The journal of the future will not be the same as today; there may have to be a completely new paradigm for communication in science. For the purpose of this chapter, I shall assume that enough of the current paradigm is continued so that we can discuss the organisation and economics of primary science publishing, both of which will also change.
Economics depends on the way that science publishing is organised, the costs of so doing and the pricing set. With reduced library budgets, dropping circulation levels and consequent increased journal pricing, pricing has received much attention. It is different from the issues of the costs of journal publishing. I shall therefore give some background of the developing organisation of journals publishing, review some aspects of costs and of pricing and then comment on some of the slogans of the decade which propose solutions and commitments.
In the Post-WW2 years revenue support for most journals has moved from individual subscribers to university, government and industrial research library purchase. An average 80% of the number of subscriptions comes from these sources.
The consequent organisation of journals publishing developed to one including a variety of publisher (university press, learned society and commercial), library (university, government and industrial research) and national deposit (direct government-funded or indirect through universities). The introduction of library subscription agents (or vendors), current awareness, abstracting and indexing services has contributed to the still-growing complexity of scientific communications.
All those players increasingly use electronic communication to run their part of the system. The development of electronic document delivery (docdel), or individual article supply (IAS), electronic versions of journals and electronic-only journals offers continuing change.
Many current developments rely heavily upon the use of networked infrastructure using the Internet protocol and create the possibility of retrying some of the early experiments into electronic publishing that encompassed pre-print servers (then called poster papers), on-line discussion attached to every article, annotated abstracts and research community newsletters (Reference ) The asynchronicity and non-localisation of network applications means that informal communication previously transmitted 1-1 or at conferences and meetings can now be integrated with the formal part of scientific communication should that be desired.
That development presents an array of services across which libraries distribute their purchasing power to meet the information needs of their customer-scientists. This happens both directly in money transfer from journal purchase to current awareness services, tables of contents services and docdel services, and indirectly through the use of staff time (the majority of the expenditure in libraries on average) on search assistance, interlibrary loans (ILL) and on training on Usenet discussion groups, pre-print servers and Internet journals.
The aims of the journal from the angle of communication are fourfold: to encourage research, to aid the flow of information, to establish priority as quickly as possible and to report separate parts of a research programme (Reference ). Two essential components are the readers ability to know the reliability of what they are reading, by knowing that the research has been filtered, and the quality of presentation including mathematics, chemistry, and images etc.
The archiving aspect is one where journals publication is seen as keeping the record of scientific endeavour and therefore building a collective knowledge base. The product itself forms the raw material for archive.
Journal publishing has a social role, both in providing the information and in distributing rewards. The former gives readers information about the direction in which a discipline is headed and what the key researchers in the field are doing (particularly important for chemists (Reference ). In this sense, tables of contents are not just information about but are information used by researchers. The distribution of rewards concerns the cycle of credibility: the prestige, recognition, positions gained and grants that ensue by publishing activities (Reference ).
There is also a wider social role in that published discoveries are communicated to the public via science writers in newspapers, television programs etc., which in turn brings recognition to science and supports its continued funding.
The refereeing process is interactive and leads not only to decisions about whether or not to publish but also how to improve the work and the presentation. Nature, for example, insists on all directly relevant work being cited - an insistence which not all researchers like, especially where funding decisions are important. This, however, leads to a specific dependability. There will be continued need for some independent arbitration.
(It should be noted that the use of publication for performance assessment increases the administrative load as scientists submit increasing numbers of papers to high impact journals in the hope of getting a higher assessment value.)
These tasks in which publishers add quality will be needed as much, if not more, in the electronic world. Ensuring consistency and quality of content, re-design for access and the usability of the interface will be key areas in helping researchers. Very little progress has been made in the area of electronic facilities to support browsing and reading; the majority of successful uses are in electronic delivery for print (Reference ). I believe that the tasks of ensuring quality of content and presentation will continue and that the expertise found in publishing houses will be required in the future.
