On 28 June, the Research Councils UK (RCUK) posted a Draft Policy on Open Access at the RCUK web site for public comment. The comment period will end on Wednesday, 31 August 2005. The RCUK Proposal requires the provision of ‘Open Access’ to all the research papers it funds via freely available versions online. If adopted, the Proposal would make the UK a world leader in the provision of Open Access to publicly funded research. The press release summarizes the RCUK position. RCUK spokesman, Professor Ian Diamond said that Councils have already widely consulted the research community:
“We’ve held workshops, given evidence at Select Committees, met with the publishers through a DTI working group and written out to all UK Vice Chancellors to share our views as they emerge on this issue and hear what others are saying.”
BACKGROUND: Open Access (OA) is a relatively complex subject, and this post has become rather long. I want to provide a basic introduction to OA first.
Open Access occurs when copies of full-text journal articles (and other research information such as data, conference proceedings and theses) are made freely available online. Anybody anywhere with an internet connection can immediately access OA research data and OA scholarly journal articles free of charge. The most swiftly effective method of achieving OA is by placing research articles in digital web-based ‘repositories’. Ideally, each research institution (eg. a university) would maintain its own ‘institutional repository’ (IR) of research output into which its researchers would upload a copy of their peer-reviewed journal-published research articles.
The IR method of achieving OA has many advantages - a repository becomes a ready-made ’showcase’ of a university’s research output and thus there is a natural vested interest in keeping the repository maintained, and to fill it up with research papers. For researches, it offers a handy way to maintain a collection of all their work and easy way to share it with students without encountering access problems. Two other OA methods are in use. (1) ‘Central’ (often subject-based) repositories used by researchers from any institution worldwide may not have such a clear ownership responsibility or incentive. (2) Scholarly peer-reviewed Open Access ‘Journals’ are unlikely to become a widely based possibility simply because relatively few OA Journals exist. The Directory of Open Access Journals (DOAJ) currently lists around 1,650 - which represent less than 7% of the 24,000 scholarly journals worldwide. The number of OA Journals, though gradually increasing, is unlikely to rise to reach ‘critical mass’ in the near future. A personal observation here: OA ‘Journals’ - web-based e-journals - seem to be just more highly visible versions of a small central/subject-based OAI Repository and most of them deposit copies of their articles in OAI repositories as well.
Ironically, pace RCUK, it is already possible for most research to be made Open Access via copies in repositories as outlined above! According to the SHERPA/RoMEO database, which lists publishers’ policies on Open Access to copies of their articles, 91% of scholarly journals already allow copies of their articles to be placed in a web-based open access repository! This is often referred to in OA parlance as ‘archiving’ or ‘author self-archiving’. But - and here’s the sticking point - it is an entirely voluntary action on the part of the researchers who may perceive self-archiving as a lot of extra hassle. Therefore, despite the publishers’ permissions, only 15% to 20% of research papers are currently being archived worldwide in a repository. Admittedly, this could partly be because most institutions do not yet have a repository to archive in (or have not made known that they have one). Without some kind of compulsion, archiving of openly accessible copies of most research is not going to happen soon - or perhaps ever. The RCUK Proposal introduces this seemingly necessary element of compulsion: the RCUK would require the timely archiving of each research paper as a condition of its funding. [Also see Alma Swan - Faculty awareness. Key Perspectives Ltd, a report on what current self-archiving activity there is. Dr Alma Swan is co-founder and director of Key Perspectives, a UK-based scholarly publishing consultancy]
Despite the low number of Institutional Repositories (of which more detail below), amongst the top 20 research grant funded universities nearly all now have a repository. Our own University of Cambridge has its DSpace@Cambridge repository which is in Pilot Study mode with a limited set of users until end 2005 so it is hardly, if at all, publicized. However, DSpace@Cambridge states on its site that it now holds a total of 30,000 records.
To enable web search engines - and therefore web users - to easily find and index the research articles in the repositories, the articles must comply with the Open Archives Initiative’s Protocol for Metadata Harvesting (OAI-PMH) - which basically means that each article in an ‘OAI-compliant’ repository must be described using an agreed format of ‘metadata’ (the items of data describing the article such as author, title, publishing date, web address etc.). There also exist specialized ‘OAI Harvesters’ such as OAIster which can easily ‘harvest’ the standardized metadata to create a cross-repository bibliographic database: OAIster now has over 5m records from almost 500 institutions in its database (although, unfortunately, it seems that many of OAIster’s records do not provide access to the full-text article).
The metadata format used by OAI-compliant repositories is called ‘Dublin Core‘ (that’s Dublin, Ohio, USA, where the original OCLC/NCSA Metadata Workshop was held by ‘geeks, freaks, and people with sensible shoes’ in March 1995). Dublin Core specifies 15 metadata elements, all optional and repeatable, which can be applied to each article in a digital repository (the elements include author, title, publishing date, web address, key words, type of resource, etc.). Dublin Core was adopted as an International Standard (ISO 15836) in October 2003.
