and to address technical issues among like-minded publishers, but in the long term users want integrated access to all journals regardless of source. ChemPort from ACS/CAS and IOP's electronic journal service are perhaps better examples of what the future holds: integration of an established bibliographic database with full text articles from multiple publishers, to enable seamless searching, viewing, linking, and forward chaining within a broad-based discipline.
Work on linking is only just starting. There hasn't been any rush, as current articles cite older ones, and you need the cited articles in electronic form to build the link. Now that the backfiles are building, and standards like the Digital Object Identifier (DOI) are in place to uniquely identify documents and facilitate copyright management, linking is possible from a technical point of view. Projects like Red Sage and SuperJournal that stimulate communication among publishers build the trust that will be needed for long-term collaborations on linking.
Libraries can expect that initially there will be a many aggregators offering a variety of electronic journals services. But over time, several key subject-based integrated services will emerge that meets the needs of users in each discipline. Service will vary per discipline in terms of value added features, and these will preserve the character and branding of component journals. It's unlikely that one monolithic and homogeneous "service" will emerge.