Electronic Journal Publishing: A Reader Version 2.0

Published by INASP, 2001

©INASP 2001

http://www.inasp.info

 

Getting Started in Electronic Journal Publishing
Sally Morris

 

4th Edition

 

Why do electronic publishing at all?

 

Publishing a journal electronically sounds very attractive. There are a number of good reasons for doing so, but it does have disadvantages too. Before committing yourself and your organisation to the effort and expense involved, it is sensible to look carefully at both the advantages and disadvantages. In the end, the decision will depend on what your main objectives are, so it is important to be clear about why you are publishing, or planning to publish, in the first place: what information you want to disseminate, and to whom.

 

Electronic publishing is often considered to have four great advantages over print: international reach, speed, additional capabilities and lower cost. All of these are true up to a point.

 

International reach. It is of course true that readers all over the world can in theory get hold of and read your journal (provided that they have the necessary equipment and, for online journals, Internet connections); you do not have to arrange for printed copies to be mailed to different parts of the world, which can be both expensive and slow. However, the real question is whether people will access your publication in the first place. To do so, they will need to know of its existence (and, for an online journal, its website address). Unless you carry out international promotion for your journal, electronic availability will not necessarily make any difference to its readership.

 

Speed of publication. As soon as an article is ready for publication, it can be published; there is no need to wait for it to be printed and bound. And as soon as it is published, it is available to readers; there is no need to wait for it to arrive through the post. What is more, you do not need to wait until you have enough articles to make up a complete printed issue. You can, if you wish, publish articles individually as soon as they are ready (or even, if you wish, in accepted – but not yet finally edited – form). These time savings can be very helpful to authors, particularly in fast-moving subjects such as the sciences, where priority of publication is extremely important.

 

Many of the important processes of publication, however, will not in themselves be speeded up if you want to produce a quality journal. If it is a research journal, you will still need to arrange for peer review to be carried out, and peer review and subsequent revision is often the most time-consuming step between submission of an article and its eventual publication. Although the transmission of papers to referees, and of their reports to the editorial office, can done instantly by electronic means [1], the referee's reading, thinking and writing of reports will take exactly the same amount of time.

 

Similarly, the editorial processes are important if published articles are to be consistently and accurately expressed, and clearly understood. Good editing also makes it easier for readers to find their way around journals and the individual articles. If you want to make an electronic journal even more useful and interesting than a print one, additional work will be required such as adding links, supplementary material, or 'non-printable' features such as moving images or animations. It is not helpful to the scientific community if publishers do not spend the time and effort required for careful editing. This may, in fact, be the 'bottle-neck' which prevents you from publishing more articles; unless you can afford to increase your editorial resources, you will not be able to edit more than a certain number of articles without damaging the quality.

 

Additional capabilities. Of all the features which are possible with electronic journals, linking – the ability to go directly from a citation in one article to the article to which it refers - is the one most valued by authors and readers. [2], [3] See section on Linking below for more details.

 

Electronic publication also makes it possible to include material which you would not include in a print journal, either for space reasons or because the material is difficult, costly or even impossible to print. Research articles could be accompanied by the full data-sets from which the results were reached – ideally in a form which readers can manipulate for themselves, for instance to compare with their own or others' data. Colour illustrations may be prohibitively expensive in a printed journal, but present no such problems in an electronic journal.

 

Moving images, sound, or animations could also be included, although relatively few authors automatically think of providing such material at present. You would also need to consider whether your readers (or, indeed, your referees) would be likely to have the necessary software to make use of the additional material. Non-textual files will also present the editorial office with new challenges in working out how to check and, if necessary, correct such material.

 

However, it is worth noting that respondents to a recent survey2 rated linking very highly, but all other additional features much less so. It is probably therefore not worth spending a great deal of time and money on these.

 

Reduced costs. If you decide to go without a print edition at all, you will save all the costs associated with printing and delivering a paper copy. However, for the relatively small print-run of a typical academic journal, print and delivery costs represent a small proportion of the overall costs, while all the editorial and 'first-copy' costs stay the same. You will also have new costs associated with electronic publication – these are discussed in more detail below. In particular, the costs of data handling and administration are usually increased. Tenopir and King [4], [5] estimate, from statistics they have gathered, that the overall saving for a journal with 500 subscribers is no more than 4%, though it could be as high as 25% if you are lucky enough to have 5000 subscribers. However, they acknowledge that these figures may not take full account of the new computer-related costs, which increase for larger circulation journals. Halliday and Oppenheim [6], [7] have estimated in considerable detail the underlying costs which are not reduced by electronic, as opposed to print, publication; they also try to estimate the new, additional costs of electronic publication. [8], [9]

 

If you decide to offer both print and electronic versions, your print and delivery savings will of course be less, but your new, electronic, costs will remain the same. Overall, Tenopir and King4 estimate that costs may increase by approximately 20%. This is therefore an expensive option, although many publishers have found that their readers are unwilling to go without print copies entirely, and they have therefore been forced to take this route at least in the short term.

 

What really seems to happen, when readers use an electronic journal, is that they print out the articles they want to read, so that they can carry them around, annotate them, and perhaps file them2. Reading a complete article from a screen is not something with which the majority of today's readers are comfortable – although it may be that the younger generation, more accustomed to reading from a computer screen, may behave differently in future. Thus in a way you are simply moving the cost of printing, from the publisher to the reader.

 

The costs of administering an electronic version will be considerably lower if you are able to make your journal available to everyone, free of charge. This means that you will not need to have a system in place which checks the identity of everyone trying to access your site, and responds in different ways depending on whether or not the system recognises them as a subscriber. Such a system is complicated and expensive, not only to create but also to maintain; see the sections below on the Electronic journal 'system', on Subscriptions administration and customer service, and on What are the business models for recouping some or all of your costs?