The particular paradigm of the journal containing papers and appearing in issues will change in the electronic arena. We do not know what it will change to, but there are certain aspects which should be kept (Reference ):
1. A title, by which it can be referred and its principally containing original research.
2. A system of peer review.
3. Dates so that precedence can be established (received and accepted).
4. An archive of accepted content (even if later amended).
5. If there is a print/CD record and an electronic version, then a statement of the relationship.
There are three reasons why one might expect papers to get shorter in electronic communication:
1. The speed of publication affects the length; one might expect that the use of a faster medium would shorten the length of papers, much as we see in Letters sections.
2. The utilitarian measures used to assess performance of researchers leads to an increasing pressure to publish papers, which are sometimes shorter than they would otherwise have been.
3. There may be less need to fully present the background, when other papers can be readily hypertext linked and what their authors say can be seen in full context, rather than in summarised form.
The result of these might well be an increasing number of shorter papers published. The logical conclusion from a record of scientific endeavour approach might be a fully searchable electronic laboratory world notebook! That, however, would not match my requirement of the readers easily knowing whether the content can be relied on or not.
Some believe that the lack of limit on page size will encourage papers to get longer, and there is evidence for this in the possibility of electronic supplementary material offered by some journals (Reference ).
Exploiting the new opportunities of the electronic medium in the journal paper.
There are many opportunities in the electronic medium to make communication more effective. These include integrating multimedia material and deposit of supporting data. The former would otherwise be communicated through separate channels (Reference ), incorporating animation and sound and using computer power for data visualisation (e.g. 3-D molecule rotation). The latter facilitates linking references to text, links to databanks and bi-directional communication. In some areas a paper is no longer accepted unless there is deposit of supporting information into a databank, e.g. in genetics research (Reference ) There are some particularly interesting enhancements in using computer-based systems; for example if there is a link from a reference to the paper cited, then the computer can calculate all the papers which cite the one the reader is looking at. This allows the scientist to follow through research results much more easily than at present.
Scientists concern themselves with who publishes a journal to the extent that they can satisfy themselves of the reliability and quality of the information. Primarily they want to be able to access all the relevant research in their research domain quickly and easily. Individualised journals are possible in the electronic arena, by setting a use profile which can automatically be altered using relevance feedback and other new tools. This can reduce the information overload felt with the demand to keep up-to- date.
Changing science environment
In looking at the organisation of scientific publishing, one cannot ignore the scientific environment in which it operates. The limits of library budgets are not vendettas against one way of organising the system as much as a reflection of the current priorities of governments. At the end of the millennium, rich-world governments are not placing science at as high a priority as they did following the rapid developments during WW2 and before the loss of belief that science would solve many problems which they now recognise require difficult political choices. The consequences are reduced funding, moves towards creating fewer larger centres of expertise to cut down on overhead costs, and funding moved from basic towards the more applicable areas of research. These might be expected to change the requirements of scientific communication. In the less rich two-thirds world, many scientists having trained in centres of expertise, are researching without supportive infrastructures. This too may contribute to changes in the way that science is done and communicated.
It is in this changing context of players and stakeholders and of the development of the journal that we consider the possible organisation of scientific publishing.
The first scenario is a continuation and extension of what is happening now, with replication of many parts of what is published and with libraries and individuals distributing their resources across those many services: abstract and index, current awareness, table of contents searching, citation linking, docdel, electronic versions, preprint servers, etc. The existing economics cannot be maintained in this scenario because of the distribution of revenue through a system only sustainable with fewer players. Nor can a individual article model be sustained without that cost covering all the rejected papers, as well as the ones published.
The second scenario says that many of the newer services are redundant in an all-electronic publishing scenario - that current awareness, citation linking, searching etc. could be organised by distributed search software onto full-text multimedia databases stored just once (whoever the publisher might be) and so reduce the costs of the total system.
Both scenarios seek to provide added functionality while also reducing the information overload and providing quicker and more effective access to the information.
The directions above could result in very different economic systems. In particular what is being traded varies from services through to purchase of or rights to use documents or even parts of documents, for example for permission to use figures, multi-media and quotes.
In the former multi-player model one might expect an approach more centred on licences and associated with particular rights, for example to print, to use for teaching, to exploit in current awareness services, to add hypertext links etc. This would move journal publishing business models much closer to the world of books and television in the way that they trade in rights.