For a more detailed introduction to Open Access, see Peter Suber’s excellent Open Access Overview: Focusing on open access to peer-reviewed research articles and their preprints and the Joint Information Systems Committee’s (JISC) Questions and answers about opening up access to research results.
[Back to the RCUK Proposal] The Proposal requires that a copy of every research paper arising from RCUK-funded work be deposited in Openly Accessible repositories at the earliest opportunity. Note that the Proposal specifically applies to articles that have been published in peer-reviewed journals or presented at conferences. The Open Access version of an article (which would normally be a simple ‘text and tables’ version, not the ‘full PDF’ format as published) should be regarded as a supplemental version for those who cannot, for whatever reason, gain access to the published journal - after all, no library can afford all 24,000 or so peer-reviewed journals. However, the original journal-published article would always remain the definitive version.
The RCUK Proposal, if adopted, would apply to new research grants awarded from 1 October 2005. But, given the long-term nature of most research, the impact of the policy would not be felt immediately. Therefore the Proposal also encourages those currently engaged in RCUK-funded research to consider offering an OA version. The Proposal would cover a substantial proportion of all UK research: in 2003/4 the RCUK provided around half of all UK public research funding [source: Leslie Carr of Southampton University].
The RCUK Proposal has largely been built on a previous report published almost one year ago (20 July, 2004) by a House of Commons Science and Technology Committee: Scientific Publications: Free for All? [A PDF version of the report is here]. This was a deeply researched report based on six months of inquiry, four rounds of public hearings, and hundreds of written comments. It included the following recommendation [my emphasis]:
“44. Academic authors currently lack sufficient motivation to self-archive in institutional repositories. We recommend that the Research Councils and other Government funders mandate their funded researchers to deposit a copy of all their articles in their institution’s repository within one month of publication or a reasonable period to be agreed following publication, as a condition of their research grant. An exception would need to be made for research findings that are deemed to be commercially sensitive. (Paragraph 117)”
However, a Government response (8 November 2004) rejected the House of Commons Science and Technology Committee recommendations [A PDF version of the response is here]. For instance this quote about mandating the depositing of copies of research in repositories:
“(44) Institutional and thematic repositories can play a significant role in the dissemination of research outputs. However the Government has no present intention to mandate Research Council funded researchers to deposit a copy of their published material in institutional repositories.”
It should be noted that the House of Commons Science and Technology Committee, unlike the RCUK, concentrated solely on Scientific, Technical and Medical (STM) research - explicitly excluding from its considerations journals and other publications in the social sciences, arts and humanities (report, paragraph 6). However, the bulk of arts and humanities research is not directly funded by the UK Research Councils, but often by institutions themselves [source: Michael Fraser of Computing Services, Oxford University], therefore the RCUK Proposal would, unwittingly, also be biased towards STM research.
The Science and Technology Committee Report seemed to concentrate more on the so-called ‘author pays’ method of open access where authors (or their funders, in practice) pay the cost of publishing in Open Access journals such as those published by BioMed Central (BMC) or Public Library of Science (PLoS). The RCUK Proposal also mentions helping to pay for OA Journal costs where relevant - such as when the author’s institution does not have an OA Repository. But the RCUK could more usefully also offer to help fund the setting up of repositories for institutions which lack them. In fact, the RCUK Proposal even says:
“(14d) This position statement carries no obligation to set up a repository where none exists at present; and the requirement to deposit will therefore not apply to authors who do not have reasonable access either to an institutional repository or to a subject-based e-print repository […]”
This introduces a reasonably sized loop-hole in the Proposal as many publishing researchers don’t have Open Access repositories in their institutions or disciplines. The Institution Archives Registry maintained by Tim Brody at Southampton University is the nearest thing to a definitive list of all OAI-PMH compliant repositories worldwide (note it depends on institutions voluntarily registering their repositories). Currently the IAR lists 55 repositories registered by UK institutions. Analysis of these 55 reveals that 34 of them are of the ‘institutional or departmental repositories’ type we are concerned with. Of the remainder, 8 are cross-institutional, 1 is for e-theses only, 2 are databases, 6 are e-journals, and 4 are ‘other’ [analysis provided by Prof Stevan Harnad, University of Southampton].
So, there are 34 registered ‘institutional or departmental’ OAI compliant repositories in the UK. But there are around 140 to 160 (depending on your source and definition) universities in the UK - therefore the 34 registered repositories represent, at most, only one quarter of all UK universities, leaving a majority of them without an OA repository. However, if self-archiving is made mandatory or suddenly takes off, the number of IRs could swiftly increase - and, indeed, the RCUK Proposal asserts this probability.