 

Convenience. One of the great advantages for the user is that they can access the journal's content at their computer – even, if they are lucky enough, at their own desktop. They do not need to go to the library, and risk finding that someone else is using the journal issue they wanted. What is more, they can access the journal at any time, and not only during library opening hours. Convenience was the most highly rated feature of electronic journals in a recent ALPSP survey.2

 

Searchability. One of the important benefits of an electronic journal is that it makes it much easier for readers to search for articles. This is particularly true if you organise your system in such a way that users can search for particular words or phrases. Even if this is not possible within your system, it is very quick and easy for a reader to browse the tables of contents of previous issues, for example to find an article which they can only remember inexactly. This is far simpler than looking through every issue in turn on the library shelf! Depending on the sophistication of your system, it may also be possible for readers to jump directly to a particular section of the journal (such as book reviews), or even to a particular section of an article (such as 'Materials and Methods' or 'Results').

 

Linking. One of the features which is most attractive to both authors and readers2 is the ability to add live hyperlinks within the journal material, either within your own publication(s) – such as links from the table of contents to the individual articles – or to other publications – such as links to articles cited in the reference list. Linking is becoming possible via secondary abstracting and indexing databases such as MedLine; at a minimum, you would be able to link to the abstract of the cited article, and depending on the other publisher's policy, your readers may then be able to follow that link to the full text if they wish. Developments such as the Digital Object Identifier (DOI) [10], [11], [12], [13] are specifically designed to make this easy. A collaborative project called CrossRef [14] uses the DOI technology to operate a system of managed and consistent linking between citations and the cited articles in electronic journals; participation is inexpensive for a publisher with only a few journals.

 

Archiving. This is another issue which requires careful thought; it is worrying publishers and libraries the world over. Print journals are relatively easy to archive, although they take up space and may deteriorate over a long period of time, particularly if it is not possible to store them in ideal conditions. Furthermore, a library which has subscribed to a journal in the past will always have the copies to which it subscribed, even if one day it has to cancel its subscription.

 

The same is not true for electronic publications. [15], [16] There is no easy way of archiving an electronic-only journal, particularly one which is delivered online. Even if the archiving library makes backup files of the journals, these may need to be updated over time, as technology changes. Links to other sites – one of the most attractive features of online journals – may become inactive due to the impermanence of World Wide Web addresses (URLs).

 

In some countries, national libraries are looking at how they can maintain such an archive, but even they have not solved all the problems. Any form of deposit of electronic journals (for example, with national libraries) will be of no use unless the archive includes the necessary software to retrieve the content. If you decide to publish an electronic-only journal, do consider carefully how you will approach this problem; otherwise, there is a real danger that the material you have published will not be available to future scientists. A paper version does, of course, provide one solution. One interesting project focusing on the archiving of journal articles is J-Stor. [17], [18], [19], [20], [21], [22]

 

Much work is going on in various countries to explore the requirements for legal deposit of electronic materials (e.g. with national libraries) in the same way as is currently required for print. In addition to working out how to preserve these materials adequately in perpetuity, questions of access also have to be addressed in order to protect the businesses of the depositing publishers. [23], [24]

 

It is important to consider, in developing your licence (see section on Licensing considerations below), what approach you will take to providing access to material if the customer should cease to subscribe. Librarians are anxious to ensure that, as with print, they continue to have access rights to the material to which they subscribed. However, not all publishers may wish (or be able, within their system) to continue to make access available to previous material even when the library has stopped subscribing. One alternative is to provide the customer with electronic files (possibly on CD-ROM) when they cease to subscribe.

 

A more radical alternative, though, is to make older back issues available to everyone free of charge. The Public Library of Science initiative [25] urged authors and editors only to work with journals which made their back issues freely available, in a particular type of archive, 6 months after publication. Partly as an indirect result of this initiative, a growing number of publishers [26] are now making their back issues freely available, though usually on their own site; the interval between initial publication and free access varies between 6 months and 3 years according to the pattern of use of older material in different disciplines (and, of course, according to each publisher's requirement for cost recovery).

 

Alternatives to formal journals. In certain disciplines, researchers are accustomed to distributing their work to each other informally before publication, in the form of 'preprints'. This is of course much easier to achieve electronically (e.g. via e-mail) than it was in print form. Electronic media also make it very much more feasible to store these preprints in a searchable database (often institution-based). A preprints database is very well established in physics and related disciplines. [27] In addition, some writers are now urging authors to archive the final, published version of their papers in publicly accessible databases (sometimes known as eprint databases); this is a more acceptable approach in subjects, such as medicine and the life sciences, where unrefereed material could be problematic. [28], [29] Software is being developed to make it easier to search for papers in more than one preprints/eprints database at the same time. [30]

 

The problem with this, of course, is that it could undermine the very existence of the journals in which the papers were originally published, since the individual articles are available free of charge to anyone who wants them. Of course, in preprint form the articles have not had the benefit of peer-review, nor of the improvements which can be introduced by good editing. Even more important, perhaps, is that they do not have the valuable 'branding' which comes from inclusion in a recognised journal; this enables the potential reader instantly to have some idea, not just of the quality, but also of the subject matter and the editorial 'flavour'. 'Eprint' archives, however, do not have these disadvantages, and could indeed threaten the survival of the journals on which they draw unless an alternative way of recovering journal costs is found.

 

Some writers, [31], [32] believe that peer review and associated processes should be paid for by authors (or their funding bodies) so that the peer-reviewed and edited versions of the articles can be made available free, although others4 recognise that it might be difficult for this to produce enough income for the publisher. This is more difficult for researchers in the humanities, where research funding is low, than in the sciences. It is also more difficult for authors from less prosperous countries (indeed, many of the journals which do recover their costs in this way waive the charges for such authors). A recently launched alternative [33] which looks promising is to ask institutions to pay, not for journal subscriptions, but to make the output of all their researchers freely available in 'open access' journals. Some publishers35 are also experimenting with making the articles themselves free, but charging for related services such as reviews.