In the second case the business model might maintain more of the current model, including many pay- per-use sections as well as subscriptions for regularly accessed content and licences.
Which is likely to happen? There is evidence both ways. New players are beginning to offer value- added services (e.g. in citation-linking, complete coverage of disciplines etc.) and intermediaries are seeking to integrate some of the functions previously done by others (e.g. delivery of original text on behalf of publishers). On the other hand, existing publishers (particularly learned societies) are themselves integrating services, while some larger commercial publishers are investing in software companies and intermediaries with the apparent intention of moving towards the second scenario.
Faced with such change, predictions are not easy, which is why I personally continue to advocate collaborative work to ensure that the needs of scientists are not lost in the changes taking place in the means to deliver what they require for science.
Different methods are used to examine costs depending on the reason for the enquiry (Reference ). If it is cost control, then a study would look at components of the system which consume the most resources and the method used would be cost per function; whereas if the focus was on cost for each institution, then cost per participant measure would be more appropriate. Among other measures is cost by item; what is the unit cost of the end-product? In many comparisons of cost between print and electronic, different measures are falsely compared.
Will there be a reduction in cost by the use of electronic preparation? Some costs are saved by the use of author-generated text files, others saved by the use of electronic communication, at least while it remains relatively cheap for the end-user (in this case the publishers administration). Experience of putting scientific papers into SGML from non-SGML imposes an extra costs of 20-30%, but the cost to the publisher might alter when authors do this themselves using new tools. Academics have widely varying responses and abilities in attending to the extra work entailed; some want to do this, others do not. Other fixed costs include running an editorial office, the administrative costs of editorial processing, the sub-editing etc. as well as marketing. While some have thought that the existing economics of journal publishing for publishers can be maintained in some form by moving the costs and cutting on other costs (Reference ) , the range and extent of potential change considered above makes this unlikely.
At the authors end, to prepare papers is already time-consuming, particularly when different journals have different requirements. For scientists to provide perfect electronic copy will require new tools to help them in ensuring consistent structuring, internal referencing and use of standard nomenclature; moreover they will need to pay for the sub-editing currently done by others if they want to be read (particularly affecting non-English authors).
One proposal is that authors not only pay the direct costs for all the preparation of their own papers, but that they also pay to cover the storage and distribution costs (Reference ).
At the readers end, to access and use electronic journals requires computers of sufficient specification, appropriate software, communications and access to network infrastructure. Each of these costs money, though the burden falls in different places depending on country and sector. For scientists in the developing world, the costs of up-to-date equipment and infrastructure required to present good graphical displays are often prohibitive.
The marginal costs of electronic versions of papers and simple electronic journals (as currently extant) appear to be cheaper if the cost of computers, software, telecommunications, network infrastructure and government funding is excluded. Clearly one of the most undesirable things to occur would be a time or byte charge for use of networks - scientists want to spend lots of time with material of interest and publishers want readers to spend time on their material rather than on others as it is then more likely to be cited.
Pricing for journal content has been by subscription per title and also recently by paying a copyright fee appropriate to the circumstances of taking images of pages. While the main market for journals has been libraries post-WW2, those individual scientists who want journals have most commonly had them priced to cover the marginal costs. Many publishers have also helped organise pass-it-ons or differential prices for scientists unable to pay for journals in Eastern Europe and the non-rich world (Reference ).
In the electronic arena, there are many more possible models for pricing. In addition to subscriptions of single titles, there are subscription to a corpus of information (by one publisher (Reference ) or consortium), pay-per-use for viewing, printing or searching, or use of sophisticated software only occasionally used, colour printing etc.; and licences granted per site, sector or nation (Reference ). One model is to type these as follows (Reference ).
By intended audience: e.g. single user, multiple user, single institution, multiple institution
By type of access: e.g. on-line timeshare, locally loaded databases, CD-ROM databases
By type of information e.g. citations, full-text, data.
Any additions the reader of this chapter might make begin to emphasise the potential complexity (Reference ).
