Select Committee on Science and Technology Written Evidence


Memorandum from Professor Michael Ashburner, University of Cambridge

  "Science is public knowledge". John Ziman

  Michael Ashburner is Professor of Biology in the Department of Genetics, University of Cambridge. He was, for the period 1994-2001, first Research Coordinator and then Joint-Head of the European Bioinformatics Institute, the Hinxton Outstation of the European Molecular Biology Laboratory. He is on the editorial boards of several Open Access scientific journals.

  1.  Preamble and some background on the "sociology" of publishing scientific research.

  1.1.  What defines science ? John Ziman, Emeritus Professor of Physics in the University of Bristol, has given us a definition of particular relevance to the present Inquiry: Science is public knowledge[188].

  1.2.  In the UK very considerable scientific research is done in our Universities and is supported by public money. The results of this research must be openly published, if not they do not meet Ziman's criterion of being science. Academic scientists have a duty to publish. Moreover, it is by their publications that they, and their institutions, are judged. It is by their publications that their "scientific reputation" is built, that they are judged for promotion through the academic ladder (although here, other factors such as teaching and administration will also be relevant), that will form the main criteria for the award of prizes and election to bodies such as the Royal Society.

  1.3.  It is, therefore, both integral to the role of scientists, and very much in our best interests, to publish our work and for our publications to be read.

  1.4.  By and large scientists publish the results of their research in scientific journals. These publications are, generally, of three types, and only the first (but by far the most important) of these is relevant to this Inquiry: primary research articles, reviews and commentary. Primary research articles are where the results of research are announced to the public. Typically, they include not only the results of research and their place or relevance in the particular field, but also (if reporting experimental results) a detailed explanation of the materials and experiments that have been done: classically this detail should be sufficient for someone else to replicate the research reported.

  1.5.  Within any given scientific subfield any practicing scientist will have an intuitive feel for where to publish any particular paper. Scientific journals vary enormously in their scope, from the whole of science—for example the weeklies Nature and Science—to the most specialized. Scientific journals also vary in their quality. There are, indeed, "objective" measures of the quality of scientific journals—eg the numbers of times papers published in them are cited in other scientific papers (measured by a journal's "Impact Factor"[189]). These are very biased and flawed measures[190], they appeal to administrators, but not to scientists. Within their field scientists have a good "feel" for where to publish any particular paper, based on their knowledge of the journals in their field, their judgment of the "importance" of their paper and of the particular audience they want to read it.

  1.6.  What makes a "good" scientific journal ? This is an extraordinarily difficult question to answer. There are some qualities all good journals have in common: they are efficient in handling publications (that is to say they process manuscripts rapidly), they publish on schedule etc. Most importantly, however, is that they are edited by people who have a good knowledge of the field and of its practitioners, for what makes a "good" journal is that it publishes "good" papers—papers of both sound technical quality but also reporting interesting and novel work. All "good" journals have two things in common: they subject the papers submitted to them to rigorous peer review, and they only publish a fraction of manuscripts submitted to them.

  1.7.  Some journals, especially the weeklies Nature and Science, triage manuscripts in their offices and only send for peer review a fraction of the manuscripts they receive (for Nature this is done in-house by the editorial staff (generally PhD level scientists), for Science this is done by sending each manuscript to two members of its external Board of Reviewing Editors). Some other journals follow a similar policy; but I would guess that the majority send all manuscripts received for peer review, typically to two or three scientists in the field. Typically, reviewers are anonymous, and they should have no conflict of interest with the authors.

2.  Who publishes the primary scientific literature?

  2.1.  The first scientific journals were established in the 17th century by the grand societies of Britain and France and very many scientific journals continue to be owned by scientific societies. In some cases a Society will be its own publisher, in others publication will be contracted to a commercial publisher. For example the journals of the Genetics Society are published by the Nature Publishing Group. It is important to note that even relatively low circulation journals may be very profitable[191] and these profits (or, the share of profit, if the journal has a commercial publisher) are an important source of income for the Society. These journals cannot make a loss, but one can argue that the motive for their publication is more altruistic than simply making money—after all these societies have the broad function of promoting their particular field of science (they support conferences, give travel grants to students etc). Many journals are published by commercial publishers, and they are published for the sole purpose of profit to their owners. In between these poles are "not-for-profit" organizations, in my own field the Cold Spring Harbor Laboratory in New York is an important scientific journal publisher.

