Scholarly communication exists in many forms but the most important is the publication of papers in peer-reviewed journals and it is this that I shall concentrate on in this article.
Why is scholarly communication so important?
Scholarly communication is an essential part of the scientific process; indeed it can be seen as the lifeblood of science. The publication of papers in a peer-reviewed journal provides a degree of accreditation of the work of a scientist and, equally importantly, it makes the results of that work available to the wider user community. The publication of papers is critical for driving future scientific research, and of more importance to individual scientists, it is largely responsible for their career development. The publication record of its staff heavily influences the future success of a research institute and it is for these reasons that scholarly communication is so important.
There are three main parties with an interest in scholarly communication and these can be represented by a construct known as the publication chain:
Authors ---------- Publishers ----------- Readers
Scientists are stakeholders in each part of the chain. Scientists write the papers, they work for the publishers as reviewers of papers, and they are the primary readers of the papers. Libraries supply information to the scientific community so they form part of the readers group.
What is the current model of scholarly communication?
The current model of scholarly communication has been characterised
as a 'pay-to-read' model where:
Scientists do the science - they seek to publish the results in a peer-reviewed journal - they select a target journal and submit the article, with a signed copyright agreement - the article is reviewed and it is either: accepted; accepted subject to revisions; or rejected, if it is rejected the scientist goes back and selects another journal - the article is published within the journal - the library, or individual, subscribes, or uses a document delivery service, to gain access to the article.
One of the major problems with this model is the time taken from submission to publication. After submission of an article time can be lost at each stage of the process and it is common for an article to appear in print 6-8 months after submission. This can be detrimental to the scientific community at large and for the individual scientist. The other significant problem with this model surrounds the copyright agreements that the authors sign upon submission of an article. These agreements give the publishers all of the economic rights and exclusive usage rights on that article. These agreements limit a scientist's ability to use their own work without permission from the publisher. A scientist cannot place a PDF of their paper on their own, or their institution's, website without specific permission and use of the paper for teaching and future research also requires specific permission from the publishers.
Following pressure from the large American universities, where faculty members are forbidden from signing agreements that give publishers full usage rights to the article, the position of many publishers is slowly changing with regard to authors retaining some usage rights over the article.
Why do we want to change the system of scholarly communication?
The current model of scholarly communication has served the scientific
community well for hundreds of years so why do we need to change it now?
Increasingly the scientific and library communities believe that the model
is weighted too heavily in favour of the publishers and that there is a
need to restore a balance within the publication chain. Only by restoring
the balance will the scholarly communication system meet the needs of every
group within the chain. I shall look briefly at the issues for each of
the groups within the chain.
Authors - Scientists have to publish and to do so they must sign the copyright agreements provided by the publishers. If they do not sign the agreement then the publisher will not accept the article. The publisher thus holds the stronger bargaining position and it is virtually impossible for an individual scientist to renegotiate this agreement. Scientists provide the peer review function for the publishers with little reward or recognition for doing so. Arguably it is the peer review process that adds the most significant value to a journal with the publisher receiving the majority of the rewards for overseeing the process.
Publishers - The scientific technical and medical (STM) publishing marketplace is dominated by a small number of large commercial publishers, with further consolidation occurring regularly. There is little effective competition within the marketplace and this has led to a situation where publishers can set prices at almost any level they wish. Profits for the commercial STM publishers have regularly been between 30 and 40% of turnover, which could be considered excessive, especially given that the scientific community adds most of the value to their product. The copyright agreements give them exclusive usage rights and an ongoing income stream when an article is requested from a document supply service.
Readers - Journal price inflation for institutional subscriptions was 7% in 2002 and in the preceding years was regularly in double figures. This is significantly higher than the inflation rate within the United Kingdom as a whole. Library acquisitions budgets have not kept pace with this inflation rate leading to a very significant reduction in the real spending power a library has. Libraries therefore have to cancel subscriptions and to protect their income the publishers have to increase the price further, leading to a vicious circle of cancellations and price rises. To illustrate the point the John Innes Centre Library spends 80% of its non-staff budget on acquiring journals and another 6% on document delivery services. Many scientific libraries find themselves in a similar position and with the current model the situation is likely to get worse. Furthermore, publishers regularly change reprint policies, generally with the effect of reducing the number of free reprints they distribute to the authors of the article.