 

Obviously, it is essential to have a viable business model for cost recovery (or subsidy) before committing to an open access journal. However, the benefits of open access to scholarship are so clear that the Open Society Institute has recently launched a 3-year initiative which will financially support the development of new business models and the creation (or conversion) of open access journals. [34]

 

Uncertainties. Electronic (particularly electronic-only) publication gives rise to a whole range of new questions, such as: How do you cite a reference in an electronic-only journal? Or in a preprints database? What do you give as the publication date for an article which is published online before the print issue is available? There are more fundamental issues, too, about what should count as 'publication' in an environment of potentially continually updated documents; [35] these are all difficult and, so far, unanswered questions.

 

If you decide to do it, what decisions do you have to make?

 

If you reach the conclusion that electronic publication is the right course of action for your journal, there are a number of choices which you will need to think about.

 

Journal website, tables of contents, or full text? In order for your journal or journals to have an online presence, you will need to create an appropriate website (in fact, if the publisher has its own website, this is a valuable marketing tool even if an online version of the contents is not available). This should contain information about the journal – its coverage, the editor(s) and editorial board members, how to subscribe and how to contribute an article; a picture of the journal cover might be a useful addition. If online content is available the journal website will also be the point of access for this.

 

You do not have to make the entire contents of the journal available electronically. There can be considerable value in just making the tables of contents available, possibly also including the article abstracts; this can be done before the printed issue is even available. This will give readers the advantage of being able to search quickly and easily for articles of interest. If you decide, initially at least, only to provide the tables of contents electronically, you will not need to think about charging for access and restricting it to subscribers. It can be good publicity for your journal, and your authors, to publicise the availability of the contents listing of each new issue as widely and as quickly as possible.

 

You could also consider offering a service to readers whereby you email them with details of the table of contents of the next issue; this is equally possible whether or not the full text is available online. If you offer this service to non-subscribers as well as subscribers, you will be able to identify people who might be interested in subscribing in future.

 

All articles or just some. You do not need to make the complete journal contents available electronically. You might decide only to offer the research articles, while including other material (editorials, book reviews, news, calendars of events, correspondence) only in the print version. Or, if you decide to view the electronic version primarily as publicity for the printed journal, you might choose to provide selected articles, or a sample issue, to illustrate the quality and variety of the journal's coverage.

 

Article by article, or issue by issue. If you do decide to publish the entire journal online, you have the possibility of publishing individual articles as soon as they are ready, rather than waiting until you have a complete issue. Indeed, some publishers are now releasing articles in their pre-publication form. [36] Publication of an article as quickly as possible can be very helpful to the author, particularly in fast-moving science disciplines where priority is important. However, it is not necessarily helpful to readers; they may find it more convenient to receive (whether electronically or on paper) a collection of articles packaged together in a journal issue [37]. Article-by-article publication also raises some difficult issues: what is the official 'publication date' if the electronic version is available before the print version? How do you distinguish pre-publication versions, if they are not quite the same as the published version? And how do you do your internal financial calculations – when is the pre-paid subscription revenue 'earned' by publication?

 

Electronic journal 'system', as opposed to just mounting files. It is perfectly possible just to create a website which simply contains electronic files of the journal's content. However, this will not be very attractive or usable for readers. At a minimum, you will really need to provide some kind of hierarchical structure so that readers can find their way around. At the very least, this structure will need to identify the elements Journal, Issue, Article, to enable logical navigation. Depending on the nature of your journal, you may want to distinguish different types of article too. Better still would be to continue the structure within articles (Abstract, Introduction, Materials and methods etc. – not forgetting References) so that readers can move directly to their chosen section. As mentioned above, a search facility is also valuable. Once you have a large number of articles available, the most sensible way to hold them is in some kind of database to which your access system can 'point'.

 

If you decide to sell subscriptions to your electronic journal (or to restrict free access, for example to members of a society or subscribers to the print version) you will also need to have a system which enables users to identify themselves so that the system can permit or prevent access as appropriate.

 

A more sophisticated electronic journal could offer other things, such as the ability for readers to bookmark or even annotate their favourite articles. E-mail alerting when a new article or issue appears, possibly even selecting only those articles of interest to the individual reader, is also possible (although it appears that too frequent messages can be irritating to customers!). Readers could also post comments relating to individual articles, which would then be available to everyone. [38], [39], [40]

 

Print and electronic, or electronic only. As mentioned above, some costs can be saved by not having a printed version at all. However, it is sensible to make sure that your existing or potential customers would be happy with this; do some market research if possible. Bear in mind, too, the archiving problem outlined above.

 

It makes a difference whether you are looking at an existing print journal, or launching a completely new one. With an existing journal, you can ask your customers whether they would welcome an electronic version, either instead or as well; library subscribers may be concerned about the loss of a print copy for their shelves, unless you can offer very substantial cost savings. Authors may be less willing to submit their papers to a journal which is only available in electronic form (this is partly because of the lack of certainty about its long-term future accessibility and preservation); it is always difficult to obtain good papers for a new and unknown journal, and this may make the problem worse. On the other hand, with a new electronic-only journal you can include features which are not possible in print; you can also experiment with new cost-recovery models, such as charging authors or their institutions rather than readers (see below).

 

If you do decide to publish both print and electronic versions, they may well not be exactly the same. Quite apart from additional features in the electronic version, it is impossible to guarantee that minor errors will not creep in to the electronic files. It may be helpful to readers if you state clearly that one or other version is definitive – a growing number of publishers are now making the electronic version the definitive one.

 

Possible extra features in electronic version. Do your customers (both authors and readers) want them? (Again, market research is the only way to find out.) Can you afford them? Some will add to the cost of producing the electronic journal, not only in technology but also in your time. Make sure that you are not wasting time and money creating features which your customers do not value highly.

