Return to Nature Debates: e-access

No Free Lunch!

Martin Frank
Executive Director, American Physiological Society

In the past few months, there has been a movement to encourage scientists to boycott (meaning to cease subscribing to, submitting to and reviewing for) any journal that refuses to provide free access to content within six months of publication through PubMed Central (PMC) or other free online archives. The American Physiological Society (APS) agrees with the need to make content freely available to the scientific community at large and, as described below, has implemented several avenues to do so. However, the society believes that the current Public Library of Science (PLS) proposal is seriously flawed in several ways, including its inability to protect the integrity of published scientific data.

Summarized below is a brief history of this movement, the problems that APS sees with the PLS proposal, and the efforts that APS has been making to disseminate freely the contents of its recent and archived publications.

About two years ago, Pat Brown, Stanford University, and Harold Varmus, then the Director of the National Institutes of Health (NIH), proposed to change how biomedical scientists publish their research. In 1999, the goal was to create �E-Biomed�, the biological counterpart of the physics archive at Los Alamos National Laboratories. The current effort to change scientific publishing centres on several elements including the PLS, BioMed Central (BMC) and PMC.

Supporters of E-Biomed proposed using the Internet to publish pre-prints of manuscripts, in a reviewed and an un-reviewed format. This electronic forum was intended to replace the existing system for peer review and publication of biomedical research by non-profit and commercial publishers alike. Its promoters argued that because the US government is the primary funding source for most of the biomedical research, the resulting scientific content should be made freely available to the entire scientific community. After all, the argument ran, why should private entities � even scholarly societies � be allowed to profit from work supported by US taxpayer dollars?

The proposal stimulated enormous discussion about the use of the Internet to disseminate scientific content and the future of scholarly publication. It is no secret that the APS opposed the original E-Biomed concept, and although it is important to review why the society opposed E-Biomed, it is equally important to take note of the catalytic role this proposal has played in furthering the cause of free, electronic dissemination of scientific content.

The APS objected to E-Biomed because it would have undermined both our ability to safeguard the integrity of journal contents and the economic viability of our scholarly journals and the service activities that they support. As with many other scholarly societies, APS uses journal revenues to run and subsidize other programmes, particularly in the areas of education, outreach to under-represented minorities, public affairs, student awards and scientific meetings. Yet the E-Biomed backers told these societies that it was wrong for them to generate a profit through subscription fees from research funded by government grants. Instead, the E-Biomed proponents suggested that societies could overcome this problem by increasing dues and meeting-registration fees to compensate for the lost subscription revenue.

An alternative option suggested by Varmus was for publishers to ask authors to shoulder the full cost of peer review and other costs related to transforming a manuscript into a finished publication. For APS journals, we estimated that this would involve a surcharge to authors of $2,500 to $3,000 per manuscript � more with colour figures � to distribute the research content for free to the community. This aspect of the E-Biomed proposal failed to recognize the fact that foreign scientists submit about half of the articles to many society journals. These scientists could not charge such fees to an NIH grant and they do not have such resources available from their countries.

Another concern about the original E-Biomed proposal was that it would have placed the NIH in the position of not only reviewing and funding scientific proposals but also of serving as arbiter for their publication. Furthermore, it would also have put the NIH in the position of forcing the scientific community to abandon its traditional means of information dissemination, and the measure of quality and impact afforded by ISI journal impact factors.

As a non-profit scientific society that publishes peer-reviewed journals, the APS took exception to the fact that throughout 1999, the promoters of E-Biomed seemed to be targeting non-profit society publishers. Scientific societies that publish journals traditionally have had the lowest subscription prices for the highest impact journals. It was difficult to understand why society journals were being targeted when the impetus for E-Biomed was that librarians were concerned about the over-priced, relatively low-impact commercially published journals.