A solution for UK universities without repositories is to ask the RCUK, or JISC, to help fund one. Another solution, which could act as a stop-gap, is the proposed launch of an OAI-compliant “universal repository” by Brewster Kahle [co-founder] of the Internet Archive (IA). The Universal Repository would accept e-prints from any scholar in any discipline with no repository of their own. Not only would it host new content for scholars with no other place to deposit their work, but it will offer to preserve all the other OAI-compliant repositories. The Internet Archive has a proven commitment to Open Access and long-term preservation: it has been archiving nearly everything on the open web since 1996 (and is well worth exploring).
The RCUK Proposal comes after three previous statements on UK Open Access: the Wellcome Trust position statement in support of open and unrestricted access to published research (October 2003; revised June 2005), The Scottish Declaration on Open Access (October 2004) by a group of 20 Scottish institutions, and the Russell Group Statement on Scholarly Communication and Publishing (June 2005), the Russell Group being an association of 19 major UK research-intensive universities. The Proposal from the RCUK should act as a further motivator for other institutions to develop equivalent policies which take into account the full range of subjects and research outputs. [sourced from: Michael Fraser of Computing Services, Oxford University].
There are two excellent commentaries on the RCUK Proposal: by Prof Stevan Harnad in the American Scientist Open Access Forum (which he moderates) (28 June) and by Prof Peter Suber in his Open Access Newsletter (2 July). Incidentally, this post owes much to these two high-profile proponents of the Open Access movement, from whose writings I have picked up most of my knowledge about OA.
Two Education Guardian writers have written about the RCUK Proposal: Richard Wray Funding aid for open access (29 June) and Donald MacLeod Research councils back free online access (29 June).
The Association of Learned and Professional Society Publishers (ALPSP) quickly produced A response to the RCUK Proposal (30 June although, strangely, it is dated 19 April) which is largely against the Proposal on the grounds of it leading to cancellation of journal subscriptions. There is no evidence for this so far. For example, in the physics area the use of openly available archived ‘preprints’ and ‘postprints’ (pre- and post-refereeing) has been standard practice for over 10 years (for example, in the seminal arXiv subject-based repository originally created in 1991). During this considerable time, physics journals do not seem to have suffered any loss of sales. The ALPSP response has been comprehensively and robustly critiqued by Professor Stevan Harnad, a passionate proponent of OAI Institutional Repository archiving as the best method of quickly achieving Open Access to current research.
Another personal observation here: Ironically, the perceived need to mandate Open Access to articles may now lead to journal publishers closing ranks and placing obstacles in the way of what, up until now, 91% of them have voluntarily allowed! This has already happened in the U.S. with the National Institutes of Health (NIH) proposal to archive all research in their PubMed Central Repository (PMC) being reduced to having a 12-month embargo on becoming Open Access! True OA should be OA immediately on publication.
Graham Taylor, director of educational, academic and professional publishing at The Publishers Association, questions the RCUK Proposal in a Guardian article: Don’t tell us where to publish (1 July). He is not happy about the RCUK Proposal and his article, therefore, has not escaped the finely tuned critical faculties of Prof Stevan Harnad who really is in top form these days: Critique of Graham Taylor’s critique of the RCUK policy proposal (5 July).
BioMed Central (BMC), an Open Access Journal publisher, welcomes the RCUK Proposal: BioMed Central applauds Research Councils UK draft statement on Open Access (5 July) - but, naturally, they emphasize the point about funding the costs of publishing in OA Journals (the ‘author pays’ route to Open Access) whilst the academically preferred route to Open Access seems to be via self-archiving of articles in Institutional Repositories.
Back links to previous relevant postings:
Open Access Journal launch and Impact Factors (June 2005)
Open Access round-up (June 2005)
e-Journals@Cambridge - improved online journal access (May 2005)
DSpace@Cambridge and open access to research (November 2004)
DOAJ introduces article level search (June 2004)
For reference, Research Councils UK (RCUK) is a partnership between the UK’s eight Research Councils. Through RCUK, the Councils work together to champion the research, training and innovation they support. The Councils are independent non-departmental public bodies, funded by the Science Budget through the Office of Science and Technology. Note that the new RCUK draft OA policy was issued in “partnership” with the British Library (paragraph 4), and counts on the British Library to play a role in the long-term preservation of OA research literature (paragraph 31).
The eight UK Research Councils are:
- Arts & Humanities Research Council (AHRC)
- Biotechnology & Biological Sciences Research Council (BBSRC)
- Council for the Central Laboratory of the Research Councils (CCLRC)
- Economic & Social Research Council (ESRC)
- Engineering & Physical Sciences Research Council (EPSRC)
- Medical Research Council (MRC)
- Natural Environment Research Council (NERC)
- Particle Physics & Astronomy Research Council (PPARC)