 

Singly or 'bundled'. If you are fortunate enough to publish a number of journals, there may be benefits in selling them as a single package (with or without the option for customers to buy them separately if they choose). The evidence shows that usage of titles which customers did not previously subscribe to can be unexpectedly high. [41], [42] If you also publish books or other content related to the journal's subject, an even richer package could be created; you could also include links to material on other sites. Another possibility would be to collaborate with one or more other publishers to create a larger package.

 

Online or CD-ROM. One tends to assume that electronic journals are online journals. However, in some circumstances a CD-ROM may be more practical. This is particularly true where there is not general availability of suitable computers, or – most importantly – of Internet connections, which are of course dependent on a widely available and reliable telephone system.

 

Of course, CD-ROM journals do have greater limitations. You will still have to mail the CD-ROMs to your subscribers; it will not be possible to publish article by article, and even a complete normal journal issue will by no means fill a CD-ROM. However, they may be an attractive option in some circumstances, although care should be taken with format. Although CD-ROM format was standardised, [43] today there is ever greater variety in CD-ROM formats (different speeds, data on both sides, etc), and it is important to make sure that your CD-ROMs are in a suitable format for the majority of your customers. CD-ROMs are also an alternative way of providing a relatively stable medium-term archive, although no one can be absolutely certain how long they will remain readable.

 

Do it yourself, or use someone else's service. One way to avoid the costs of developing an electronic journal service from scratch is to use one that someone else has already developed. Learned societies (particularly the larger ones) may offer their services for a charge; so may university presses and other publishers, both non-profit and commercial, as well as some subscription agents

 

(although it is important to make sure that the journal is also accessible to customers of other agents). Other organisations are specialising in offering journal hosting services, sometimes specifically to smaller not-for-profit societies. [44], [45], [46] However, your journal(s) available on the same site as a number of other journals will very likely increase their visibility to potential readers, and thus their use. With electronic journals, it does not really matter where such a partner is based, although there may be language or other reasons why you prefer to use someone in your own country, and day-to-day communications will certainly be easier. Universities and other organisations might also be willing to include your material on their website; bear in mind, however, that if they are not already managing journals or similar content, their systems are unlikely to be suitable without further (and costly) development.

 

There are a number of organisations which have been set up specifically to assist with the electronic publication of journals from developing countries, recognising that the journals may have low international visibility and that their publishers may be hampered by a lack of money and know-how. One such organisation is Bioline Publications [47], [48], [49], [50], 13 which currently makes available 24 bioscience journals on behalf of collaborating publishers, and has also launched 3 online-only journals of its own [UPDATE]. Bioline has also established the Electronic Publishing Trust for Development [51] to support the electronic publication of journals published in developing countries. Another is ExtraMED [52], [53], which produces 10 CD-ROMs a year containing the full page images of nearly 300 medical and health science journals from developing countries and promotes them to libraries worldwide; similar services exist in science (ExtraSCI – UNESCO) and Agriculture (AgROM Extra - FAO). INASP itself produces African Journals On-Line, which mounts the tables of contents of numerous African-produced journals and sells photocopies of articles. [54]The Open Society Institute runs the Center for Publishing Development in Budapest [55] which helps to develop local publishing initiatives in Eastern Europe. Many other such initatives are listed on the Liblicense [56] and ALPSP [57] websites. It is always difficult for learned societies, even in the West, to find sufficient funding to support the development of their fellow publishers elsewhere as much as they would wish; the idea of 'twinning' publishers from more and less developed countries is being explored, for example by the Association of Learned and Professional Society Publishers though its 'ALPSP Partnership Project'.

 

The International Council of Scientific Unions has an excellent short publication, Guidelines for Scientific Publishing [58], which was updated in 1999. Another useful publication is Serial Publications: guidelines for good practice in publishing journals, the second edition of which (published in 2002) takes account of electronic journals. [59] There is also a helpful chapter on Electronic Publishing in Journal Publishing by Page et al. [60] In addition, in 2002 SPARC (an initiative whose mission is to reduce the costs of journals to libraries) launched Gaining Independence, a guide for those wishing to set up their own, reasonably priced, electronic journals. [61] A brief checklist from ALPSP, Putting your journals online, is available at http://www.inasp.info/psi/ejp/checklist.shtml

 

Sell it yourself, and/or go through intermediaries. Making the journal available electronically is only half the story. You then need to sell it, which includes not only making the sale, but also arranging electronic access and managing whatever access control mechanisms you have selected. In addition, you would be wise to consider the licences which you want your customers to have; unlike print journals, electronic journals provide users with a wide range of possible uses, some of which you may not want to allow (see the section on Licensing Considerations below). One possibility is to have one or more intermediaries handle this side for you; many subscription agents are now offering this service, for a fee, for electronic journals. This can make very good sense, as most libraries already deal with subscription agents for their printed journals. Furthermore, through some subscription agents' systems, users can obtain access to all their electronic journals through a single access mechanism; this is obviously preferable, for the user, to the alternative of having to access a different website, probably using a different password, for each journal. It is obviously easiest if the agent is also acting as host for your journal; if they are not hosting the journal, you will need to provide them with basic 'header' information – tables of contents and, preferably, abstracts – in electronic form so that their users can search these. Taking this route will increase your overall costs, since you will need to pay the intermediary, but it is very likely also to increase the accessibility, and thus use, of your journals.

 

What changes do you have to make to your procedures?

 

Whether you decide to do it all yourself or not, electronic publication will make it necessary to change the way you do things. [62], [63] It is important not to overlook this aspect.

 

Editorial. The peer review, revision, and editing of articles, as already pointed out, does not become any less important if they are to be published electronically. In addition, editors may need to add (and verify) electronic links within the article, within the journal, to other publications and to additional material. An increasing number of authors – though by no means all - are able to submit their articles in electronic form, which can save some of the costs of typesetting (or its electronic equivalent). You will only be able to reap the benefits if you have a suitable computer in-house, so that the copy-editing and preparation can be done electronically; if you have to pay your typesetter to transfer your copy-editing and preparation markings to the electronic version, you will see little or no saving. Article-by-article publication makes it particularly important to have a good internal tracking system so that you know what stage each article has reached; you may wish to add articles to your database before they are finally released for publication. There are a number of (fairly expensive) commercial systems which enable you to manage the articles themselves, as well as information about them, at all stages in the publication process. If you do not have one or more computers with sufficient capability, you will need either to buy them or to subcontract the work to someone who has – publishing an electronic journal is more difficult, but not impossible, without any electronic equipment of your own at all.