  2.2.  In the 1950's certain commercial publishers had the foresight to see that there was easy money to be made from publishing scientific journals. Perhaps the most famous of these was Captain Robert Maxwell, who founded the Pergamon Press. Over the last decade or so there has been a massive consolidation in the STM publishing field. For example Elsevier has purchased, in the last few years, both Academic Press (Harcourt) and Pergamon Press, as well as several smaller publishers—at least one merger being referred to the Competition Commission[192]. This has resulted in the situation where much of the intellectual property of science over the last century is now owned by a very few commercial publishers.

3.  Why did the scientific community get angry?

  3.1.  About five years ago the scientific community began to realize that both it—and the public at large—was being "ripped off" by the large commercial publishers of scientific journals. The argument was as follows: Our research is funded by the public purse (by and large), our objective is to make the results of this research as widely available (internationally) as possible, we provide the primary content of these journals free of charge (indeed we may even have to pay "page charges"), we maintain the standards of the journals by reviewing papers (without payment) and, in many cases, by acting (without payment) as editors or editorial board members; yet we are obliged to sign over to the publisher our copyright (except for US Government employees), and we—or our institutions—are then charged a high cost to buy back access to "our work"; these costs are, in many cases, so high that many institutions simply cannot afford them.

  3.2.  It also became clear that the Internet offered both an alternative model for publishing and one that could have one enormous benefit to science and society. This benefit would be access to the full content of scientific papers on the Internet. This is important for two reasons: the first is that it would widen access to individual papers. The second is that it would open the full content of scientific papers to computational searching and indexing This is, simply, vital for scientific research. The volume of research, internationally, is now such that no single scientist can "keep up" with his or her field. We rely on computational search tools to discover what has been published. In medicine and biology by far the most important of these is MEDLINE, a product of the US Government's National Library of Medicine that is freely available to all[193]. MEDLINE publishes, and indexes, abstracts of the biomedical literature. Its current content is some 11 million abstracts. (As an aside, Elsevier own a competing product, (EMBASE[194]), which is only available by purchase, costs the University of Cambridge £5,700 p.a. Elsevier have argued in the US that MEDLINE should not remain free to the community, on the grounds of unfair competition.) The community realized just how much more powerful indexing of the scientific literature could be if the full content, and not just abstracts, of their papers were available for searching.

  3.3  These two general considerations, grave concern about the pricing policies of some major commercial publishers (and with their massive profits), and the restriction they make (through their ownership of copyright) to full Internet access to scientific papers, lead to the demand for an alternative model. This is Open Access publishing.

  4.  The Open Access alternative.

  4.1.  In 2001 a group of ten scientists (including myself) published a letter in Science calling for an alternative model for publishing primary scientific papers[195]. This letter called on the publishers to make their full content (of primary scientific papers) open to all without charge six months after publication. A site at the National Institutes of Health was set up to enable this (PubMed Central[196]). This site now includes the content of 144 biomedical journals, and is systematically digitizing back issues of many[197]. (By way of contrast, Elsevier's Science Direct site has 1,800 journals[198].).

  4.2.  Our letter called for a boycott of publishers who would not agree to this policy, and asked the community to sign an open letter saying that they would support this boycott. Over 30,000 scientists signed this letter. This letter had very wide publicity and engendered immense hostility from the commercial STM publishers[199]. It did, however, have an immediate impact, both practically (an increase in the number of journals available on PubMed Central) and in raising the consciousness of the community about this issue. It lead directly to the founding of new Open Access publishing, both by some innovative and creative commercial publishers (eg BioMed Central[200]) and by the Public Library of Science (PLoS[201]).

