Most staff and students in UK higher education now have online access to hundreds of academic journals, thanks to the HEFCs Pilot Site Licence scheme. Many more journals are also available in electronic form, access to which must be negotiated separately. The total number of electronic journals is now so large that the most ostrich-like of librarians can no longer ignore them. A recent posting to lis-elib maintained that "There will be 3000+ e-journals based on existing publications alone (i.e. parallel print/e-journals) by the end of this year (until now we have been talking about 100s)" . I shall discuss here some of the problems encountered in trying to provide access to all this material - I wish I could also provide the solutions! The Universities Science and Technology Librarians Group discussed the topic at a meeting in March 1996 , and I'm grateful to my colleagues for the long list of problems raised then.
In an ideal world, the electronic journal is surely the answer to so many of our traditional problems, with speedy delivery, availability unlimited by time or geography, and searching facilities. And think of all the shelf space saved! In reality, we're in a transitional period, having to cope with all our print journals at the same time as coming to terms with a new medium. I suspect few libraries have yet replaced print journal subscriptions with an electronic service, as they have already done with bibliographic databases.
What do I mean by an "electronic" or "online" journal? "Electronic" - one where the text is read on, and/or printed from, the end user's computer rather than as print on paper. "Online" - the data is downloaded directly from the host computer rather than via an intermediate medium such as CD-ROM. Not being a CD-ROM fan, I shall here discuss only the problems of online journals. I'm also mainly talking about "academic" journals, from reputable publishers such as learned societies and containing scholarly, peer-reviewed articles, rather than the newsletter type of publication. Most of the journals we're currently concerned with are electronic full text versions of an existing print-on-paper journal. There is however a continuum from titles providing Tables of Contents only, to those adding abstracts of articles, selected full text articles, all articles as full text but excluding letters, reviews, etc., and complete full text equivalents of the print version. Some add content to the electronic version, such as responses to articles and hypertext links to related resources.
Electronic journals fall into several groups according to the method of payment and/or licensing. The HEFCs Pilot Site Licences provide subscribing UK HE institutions with free access either to all a participating publisher's online titles, or to those to which the institution subscribes in print. Many other publishers are allowing free access to the electronic version of titles one subscribes to in print, sometimes just for a trial period; some are making an additional charge for the electronic version from the start. Some academic journals are freely available online, for example those funded under the eLib programme .
The medium should not overshadow the fact that most of the electronic journals we deal with at the moment are still traditional journals, and many of the procedures for dealing with them will be little different from those for print on paper. A simplified flowchart for traditional academic library journal processing will be something like this:
Traditional academic library journal processing
Each step has its own problems with the implementation of electronic rather than paper journals, which I shall discuss in turn.
Identification of suitable titles is in some ways easier in the electronic medium, with the availability of services such as the ARL Directory  and e-mail notification services like the NewJour list  . All the usual journal selection criteria will apply - relevance to the needs of one's users, peer review, status of publisher, etc. - and additional ones will apply to the electronic medium. Foremost among these is whether users will be able to read the journal and print articles from it - do they have appropriate hardware, networking and software?
Many publishers are making electronic journals available free for a trial period - analogous to the sample issue - and it's important to get users to try these out and give you feedback. A problem here is that of control over the timing of the trial - academic staff must be free to take part, which may be impossible at certain times of the year.
Selection is in some ways easier if you already subscribe to the print version. If the electronic version is free with the print subscription, as many are, then the feeling tends to be "We might as well take the electronic version too".
Checking-in of issues of a print journal is generally straightforward, whether or not you have an automated serials management system. Checking that electronic issues are available online is likely to be far more time-consuming, especially since they are often available before the corresponding issue appears in print (for a dual print/electronic title). Another problem is - what is an issue? The Chicago Journal of Theoretical Computer Science  publishes articles as and when they are accepted. While it maintains the idea of the volume in that articles are organised on the Web page by year, the "issue" concept no longer exists, and one has no way of knowing when the next article may appear. This makes claiming missing issues impossible!
Most licences for electronic journals will allow access by bona fide members of the institution holding the licence, and it is up to the institution to ensure that only these people are given access. External users needing to consult a print journal are admitted to most academic libraries, with or without various registration procedures, but access to electronic journals will often be impossible for such people. The terms of the licence may prohibit it, or the institution's own arrangements for access may do so, for example where the user has to have an account on its computer network to be able to use any of the machines on campus.