Other factors also influence the desire to change the model. These include a push to 'open' science, particularly publicly funded science, to the wider corporate and public worlds through activities such as knowledge transfer and public understanding of science. Opportunities have been presented by the growth in the Internet and other new technologies. One of the major opportunities has been to reduce the delay from submission to publication but as yet little has happened in this direction and there is a growing sense of frustration at this. The current model of scholarly communication makes all of these opportunities difficult to deliver. New technologies allow readers to focus more closely on an individual article and thus there is a growing question surrounding the value of the 'brand' that a journal represents.
What changes are possible?
The lack of competition in the market is one of the major problems and one of the first initiatives by SPARC(1) was to try and introduce competition through the establishment of competitor titles. An example of this is Organic Letters published by the American Chemical Society at £1770 a year competing directly with Tetrahedron Letters published by Elsevier Science at £5600 a year. To facilitate competition they have also overseen the resignation of a number of editorial boards en masse to set up competitor titles. The Public Library of Science(2), signed by over 30,000 scientists worldwide, and the Budapest Open Access Initiative(3), signed by over 2500 scientists worldwide, illustrate the strength of feeling within the scientific community surrounding this issue.
A number of alternative publishing models have been proposed as a means of bringing about change. The primary example of this is the 'pay-to-publish' model with BioMed Central (4) being the primary exponent at this point in time. With this model authors pay a processing fee on submission of an article, generally around the $500 mark. The articles are fully peer- reviewed and are published on the web immediately on acceptance. Key features of this model are:
- A greatly reduced time from submission to publication;
- The author retains the copyright on the article giving them full control over its future usage;
- The article is available to anyone with an internet connection; and
- It has the potential to allow an accurate assessment of the use of an individual article.
Pre-print servers represent another publication model but they lack peer-review so may not work well within the broad area of life sciences, although they have worked successfully within disciplines such as physics.
What is stopping us from changing?
The pay-to-publish model would create a much more balanced relationship between authors, publishers and readers so why haven't we seen significant changes occurring? It is only recently that this issue has come to the fore so we are still in an awareness-raising stage but there are three potential barriers to acceptance of the change:
- Article processing charge - Page charges have never been popular and the article processing charge is seen as being a page charge system. It should be remembered that some journals use page charges, particularly for colour illustrations and the charges vary considerably from journal to journal. If an institution becomes a member of BioMed Central then the article processing charge is waived.
- Impact factors - BioMed Central journals do not have impact factors and will not have them for at least another 2 years. Impact factors need a publication period of 3 years for analysis purposes, however Thomson ISI who control the production of impact factors have included BMC journals within Web of Science, which is a prerequisite of gaining an impact factor. Funding bodies, assessment exercises and the need to build a CV all place pressure on scientists to publish in high impact factor journals, so publishing in a journal with no impact factor is a difficult step to take. The biggest difficulty with impact factors is that they measure the success of a journal as a whole and not the individual papers within that journal. However, until funding and assessment bodies take this on board, they will remain the critical benchmark of scientific performance.
- Learned Societies - The future of learned societies represents a concern for many within the scientific community if 'pay-to-publish' becomes the norm. Learned societies generally use the income from their journal to subsidise other activities and without this income these other activities may be threatened.
What can be done to bring about change?
The biggest step we can take is to highlight the difficulties inherent in evaluating scientific performance using impact factors as the key measure. Successful achievement of this will give the scientific community confidence to publish in whatever fashion they choose.
Ultimately it will be the scientific community that decides whether a change is necessary and what form that change will take but I hope I have made you aware of some of the difficulties within the current system. For further information please feel free to contact me directly or talk to the library within your own institutions.
Useful internet addresses:
(1) SPARC: www.arl.org/sparc/home/index.asp?page=0
(2) Public Library of Science: www. publiclibraryofscience.org/
(3) Budapest Open Access Initiative: www.soros.org/openaccess/index.shtml
(4) BioMed Central: www.biomedcentral.com
Kenneth Dick BA (Hons) PGDip MCLIP
Acting Librarian, John Innes Centre, Norwich