 

It is essential that, when acquiring copyright or the right to publish from your journal contributors – which you should always do – you make sure that you have secured the electronic rights. It is also important to double-check that you or the author have obtained electronic rights for any other material you wish to include, such as illustrations or quotations from other publications. One sample of an appropriate agreement has been produced by ALPSP. [64]

 

Production. The production processes will obviously need to be different for an electronic (or, indeed, a parallel print and electronic) journal. Files will need to be put into an appropriate format such as PDF or HTML, so you, your staff or suppliers will need to learn how to do this. PDF files reproduce the appearance of the printed page; this can be less readable on screen, depending on the journal's layout, but it is good for printing out. Your or your supplier will need specialist software [65] to create PDF files, and your readers will have to acquire special (though free) software to read them. HTML files produce a screen-based layout, which can be modified (as, unfortunately, can the text itself) by the user. It can be easier to read on screen, but produces less attractive printouts which do not mirror the print edition. An HTML file is also smaller than the corresponding PDF, and therefore faster to download; this may be important if Internet connections are slow.

 

The electronic files must be carefully checked for accuracy, as it is surprisingly easy for errors to creep in when they are converted. Each file will need appropriate 'labelling' – at the very least a file name or other identifier which enables you, and your system, to know which article it is, and preferably a minimum set of 'metadata' such as journal title, volume, issue, article title, article author(s), date of publication, and identifier. If your processes have been less than systematic in the past, this is the moment when you will really have to change that!

 

If you intend to publish parallel editions (e.g. print and online, print and CD-ROM) it will make sense to prepare the content in a single 'medium-neutral' format which can then be converted into the different formats required. SGML, and its close relative XML, is a particularly valuable format to use for this purpose, but it is quite difficult to learn and to use (although some commercial tools are available to simplify this). Even a simple word-processing format, such as Microsoft Word, will take you a long way.

 

If you have chosen to publish article by article, your suppliers will need to have a suitable system for tracking each article individually; this may increase costs.

 

Marketing and promotion. Marketing, in the true sense of discovering what your market wants and then providing it, is every bit as necessary for electronic journals as it is for print journals. Although it may seem tempting to miss out the step of conducting market research, this is extremely unwise. In addition to talking to as many potential authors and readers and possible, try to carry out systematic surveys by questionnaire. If you see electronic publication as being a way to make your journal more international, your market research should be international too. This will cost money, but not as much as you could lose if your journal failed. This process should not stop once your journal has been launched; there are continual opportunities for change and improvement, in line with your authors' and readers' changing needs. Questionnaires to both actual and potential authors and subscribers are worth repeating every few years (if you can do this by email, so much the better); customers who have cancelled their subscriptions can be particularly valuable sources of information about what you are doing wrong!

 

An electronic journal will not sell itself, any more than a print journal. You will need to make sure that those who should be interested in reading it, and writing for it, are aware of its existence. This is not simply a matter of mailing a leaflet (indeed, many publishers report that this is relatively ineffective). If you can obtain suitable lists of email addresses, this is much cheaper than a postal mailing; be careful, however, about sending 'promotional' messages to email discussion lists, as this is considered unprofessional, and may generate a strong negative reaction, although straightforward information about journals is welcomed on many lists.

 

What is much more important is to make sure your journal is visible, both nationally and internationally. Make sure that your journal is covered by the major Internet search engines [66]. However, one of the most fundamental ways of ensuring visibility to your community is to make sure your journal is included in secondary (abstracting and indexing) databases. The major services used to be resistant to including electronic-only journals, but this is changing. The most important of all is ISI [67]), although their acceptance procedures are quite stringent and the number of journals they include is restricted [68]. However, you cannot include your journal in too many such databases; identify as many as you can which are relevant to the subject matter. You may be asked to provide a free subscription, but this is a very small price to pay. If you can supply 'headers' – bibliographic details and, ideally, abstracts - in electronic form, so much the better. Some services will even pay for this, if they find that it saves them time and money; unfortunately, however, there is as yet no standardisation about the format in which they require them. You should also make sure that your journal is listed in Ulrich's International Periodicals Directory [69], which is the major reference source.

 

It is particularly difficult for journals from the developing world to achieve the visibility which results from inclusion in the major databases. There may be concerns about the quality of refereeing, the absence of high-profile writers, and the regularity of publication – all of which problems are in turn exacerbated by low visibility. Services such as ExtraMED, ExtraSCI and AgROM Extra (see below) are doing their best to counter this problem. A conference in Nanjing in November 1999 addressed this very issue. [70]

 

If your (or your partner's) system has such a capability, you may want to offer free access for a short time, or to a certain amount of material, to attract new subscribers. Indeed, many publishers have made their electronic journals completely free to everyone for the first year or more. This can be a good way both of attracting interest, and of getting feedback, although you should be aware of the possibility that customers will resent being charged for something which has hitherto been free; if you intend to introduce charges at a later date, it is helpful to inform customers of this at the beginning.

 

Once a journal is made available online, the process of gathering feedback from readers and authors can be much easier than for a printed title, as the Internet allows for much easier two-way communications. Online feedback forms, direct links to e-mail addresses, and discussion forums all permit much more effective exchange of information.