The electronic medium may present problems for access and use of the journals, for both library staff and users. Publishers generally limit access by either the Internet Protocol (IP) address of the user's computer, or by password. The former is now common, and more convenient for most users, who don't have to obtain and remember a password. However, the range of permitted IP addresses for the institution's sites must be established and changes notified to the publisher; there seem to be as many definitions of an "institution" and a "site" as there are electronic publishers - multi-site institutions beware! The problem as far as users are concerned is that access will be restricted to their institution and they will not be able to access a journal from home, thus negating what many see as the major advantage of electronic journals.
Password access gets over this problem, but the library then has to decide on a mechanism to give passwords only to authorised users. The national ATHENS initiative  , whereby users only need one username and password to get access to many different services, is an important step forward here. (At the time of writing only electronic journals published by Blackwell and Blackwell Science are covered, in the JournalsOnline service  ). The IDEAL service  is, as far as I know, the only one to offer users the choice of direct login (from a computer in a subscribing institution) or username/password (which they can use from home) to access the full text of journals.
One of the stumbling blocks to promoting the use of electronic journals is the potential plethora of interfaces and delivery mechanisms with which the user may be required to become familiar. The situation has stabilised considerably in the last year or so, with primary access via the Web and delivery of full text as a PDF (Acrobat) file becoming almost the norm. However, although Web interfaces may have many features in common, their implementation by different electronic journal publishers varies considerably, with a variety of search options and navigational tools. Increasingly, multi-media features are appearing in electronic articles, such as sound or video clips, and users will need to install the appropriate software and hardware to access them.
A recent posting to lis-link  summarised the results of an informal survey on electronic journal cataloguing practice. The small number of replies would seem to indicate that few libraries, in the UK at least, are so far cataloguing electronic journals. The main reason is probably that expertise in cataloguing electronic resources has not yet built up here, and another that few records are as yet available for downloading from sources such as OCLC.
Another problem is that standards are still evolving. For instance, the 856 field (Electronic Location and Access) isn't yet part of UKMARC, so even those UK libraries with Web catalogues can't link from the catalogue record to the electronic journal itself. This will no doubt be resolved in time, and we at UCL are putting the 856 field in our records anyway. For examples of our electronic journals catalogue records, try looking for Appetite or Journal of Biological Chemistry on our Libertas system .
Those of us without a Web interface to our catalogue must provide access for our users by listing electronic journals on our Web pages. This means a duplication of effort if we also catalogue them, but this should be a temporary state of affairs until we all have Web catalogue interfaces.
Ideally, one would like an electronic journal's URL (whether linked from a plain Web page or a MARC record) to be that of the title itself. In practice, many libraries are linking only to the home page of the publisher's electronic journal service, for example, Academic Press's IDEAL service , both because the URLs of each title can be extremely long, and because in the past they have changed, and may do so in the future.
Printing the article the user wants is the equivalent of photocopying the printed page. The problems it presents will often be those of the networked rather than the stand-alone environment, and particularly those of cost. Most academic library users pay a charge per page of photocopying, but this is not yet always the case for computer printing. Where it is not, the institution may incur heavy printing costs, especially when students cotton on to the fact that such printing is free to them! Problems may also arise for users with older printers, and whose printer is not configured to suit the file format of a particular journal. Users may not yet be aware of the length of time it can take to print a journal article in PDF format, which will again pose particular problems with networked printers.
One of the biggest problems for libraries at the moment is the extra staff time needed to administer electronic journal provision, as virtually all the work involved is on top of that which library staff do already. There are also extra costs in training users; more resources are offered to them each year and consequently we have to increase the number and/or length of training sessions. And before staff can train users, they have to train themselves, or be trained, in what electronic journals are available and how each is best used; some people will find this easier than others. Staff costs, particularly for supporting users, will also be incurred by departments other than the library, such as the Computer Centre.
Electronic journals which are currently free may well start charging for access soon, and the periodicals budget will have to absorb these costs if the library decides to continue subscribing.
It is difficult to quantify the extra hardware and support costs, as these are increasing already with the growth in personal computing and Internet use generally. Computers and printers may be tied up for longer downloading and printing articles; helpdesks and enquiry points will become even busier.