5.  What is Open Access publishing?

  5.1.  Open Access publishing should be seen as one current in a wider river, a river that includes Open Source code for computation[202] and far broader societal demand for openness in Government and public affairs. It has two principles: one is that the copyright of scientific papers remains with their authors (unless owned by their employers, eg the US Federal Government); the other is that access to the full content of scientific papers is available free of charge to all. Of course, scientific publishing must be paid for. There are several possible business models. It may be paid for by advertising, it may be paid for by charging for access to articles that are not primary research papers (eg reviews, commentaries, news articles), it may be paid for by up-front charges to authors, or by any combination of these charges.

  5.2.  The PLoS model is for publishing to be paid for by the "authors", in reality their funders. The current charge for publishing a paper in PLoS Biology is $1,500. This can be waived, if an author has convincing evidence of inability to pay (eg for an author from a developing country or without grant support). (Many journals have, in the recent past, charged authors "page charges" that often exceeded $1,500; even today, the Proceedings of the National Academy of Sciences (Washington, DC) charges its authors $70/page[203].)

  5.3.  Your Committee will have Evidence from the Public Library of Science, and from others who practice Open Access publishing. Let me just make one comment from the point of view of a practicing scientist. I have no doubt whatsoever that the model is working and is being very effective. I must declare an interest (I am on the editorial boards of three Open Access journals: PloS Biology, Genome Biology and the Journal of Biology), but these journals have rapidly established very high reputations.

6.  What impact do publishers' current policies on pricing and provision of scientific journals, particularly "big deal schemes", have on libraries and the teaching and research communities they serve?

  6.1.  There is great concern in the research community about the consequence of "big deal" schemes forced on University Libraries by the publishers. This is for two reasons. The costs are adversely impacting the provision of journals, since journals published by non-commercial publishers (eg the scientific societies) must be dropped to pay for "big deals". There are also concerns about security of access, not only to current but also to past content. Already some major US Universities have cancelled their "big deal schemes", with the effect of cutting access to these journals, both current and past, to their communities[204].

7.  What action should Government, academic institutions and publishers be taking to promote a competitive market in scientific publications?

  7.1.  There is no such market now, nor could there be, in the traditional sense. Primary scientific papers are only published once and, regardless of where they are published, the scientific community, and the public at large, need access to them. Scientific journals, if commercially published, have a monopoly. I may not like Elsevier's commercial policy. I can refuse to edit their journals or review their papers, but I cannot afford (professionally) to ignore their content or not to cite the papers they publish. The commercial publishers exploit their monopoly for monetary gain. Open Access journals are non-monopolistic, since they do not hold the copyright of their papers. Papers published in PLoS Biology carry this copyright notice:

  Copyright: 2004 Ashburner et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

  7.2.  Competition between journals exists, and will continue regardless of business model, because the journals are competing for the "best" scientific papers. At the top of the league the existing competition between, for example the weeklies Nature and Science, is very intense. I see no need for Government intervention.

8.  What are the consequences of increasing numbers of open-access journals, for example for the operation of the Research Assessment Exercise and other selection processes? Should the Government support such a trend and, if so, how?

  8.1.  I see no great impact of Open Access publishing on the RAE, or similar exercises. I presume that we will soon have a wide range of different Open Access journals, just as we now have a range of different commercial journals. These will vary in standard and, just as now, assessments of research may (although mistakenly in my view) take such measures as the Impact Factors of these journals into account. The outputs of scientific research should be judged on their merits, and the best way to secure this is by peer review.

  8.2.  The Government should actively encourage both the development of Open Access journals and encourage scientists to publish in them.

  8.3.  Most academic science in the UK is Government funded, and it is obviously in the Government's general interest for the fruits of such research to be as widely available to the community as possible. This cannot be achieved if there is a very high price barrier to access—as is the case with, for example, Elsevier journals. Not only is this a bar to access by scientists (surmountable, perhaps, if they work at a relatively well-funded Institute), but it is an insurmountable barrier to access by science journalists, members of the public or, perhaps, to Members of Parliament. Papers published in Open Access journals are there to be enjoyed, without cost, by all.

  8.4.  Several private and public funders of science have publicly declared their support for Open Access journals, and encourage their fundees to publish in them. Such support was first given by the Howard Hughes Medical Institute in the USA (a very major funder of biomedical research), followed by the Wellcome Trust in the UK and some overseas bodies, including the French Centre National de la Recherche (CNRS) and the German Max Planck Society[205].