 

Sales. Printed journals need little or nothing in the way of salesmen. Libraries decide which journals they want, and then choose their preferred channel for buying them. However, site licensing of electronic journals (see section on Site Licences below) does require detailed discussion and negotiation with each customer; this is time-consuming, and if you decide to do it yourself you are likely to need extra staff with the appropriate skills. The alternative is to have someone else, such as a subscription agent, do it for you – but you need to be certain that you are willing for someone else to negotiate prices on your behalf. Bear in mind that the price for a single library may represent subscriptions for several journals; if libraries join together to form consortia, this is even more likely to be the case.

 

Selling to library consortia (and other large customers) is a particularly difficult and time-consuming process, and – as mentioned above – the publisher with only a small number of journals can easily get squeezed out. Initiatives are, however, being developed to help smaller publishers work together to sell a single, larger collection of journals in these circumstances. It may be beneficial to try to work with other publishers in your own country or region to overcome these problems.

 

Subscriptions administration/customer service. If you decide to restrict access in any way (for example to paid or signed-up subscribers, subscribers to the print version, or members of a society or association) you will need to administer this process, providing passwords or other access controls to new users, disabling those which have come to an end, and ensuring that your system can identify authorised users when they try to access the system. All of this will entail even more complicated administration than is required for a printed journal, and will require considerable technical expertise.

 

In some countries, the tax situation will be made more complicated if you are selling electronic journals; they may attract tax where printed journals do not. You will need expert local advice on this.

 

Customers do not need any help to use print journals; however, the situation is different with electronic journals. Some customers will forget their passwords and you will need to reissue them. Others will be unable to get access even though apparently using the correct password. Even if you make your journal available free of charge, you will find that some of your customers have technical problems in accessing or using it. Thus you will need to provide some kind of helpline for customer support, whether on the telephone, on email or both; customers will become very frustrated if help is not available at the exact moment when they need it, so the helpline should ideally be manned full-time during working hours, even if not for 24 hours a day.

 

Of course, if you decide to use an intermediary to provide access to your electronic journal, all of these problems will be taken care of, although naturally at a cost.

 

What might it cost?

 

It is impossible to put precise figures on the cost of publishing a journal electronically; however, it is vitally important to work out your own probable costs in full detail, and not to underestimate either the time or the money that will be required.4, [71], [72], [73], [74] , [75], [76]

 

Time. Someone will need to be responsible for thinking through all the questions raised here, and making the necessary decisions. This is not a one-time process; your electronic journal will need to be continuously monitored and managed to an even greater extent than a print journal. There will be additional complications to your editorial and production processes, and new skills and procedures to be learned. The initial creation of a suitable system will be extremely time-consuming, and its maintenance and development will require continual commitment. If you do it yourself, your subscriptions administration and customer support will also be more labour-intensive.

 

Money. While, as we have seen, taking the 'electronic only' route will save costs on paper, printing, warehousing and despatch, these savings can very easily be offset by new costs. While some have argued that it costs only a fraction of the cost of establishing and running a printed journal, the experience of many traditional publishers does not bear this out. The biggest single cost will be the development of an appropriate journal system, with or without access controls depending on whether or not you wish to charge for subscriptions. Like all computer system developments, this is notoriously difficult to budget accurately, and it would be unsafe to give any guidelines – many people recommend adding 50% to the number (and the time) you come up with! Deciding to partner with an existing provider will make it possible to obtain definite prices – these are likely to be flat fees, rather than being based entirely on a percentage of turnover, since most publishers experience low turnover in the early years of electronic journals. Bear in mind, too, that this is not a once-for-all investment; experience shows that such systems need continual further development, which must be budgeted for.

 

You may need to allow for additional staff costs to cope with those processes which become more complicated and more technical; people with specialist computer skills are expensive everywhere!

 

It is impossible to give definite guideline figures, as there are so many variables – the number of articles received and published, the number of subscribers, the number of other journals handled by the same system, not to mention local variations in equipment, service and labour costs. However, as mentioned above, Tenopir and King4 give some useful averages for a typical scholarly journal. They find that overall savings are of the order of 4% for a 500-circulation journal, and that costs are increased about 20% by parallel print and electronic publishing.

 

What are the business models for recouping some or all of your costs?

 

Your organisation may not wish to recover any of the costs, but rather may see the electronic publication of articles as a service to the community. However, many others will want at least to recover their costs, and possibly to provide additional income to fund other activities. There are various different business models currently in use. Several of these will only be applicable if you offer both electronic and print versions of the journal. In all cases, your pricing decisions will need to be made with a view to what usage pattern you want to achieve.

 

Electronic version free to everyone. This will not, of course, enable you to recover any of your costs from users. However, if it has the effect of increasing readership (which will only happen if your marketing and promotion is effective), then it might make the journal more attractive to advertisers or other sponsors. However, publishers have found it impossible to fund specialist scientific journals by advertising alone. As mentioned above, many publishers make their electronic journals free of charge for an initial period for promotional purposes. The absence of any need for access controls does reduce the costs of the journal system.

 

Author (or institution) charges. The idea of making journals free to readers, and requiring authors to cover the costs instead (for example, out of their research funding), has received much publicity [77] and is beginning to be tested. [78] A new variant on this model is to charge institutions to make possible free access to the output of all their researchers30. As mentioned in the section on 'Alternatives to formal journals' above, the Open Society Institute is actively engaged in supporting the development of financial models which make it possible for journals, without reducing the value they add, to make access freely available to all.32 Although free access is obviously attractive to libraries and readers, it could be risky for a new journal if the charge makes the journal less attractive for authors. You would also need to calculate quite carefully (a) how much you needed to charge to cover costs and (b) how much you could reasonably ask authors to pay. It will obviously make a big difference if they are able to use part of their research grants to pay the charges; if not, and the money has to come out of their own pockets, then this approach is really a non-starter. One possibility is to use author charges to reduce, rather than totally eliminate, subcriptions.3

 

Electronic version free to print subscribers. You may want to do this either permanently, or for an initial period as a form of promotion. It will require an access control system, and you will need to assign a password or other form of identification to every subscriber. Be careful not to increase the print subscription price too steeply, or subscribers will feel – with some justification – that they are being asked to cover the cost of the electronic version, whether they want it or not; this could lose you print subscribers.