If and when electronic journals replace their printed equivalents, then the savings in terms of human and other resources may be considerable. Libraries will no longer have to shelve or bind their journals, or replace lost or damaged issues, and may need fewer photocopiers.
The equivalent of metres of back issues of journals is the electronic archive. It is assumed now that libraries will hold back runs of journals - and that at least one library somewhere will hold a complete run of a title - but the question of archiving electronic journals remains unresolved. At the moment, publishers are maintaining their own archives of electronic journals, which are quite extensive in some cases, but even given the ever-decreasing cost of electronic storage, one wonders whether this situation can continue.
Many people feel that archiving should ultimately be the responsibility of national libraries. One problem is that currently copyright deposit does not apply to the electronic medium, so they incur extra costs in acquiring electronic material. Another, which also applies to other libraries holding large electronic archives, is that of access to the holdings. Who will be allowed access to a title? - licensing arrangements have so far been quite restrictive. Will access to issues which the library has purchased be maintained if the subscription is subsequently cancelled? Will data be refreshed and ported to new hardware, or will legacy hardware need to be maintained to read older data? Such problems probably do not concern most of us in our daily work, and will not arise for a few years, but it is useful to be aware of them, particularly if making a decision as to whether or not to cancel a print subscription in favour of an electronic one.
A related question concerns inter-library loans. If a library subscribes to only an electronic version of a journal, under present copyright legislation and licences it will not be able to make copies for another library.
It has always been notoriously difficult to quantify the use of printed journals, and one feels that the electronic medium should improve the collection of statistics in this area, as it has, for example, with online circulation. The publishers are presumably collecting statistics on the use of their electronic journals, but whether or not they will make them available to institutional subscribers remains to be seen. Measuring access to Web pages is notoriously inaccurate, although some access mechanisms should improve accuracy. The problem of finding out what the user does with a retrieved article (such as how much of it they read and how useful they find it) remains; in these early days there must be a lot of exploration and trying things out, which will inflate the apparent use of a journal.
Many currently free electronic journals will start charging for access from 1998, so we need feedback now to enable us to decide whether or not to pay an extra subscription for electronic access, or even to cancel the print version. As well as quantitative data on number of accesses, we need qualitative information from users on how useful they find online access, whether or not they prefer it, the problems they've encountered, and so on. Projects such as SuperJournal  have funding to enable them to use questionnaires and run focus groups, and their results should prove very useful to all of us. In the meantime, we have to collect and make use of feedback from our users as best we can.
Amy Tucker, Electronic Journals Product Manager, Institute of Physics Publishing <email@example.com> replies on 22-July-1997:
I would like to respond to Judith Edwards' article 'Electronic Journals - Problem or Panacea'
Institute of Physics Publishing (IOPP) is part of the HEFCE agreement and provides access to 28 full text journals online to participating institutions providing they have one or more print subscriptions via the Electronic Journals service (http://www.iop.org).
In her article Judith states that:
The IDEAL service  is, as far as I know, the only one to offer users the choice of direct login (from a computer in a subscribing institution) or username/password (which they can use from home) to access the full text of journals.
Just to let you and your reader know this is not the case. IOPP abolished the need for obligatory Site and user passwords in November 1996 and since then have offered a multi-level service which includes access from the home and office:
Access to the service is verified by IP address. Judith also mentioned the problems associated with IP registration:
However, the range of permitted IP addresses for the institution's sites must be established and changes notified to the publisher; there seem to be as many definitions of an "institution" and a "site" as there are electronic publishers - multi-site institutions beware!
We allow institutions to register their site by domain name, thus solving many of the problems that occur if a number of IP address ranges are covered, or require adjustment.
I hope this information is of some interest to you.
 Steve Hitchcock, (12 June 1997). Re: Secure document delivery. Posting to lis-elib. Archived at:
 Report of USTLG Meeting 14 March 1996. Posting to lis-scitech. Archived at:
 Association of Research Libraries (1995). Directory of Electronic Journals, Newsletters, and Academic Discussion Lists. Mirror site:
 BIDS Journals Online,
 Val Wilkins, (20 June 1997). E-journals: summary of survey. Posting to lis-link. Archived at:
University College London Library,
Tel: 0171 380 7833