  8.5.   The Government should not allow scientists, publishing primary research papers on work funded from a public source, to assign an exclusive copyright of such works to a commercial publisher. (This is the case for scientists working for US Federal Government agencies, such as the National Institutes of Health.)

  8.6.  The Government should encourage the Research Councils to follow the lead of other funders in encouraging Open Access publication of primary papers. They should encourage the Research Councils to state explicitly that research funds can be used to pay for such publication should it be needed. (The costs will be relatively minor, 1 to 2% of a typical research grant over its three-year life; if Open Access publishing becomes the norm, then there may need to be a re-distribution of resources within Universities.)

  8.7.  By encouraging Open Access publishing the Government will be serving other ends. This includes widening the fruits of scientific research to the community at large. Any person with Internet access at home, school, work or their local library, will have as easy access to this work as a Cambridge Professor. Science is international. Much good science is done in developing countries. At most Universities and research institutes in these countries access to the bulk of the scientific literature is denied, because of cost (although the WHO and others have recently been making some steps to improve their access). Access to the Internet is now longer such a serious problem in Universities and research institutes in developing countries, but access to journals, either in print or via subscription to Internet editions is. By encouraging Open Access publishing the Government will be encouraging scientific and technical development in the less fortunate nations of the world.

9.  How effectively are the Legal Deposit Libraries making available non-print scientific publications to the research community, and what steps should they be taking in this respect?

  9.1.  I am not well qualified to comment on this, but there are concerns in the scientific community. If I want to read a paper published in the Philosophical Transactions of the Royal Society in 1795 I can be confident that I will find a paper copy on a shelf in a library. Paper is an extraordinarily durable medium. There is the concern that electronic media will not be as durable.

  9.2.  A second concern is that, as back issues of journals are digitized, paper copies now held by libraries will be discarded. This could be a disaster, especially if the only access to the digitized back issues is through a commercial publisher and relies on a current "big deal scheme". It is very important that these digitized archives are held in the public domain by, for example, the Legal Deposit Libraries (for this reason we must applaud Elsevier's agreement with the Royal Dutch Library, which will act as the deposit of last resort for their journals, should they go out of business[206]).

  9.3.  The UK academic community is well served by Inter Library Loan Services from Boston Spar[207]. They traditionally provide copies of papers from print material. It is important that they have the resources to provide copies of material published electronically, even when the original source is no longer available. Pressure from the commercial publishers has been directed at the British Library against document supply[208].

  9.4.  The Government should encourage resistance to pressure from commercial publishers to curtail document supply by the Legal Deposit Libraries.

10.  What impact will trends in academic journal publishing have on the risks of scientific fraud and malpractice?

  10.1.  I am not sure that Open Access publishing will have a major impact on the risks of scientific fraud or malpractice. However, if there is any impact it will be positive, for this reason. Scientific fraud and malpractice will rarely be detected if a scientific paper remains unread. Open Access widens access[209]. By widening access Open Access publications widen the probability of work being closely examined.

February 2004

188   J.M. Ziman. 1968. Public Knowledge. An essay concerning the social dimension of science. Cambridge. Back

189  Back

190   D. Adam. 2002. Nature 415: 726-729. Back

191   For example, in 2001 the Genetical Society earned £169K (net) from its journals. The Society's total income was £232K. ( Back


193  Back

194  Back

195   R.J. Roberts et al. 2001. Building a "Genbank" of the published literature. Science 291: 2318-2319. Back

196  Back

197  Back

198  Back

199   eg  Back

200  Back

201  Back

202  Back

203  Back

204   J. Knight. 2003. Cornell axes Elsevier journals as prices rise. Nature 426: 217. See also and  Back

205  Back

206 Back

207 Back

208   Annual Report 2000/2001. International Association of Scientific, Technical and Medical Publishers. Amsterdam. Back

209   S. Lawrence, 2001. Free online availability substantially increases a paper's impact. Nature 411: 521. Back

previous page contents next page

House of Commons home page Parliament home page House of Lords home page search page enquiries index

© Parliamentary copyright 2004
Prepared 20 July 2004