 

Electronic version sold to everyone (price: same/more/less than print). You may decide to set a separate subscription price for the electronic journal. If there is a parallel print version, it is sensible to consider the relationship of this price to the print subscription. Should it be the same (since subscribers are buying the same content)? Or more (since they are getting additional benefits and features)? Or less (since your costs are, or may be, less)? You might want to offer a discount for combined print-plus-electronic subscriptions.

 

Electronic version sold to print subscribers for an extra charge. A number of publishers are making an additional charge to include the electronic subscription; the supplements range enormously widely, from 10% to 50% or more. An alternative is to consider the print subscription to be the 'add-on', charging a substantially reduced rate to those who subscribe to the electronic version. [79]

 

Individual subscriptions. The majority of print journal subscription income usually comes from libraries and institutions, but many publishers also sell at a lower price to individuals. You might want to consider offering individual subscriptions to your electronic journals, on one or more of the models outlined above. It is indeed arguable that the real benefits of electronic journals (desktop access, additional features and services) are more attractive to the individual reader than they are to the library. However, you should also bear in mind that if you offer library subscriptions, particularly on the basis of site licences, individuals may already have good access to the electronic journal in any case through their place of work. On the other hand, low-priced individual subscriptions may be the most practical way of permitting users to access the journal when they are not at their place of work. If you offer more than one kind of electronic subscriptions, and if the privileges or permitted uses differ, remember that your system will need to be able to distinguish the different kinds of subscriber at the point when they access the system.

 

Member subscriptions. For those societies and associations which get much of their income from membership subscriptions, electronic journals can present something of a problem. If members see access to their own personal copy of the journal as one of the main benefits of membership, and if they now have access at their own desk through a library subscription at their place of work, there is a risk that they will be less inclined to continue their society membership. Society and association publishers will need to think carefully about this, perhaps offering additional electronic benefits or services only to members.

 

Single article sales. It is often argued that, since a given reader will only be interested in some articles, it would be better if he or she could buy those articles separately. While the costs of administering such a system might be prohibitive in the print environment, it is in principle much easier electronically, although the publisher does need a suitable journal system which can deliver individual articles and collect payment, whether by invoice or by credit card. One of the problems of this system is that the money to buy journal articles is usually in the hands of the university or other institution, and not the individual reader. Publishers have feared that single article sales could reduce the demand for subscriptions, but in fact studies so far do not bear this out – if anything, this seems to tap an additional source of revenue. [80] However, quite apart from any possibility that the publisher might receive less money overall, that money would flow in the form of many small transactions (which are more costly to handle), and over the course of the year rather than in advance in the form of a subscription.

 

Site licences. Universities, companies and other institutional journal subscribers are always very interested in obtaining site-wide access to electronic materials. Access which is only possible from one or more specified computer terminals in the library is very much less attractive. It is therefore a good idea to work out a licensing policy, and an access system, which permits this. Defining the 'site' may be quite difficult, particularly for a multi-site organisation such as a large (possibly international) company. Pricing is a particular problem, since organisations vary greatly in size and in the potential or actual number of users. If you simply have a price based on the print subscription price, then large and small organisations will pay exactly the same, while at the same time none of them will need multiple subscriptions – so you may end up losing revenue. Some method of pricing based on numbers of users might be more equitable, although it will be more complicated to work out the appropriate price for any given customer. Different publishers have looked at the funding of the institution, the total number of faculty and students, or just the numbers in the most relevant department(s). In most cases, publishers have found it simplest to establish a relatively small number of 'price bands' based on these measures.

 

Consortium licences. It is becoming increasingly common for institutions, particularly universities, to band together in order to make better use of their limited budgets, and to share the availability of the materials to which they subscribe. This can be a very effective way of getting your journal content to a much larger community, while still only necessitating a single negotiation. Combined institutions however make very powerful negotiators, and you need to be careful to make the right pricing decisions and not to allow yourself unintentionally to do business at a loss. It is also difficult for a publisher with only one or a small number of journals to sell to large customers and consortia – and equally difficult for the customer. The process is no less time-consuming for one journal than for a hundred, and it requires specialist skills and knowledge. Some organisations are trying to act on behalf of many small publishers to overcome this problem. [81] As mentioned above, the Association of Learned and Professional Society Publishers is also investigating the creation of a single 'package' of journals from many of its smaller members.

 

Licensing considerations

 

When you sell a printed journal, it is fairly clear and well understood what your customers are and are not permitted to do with the material. However, this is much less obvious in the electronic environment and it is sensible to draw up a statement of what customers may and may not do, which they will be asked to agree to. This may happen at two levels (or even both). The individual user may be presented with a short summary on screen of what is permitted and forbidden, and you might even want so to configure your system that they had to click on a button marked 'I agree' before they could go any further. The subscribing institution might be offered a more detailed printed licence, which they would be asked to sign. The latter approach does allow for a certain amount of negotiation if necessary, although obviously you would wish to design your licence so that it was acceptable to most of your customers without involving both them and you in the time and cost of individual negotiation.

 

Some of your terms and conditions of use would depend on whether or not the electronic version is free or charged-for, but others would be applicable in any case. You would undoubtedly want to permit users to search, retrieve, and print articles from the journal; you might or might not also allow them to save individual articles electronically on their own computers. You should think about the best policy on allowing users to re-send articles, particularly electronically, to other people outside the subscribing institution – a number of publishers do allow this on a small-scale, personal basis. It is unlikely that you would want users to sell articles or redistribute them systematically on a large scale, or to mount all or even any of the content on other web sites.

 

While, particularly in the case of a journal for which a charge is made, you will probably want a signature or other indication of acceptance of the terms and conditions, the use of electronic means to enforce them is unlikely to make economic sense as the applicable technologies are still expensive. Within the academic and scientific communities, a culture of trust and honesty generally prevails.