Will pricing change? The costs to libraries include the issues of pricing set by publishers in purchase or access, as well as staff, buildings, shelving, computing, cataloguing, storing and searching for access, and any replacement and binding costs, etc. Such a question is clearly of interest in the case of limited budgets, as well as strategies for reducing costs by, for example, buying less (using an access rather than holding policy), by storing less (by buying and holding only current issues and throwing it away and relying on access for older journals(Reference ), by storing holdings off-site in cheaper premises, and by enabling researchers to do their own computer searches. Research libraries already pay more for information from one journal, because they pay both for subscriptions, increasing in price, and for other services that give information about the content of those subscriptions in order to offer added value to their customers. With limited budgets this would be unsustainable were it not for all the other changes and opportunities.
The most critical element in pricing, however, is the organisation of scientific journal publishing and it is to that we return with some considerations of some slogans of the decade.
there is not enough material to make it worth scientists changing their research habits
the interface still does not provide scientists with the kind of browsing required
researching and reading on-screen constrains scientists to one place
the printer does not print the mathematics and figures in a usable enough form
up-grading equipment to use in scientists offices would cost too much.
The apparent slowness of publishers could be seen as proper caution mixed with fear and confusion about what exactly to do (Reference ). Given the increase of cost in the short-term to put everything into SGML, to invest in access to the network infrastructure, to train staff and so develop electronic publishing, is there sufficient justification for doing so without strong evidence that that is what is wanted? Or in other words, is there enough evidence to justify an increase in price of journal subscriptions and enough support from the scientific community to pay for that investment in the transition? And who will stand up and say exactly what that implementation should be? I would suggest that a collaborative approach involving government-funding is the right approach here (Reference ).
We do not know how many similar factors need to be present in order to give the same level of use. Other sub-disciplines using the same technical system are much more divided in their use, some using it prolifically, others not at all, with variable quality being given as one reason.
At present I see no great evidence for a substantial diminution of work associated with journal publishing over the next decade as it gradually moves electronic. That work has either to be done by existing staff or by new staff taken on. Few academic institutions have chosen to take the initiative in print publishing, so what indicates that they would like to do so? Some research results indicate that scientists are not at present as unhappy as would be required for them to do so (Reference ). The altruism of academics is as mixed as the commercialism of other groups. If one examines learned societies, run by researchers who are largely academics, then there is no evidence of great difference between them and the policies of commercial publishers. Indeed one might question some learned society journal pricing policies and what is done with the surplus- over-cost. Local or national funding might also inhibit some of the actions taken to ensure global access to scientists, particularly in the less-rich world. Whether or not differential charging between academic institutions and industrial research centres would be introduced when the latter are also contributing scientific research, would also have to be tackled. It is not immediately obvious which way the overall equation would work out.
The changes arouse strong emotions as economic systems change and individual institutions suffer. What we can expect is a period of tumult. I have tried to show that some of the more simplistic approaches to costs and to pricing are likely to fail in an increasingly complex world of science publishing. I have also taken some of the slogans of the decade and examined the case in the context of the economics of science publishing.
If I were forced to take a stand on the most likely outcome, I would predict that the majority of science journals will go electronic when the technology supports them to do so, the top science journals will be published both in print and electronically. Publishers will be contributing their expertise in quality, through sub-editing, presentation and administration of peer review, whether in their current organisations or not. There will be an increasing separation in the way in which the functions of a journal are provided: the communication of research, the archiving and the social/community aspects. There will be a more extensive layering of access via searching, reference linking and titles selection, with new techniques developed to do browsing. The economic models will settle down to a mixture of site or sector license, subscriptions and pay-per-use, possibly not exclusive to each other.
The most productive approach is one where scientists collaborate in government-funded projects to explore the future and in which they can say what it is that they want from the many opportunities. I do not believe that the system that has been dominant for the last 50 years will remain so, but neither is it easy to say what will replace it.
The future is, however, changing slightly more slowly but much more dramatically than we usually expect. Who could have anticipated the entire system with its many players and surprising degrees of standardisation that has developed in journals publishing - all to serve the cause of scientists and so science?
Campbell R 1993 Changing Roles in Information Distribution: The Primary Publisher.