On the plus side of the balance sheet, the E-Biomed proposal proved to be a stimulus that led many publishers to reconsider how best to provide the biomedical literature to the scientific community. The APS and other scientific society publishers had already been working since 1996 with HighWire Press to create an electronic library of biomedical literature. HighWire Press currently serves as the host for over 260 journals published by non-profit society publishers. These journals are fully searchable and linked to each other through the references in the articles, creating a virtual library of scientific information. In addition, APS and many of the other participating non-profit publishers remove access controls after 12 months or less to allow the public free and unrestricted access to the content in their journals. At present, there are over 250,000 articles available in the free-access library, making the HighWire Press collection the largest repository of free life-science literature in the world. Interestingly, this collection does not meet the PLS criteria for free access.

Although E-Biomed failed because of the reasons noted above as well as its projected role as preprint server, it did serve as a stimulus for publishers to review how journal content is priced and how to ensure free and open access to the content after some defined period of time. It also resulted in the creation of PMC as a permanent archive of free biomedical literature. Both Pat Brown and Harold Varmus currently serve on the PMC Advisory Committee.

PMC is managed by the National Center for Biotechnology Information, a part of the National Library of Medicine. PMC initially called for publishers to deposit their digital content into government computers to allow for enhanced searchability and archiving. The cost of the archiving was to be born by the US federal government, but the cost of converting digital files from the original online publication format to one compatible with PMC was to be borne by the publishers. It seemed doubly disadvantageous for scientific societies to be asked to bear these extra costs when the whole premise of E-Biomed was that they were not entitled to make money from the manuscripts that were published in their journals. Many publishers were also concerned by the lack of tangible plans for how PMC would ensure the integrity of the information in this archive.

To date, only eight established journals have deposited their content with PMC, and no doubt this has caused considerable frustration for the promoters of a federally managed, archive of free scientific literature. The only other content to be found on PMC are the articles published in a number of online journals created by BMC. BMC is a recently established for-profit online publishing house that is committed to making original research articles freely available to all. Interestingly, Harold Varmus serves on the BMC editorial directorate, which might be why articles published in the BMC journals are cited immediately in PubMed and archived in PMC. It should be noted that APS had to petition to have Physiological Genomics cited in PubMed and it took the society 14 years to get News in Physiological Sciences cited in PubMed. Similarly, it took AAAS two years to get the Signal Transduction Knowledge Environment included in PubMed.

Perhaps it was the reluctance of established scientific publishers to adopt PMC that caused a group of PMC supporters to band together under the banner of the PLS to ask the scientific community to boycott any journal that refuses to release its content into PMC within six months of publication. It is worth noting that there is a significant overlap between those who signed the PLS call for a boycott and those who serve on the PMC Open Letter Advocacy Group and the BMC editorial directorate.

The APS has no argument with the PLS�s desire to make the scientific literature more freely available to the public and members of the scientific community, especially those in developing countries. Since the mid-1980s the APS has been sending its journals � at no cost to the recipients � to institutions in the former Soviet Union, Eastern Europe and sub-Saharan Africa. With the advent of electronic publication, APS and other society publishers are making the journals available in those developing countries with adequate Internet connectivity. For those lacking such connectivity, especially those in sub-Saharan Africa, APS makes the content available through SatelLife, a satellite transmission system.

Similarly, the APS has been making the content of our journals freely accessible 12 months after publication since 1999. In so doing, APS has been able to maintain the subscription revenues that allow us to create programmes designed to benefit the membership and the scientific community. In addition, these revenues will allow the society to post online �legacy data�, articles published from 1996 back to when each of the society�s journals were first published. The APS Council has authorized the expenditure of $250,000 to prepare and scan the content of journal articles published between 1986 and 1996 for posting on the journal sites in 2001. Eventually, the APS hopes to make the entire content of the American Journal of Physiology available online back to its start in 1898. This material will be freely available, but the APS can only think of preparing it for online publication because of the revenue the society generates from its journal programme.