 

Some very useful work has been done to produce 'model' licences, or licence frameworks, which will help you to devise a suitable licence for your own journals. All have been produced in discussion between publishers and customers, and thus try to address the concerns of both [82]. While it is unwise to use any of them as they stand, without considering carefully the implications of the various alternatives they present, they may save you a great deal of time in working out appropriate wording, and they will also help you to make sure that you don't leave anything out of your licence. One such example was devised in the UK by a group of university librarians and publishers. [83] The LibLicense site [84] also provides much useful information about licensing and licences, and allows you to sign up to an interesting discussion list. There are also a variety of sets of guidelines issued by universities and others, on what to look for in a licence.[85] If you can avoid writing your own licence from scratch, do so: use all the material at your disposal to incorporate useful wording from elsewhere. Many of the documents referred to are made freely available for just this purpose.

 

Some publishers have stopped requiring their customers to sign a licence at all – instead, they have a clear statement on the journal site of what is and is not permitted. So far, no problems have been reported with this approach, and of course it saves everyone a great deal of time and expense. [86], [87]

 

Conclusions

 

All of this is not intended to put you off publishing an electronic journal; it may very well be the best way of achieving your objectives. If, after weighing up all the advantages and disadvantages, you decide to go ahead with electronic publication, it is hoped that this brief introduction, and the plentiful information contained in the Reader, will give you the best possible chance of making your journal a great success.

 

Further reading

 

Scholarly e-publishing bibliography http://info.lib.uh.edu/sepb/sepb.html (6-5-02)

 

Scholarly e-publishing links http://info.lib.uh.edu/sepb/sepr.html (6-5-02)

 

Subject index to literature on electronic sources of information http://library.usask.ca/~dworacze/SUB_INT.HTM (6-5-02)

 

Acknowledgements

 

Thanks are due to many friends and colleagues who have shared their knowledge and experience with me over the years. In particular, I would like to thank Cliff Morgan, Don King, and Andrea Powell for their constructive input to this chapter.

 

References

 

  1. See, for example, http://www.espere.org (6-5-02)
  2. Swan, Alma and Brown, Sheridan, Authors and Electronic Publishing, ALPSP, Worthing, 2002;  summary available at http://www.alpsp.org/pub5.htm (6-5-02)
  3. Baldwin, Christine and Pullinger, David, ‘What readers value in academic journals’, Learned Publishing, Vol 13 No 4 (October 2000) pp. 229-240.  Also available at http://www.catchword.com/alpsp/09531513/v13n4/contp1-1.htm (6-5-02)
  4. Tenopir, Carol and King, Donald W, Towards Electronic Journals:  realities for scientists, librarians and publishers,  Special Libraries Association, Washington DC, 2000
  5. King, Donald W and Tenopir, Carol, 'Evolving journal costs:  implications for publishers, libraries and readers', Learned Publishing, Vol 12 No 4 (October 1999) pp 251-258.   Also available at http://www.catchword.com/alpsp/09531513/v12n4/contp1-1.htm (6-5-02)
  6. Halliday, Leah and Oppenheim, Charles, Economic Models of the Digital Library, Loughborough University, Loughborough, 1999.   http://www.ukoln.ac.uk/services/elib/papers/ukoln/emod-diglib/final-report.pdf (6-5-02)
  7. Halliday, Leah and Oppenheim, Charles, 'Economic Models of Digital-Only Journals', Serials, Vol 13 No 2 (July 2000) pp 59-66. 
  8. Shaw, Dennis & Moore, Howard (eds), Electronic Publishing in Science:  proceedings of the joint ICSU Press/UNESCO expert conference, February 1996, ICSU Press/UNESCO, Paris, 1996;  also available at http://associnst.ox.ac.uk/~icsuinfo/confproc.htm (6-5-02)
  9. Boyce, Peter, various articles on the Internet Society (http://www.aas.org/~pboyce/epubs/INET99.html) , e-publishing (http://cdsweb.u-strasbg.fr/misc/epub-talk/1.html) and in Serials Review
    http://www.aas.org/~pboyce/epubs/sr-art.html (6-5-02)
  10. Baron, Joel, 'Why we need information identifiers', Learned Publishing, Vol 10, No 2 (April 1997) pp 132-134.Also available at http://www.catchword.com/alpsp/09531513/v10n2/contp1-1.htm (6-5-02)
  11. Paskin, Norman, 'Information Identifiers', Learned Publishing, Vol 10, No 2 (April 1997) pp 135-156 (includes a useful glossary).   Also available at http://www.catchword.com/alpsp/09531513/v10n2/contp1-1.htm (6-5-02)
  12. Paskin, Norman, 'E-citations:  actionable identifiers and scholarly referencing', Learned Publishing, Vol 13 No 3 (July 2000) pp159-168.    Also available at http://www.catchword.com/alpsp/09531513/v13n3/contp1-1.htm (6-5-02)
  13. See also http://www.doi.org/about_the_doi.html (6-5-02)
  14. See http://www.crossref.org/faqs.htm(6-5-02)
  15. See http://www.leeds.ac.uk/cedars/documents/CIS05.htm (6-5-02)
  16. Russell, Kelly, Weinberger, Ellis and Stone, Andy, 'Preserving digital scholarship: the future is now', Learned Publishing, Vol 12 No 4 (October 1999) pp 271-280. Also available at http://www.catchword.com/alpsp/09531513/v12n4/contp1-1.htm  (6-5-02)
  17. Butler, Declan, 'Writing is on the web for science journals in print', Nature, Vol 397 No 6716 (21 Jan 1999) pp 195-200. Also available to subscribers or for purchase at http://www.nature.com/nature
  18. See JSTOR bibliographies at http://www.jstor.org/about/bibliography.html (6-5-02)
  19. Morris, Sally, ‘Archiving Electronic Publications:  what are the problems and wh