Coles B 1993 The STM Information System in the UK London, Royal Society
East H, Sheppard E & Jeal Y A Huge Leap Forward British Library R& D Report 6202
King D W & Roderer N K 1978 Systems Research of Scientific and Technical Communication in the United States: The Electronic Alternative to Communication through Paper-based journals King Research Inc. NTIS PB 281 837 /851
Lancaster F W 1978 Towards a paperless Information System New York: Academic Press.
Odlyzko, A M 1995 Tragic loss or good riddance? The impending demise of the traditional scholarly journals Int. J. Human-Computer Studies 42, 71-122
Okerson A S & ODonnell J J (Ed) 1995 Scholarly Journals at the Crossroads: A Subversive Proposal for Electronic Publishing. publ. Association of Research Libraries.
Page G, Campbell R and Meadows J 1994 Journals Publishing 2nd Ed London: Butterworths.
Schuyler L 1995 Assessing the impact of the Internet and online services on traditional publishers, Proceedings of On-line Information Conference, December 1995, London.
Shackel B 1991 BLEND-9: Overview and Appraisal British Library Research Paper 82
See, for example, Licklider J C R 1966 A crux in Scientific and Technical Communication American Psychology 21, 1044-1051. It is interesting to note that the more innovative and exciting writing about possible electronic communication happened before particular solutions (TCP/IP, WWW) and seem less shaped by the possibilities presented by current particularities.
For further reading see Singleton A 1994 The Scientific Journal Information UK Outlooks 7 pub: British Library
Pullinger D J 1986 Chit-chat to Electronic Journals: Computer Conferencing supports Scientific Communication IEEE Transactions on Professional Communication PC29(1), 23-29
Meadows A J 1974 Communication in Science London: Butterworth
Gould CG & Pearce K 1991 Information Needs in the Sciences: An Assessment publ. Research Libraries Group, California, USA
Latour b and Woolgar S 1986 Laboratory Life: the construction of scientific facts 2nd ed. Princeton, NJ: Princeton University Press.
Gould CG & Pearce K op cit.
Pullinger D J 1994 The SuperJournal Project Institute of Physics Publishing, Bristol UK
Discussed in Serials Publishers Executive of the Publishers Association 1993.
e.g. Nature Leader 12 October 1995 and the first Supplementary Information stored on WWW site 8 February 1996
E.g. videos on request, for example from the American Astrophysical Journal.
See also discussion in Natures Leader 18 January 1996.
See for example the exploratory work done in Pullinger D J et al 1995 STAMPEDE Project Report IE750 to European Commission DGXIII on Billing Charging and Paying for STM Journal Information on Internet.
Singleton A and Pullinger D J Ways of viewing costs in Journals Electronic Publishing Review March 1984
Page G Networking and the economics of journal publishing Learned Publishing 5(3), 137 - 144
Morris S 1990 Pricing for Europe Serials 3(3)
Bodinham M and Campbell R 1992 The survival of the Journal: a simple spreadsheet model for predicting the outcome Learned Publishing 5(3), 153-159
Harnad S 1995 The Guardian
The British Medical Association has an explicit brief to help in this way.
For example, Academic Press national site licence in the UK, 1995
Hunter K has pointed out that this might cause problems in 1992 The National Site License Model Serials Review 18(1/2), 71-72.
Fayen E G Pricing Models for Electronic Information Paper presented at STM Meeting London November 1992.
IFRROs recent questionnaire to publishers being the most complete I have seen.
Lines M B 1992 The publication and availability of Scientific and Technical papers: and analysis of requirements and the suitability of different means of meeting them. Journal of Documentation 48, 201-219 [to be checked]
see for example Olsen op cit.
Singleton A 1994 Charging for Information in Electronic Products Paper presented at the 10th International Learned Journals Seminar April 1994.
As indeed the UK has in its eLib (Electronic Libraries) programme and US National Science Foundation has in its Digital Libraries Projects.
Schauder D 1994 Electronic Publishing of Professional Articles: Attitudes of Academics and Implication for the Scholarly Communication Industry. Journal of the American Society for Information Science 45(2), 73-100
Listing by Author's Name, in alphabetical order
Return to the ICSU Press/UNESCO Conference Programme Homepage
University of Illinois at Urbana-Champaign
The Library of the University of Illinois at Urbana-Champaign
Comments to: Tim Cole