According to their recent statements, the proponents of the PLS believe that publishers have no right to manage the content or profit from the dissemination of journal articles. One of the PLS advocates, Michael Eisen, Lawrence Berkeley National Laboratory, suggested that publishers were like midwives who helped with the �birth� of an article and then refused to turn the �baby� over to the �parent�. This inept analogy takes no account of the fact that even a midwife charges a fee for her services! The PLS proponents also condemn publishers for generating profits through subscription fees to individual and institutional subscribers when a significant portion of the research being published is funded by the federal government. It seems not to matter to them that society publishers use these profits to benefit their members and the scientific community.

Interestingly, some of the PLS signatories are also investigators who have profited from federal support for their research. NIH has provided initial support for many of the investigators who have gone on to convert their research into profitable biotech companies, yet they do not return any revenue or royalties to the NIH to support future research efforts of the scientific community. If it is wrong for publishers to profit from research supported by the federal government, is it not wrong for individual investigators to profit? This is an issue that the promoters of the PLS fail to address.

Along with economic consideration about when to release content to the public, there are legitimate questions about why information has to be deposited in a single government database in order to give the public access to it. Originally it was suggested that content had to be deposited into PMC to allow for enhanced search capabilities. However, just what those enhanced search capabilities would entail have not been defined or communicated to the publishers. In addition, search engines such as Alta Vista and Yahoo can already effectively search content that is not deposited on their computers. Neither have the proponents of PMC described their archiving plans. The APS is currently working with HighWire Press on two sophisticated archiving projects, known as LOCKSS and Dark Cave.

The APS also finds the call from the PLS to cede ownership or copyright of the articles unacceptable. The APS accepts responsibility for the content that it publishes in its journals, agreeing to correct the literature should there be errata and to investigate ethical questions should they arise. If the society were to relinquish the copyright after six months, as called for by the PLS, the public would have to rely on no one, or perhaps only authors themselves, to make corrections to the literature and to investigate misconduct. This also would allow others to profit from the �value� added by APS to the published journal articles by creating collections for re-sale to the public.

The APS has set its own priorities for providing public access to the content of its journals as stated above. Some of these objectives are less than what the PLS proponents seek, but some of these objectives go further than their demands. Furthermore, APS has already implemented significant portions of its programme to make the contents of its journals available, and the results can be seen at the HighWire Press website for the APS journals.

The APS believes that the proposed boycott ignores the diligent efforts of non-profit scientific publishers to make their content freely available to the public in a timely fashion while still covering the real costs of producing high-quality journals. The APS also believes that coercion is an unacceptable means of effecting change. It would be preferable for the representatives of PLS to meet with the non-profit publishing community to discuss how best to attain their goals. Just as with E-Biomed, the ideas advocated by PLS are more likely to be adopted through constructive discussions rather than by the threat of a boycott. In the mean time, the APS strongly urges our members and authors not to sign the boycott letter. After all, there is no such thing as a free lunch.


What the APS is doing to give free access to its content

APS journals are available through HighWire Press as part of an online collection of more than 260 life-science journals.

The content of the entire HighWire collection is completely searchable.

Since November 1999, the APS has made the content of its journals available free online 12 months after publication.

APS journal content is available on the HighWire Press site as part of a database with over 250,000 free articles.

APS is in the process of scanning and placing �legacy� content online so that scientists will have online, searchable access to research published before 1996.

APS is working with HighWire Press on two types of electronic archives to protect scientific content far into the future.



Martin Frank joined the Department of Physiology, George Washington University School of Medicine, Washington DC, in 1975, as an assistant professor. From 1978 to 1985, he served as the Executive Secretary, Physiology Study Section, Division of Research Grants, National Institutes of Health, Bethesda, Maryland. From 1983 to 1985, he was a member of the Senior Executive Service Candidate Development Program, Department of Health and Human Services, Washington DC. As part of the programme, he served as a policy analyst in the Office of the Assistant Secretary of Health, DHHS. In 1985, he accepted his current position as executive director of the American Physiological Society, Bethesda.

A version of this article, written for the Nature web debate, is also published as an editorial in the APS�s newsletter, The Physiologist.