University Deans are required to do many things for their university, including taking some responsibility for the research output of their Faculty. Each year, capturing all forms of research deliverables – journal papers, technical reports, conference and workshop proceedings, presentations and Doctorate and Masters theses – is a necessary and important chore. This is especially important in the UK – where the research funds allocated to each department by the Government are explicitly linked to the quality of its research over a four or five year period.
First as Chair of the Electronics and Computer Science Department, and then as Dean of Engineering at the University of Southampton, I was responsible for two of these ‘Research Assessment’ cycles in the UK. It was during the preparation of these research returns that I encountered an interesting problem: the University library could no longer afford to subscribe to all the journals in which our 200 engineering faculty members – plus a similar number of postdocs and graduate students – chose to publish. This meant that just assembling the published copies of all the publications of all research staff and students became a much less straightforward exercise. The reason for this problem is well-known to librarians – it is the so-called ‘serials crisis’. This crisis is dramatically illustrated below in a graph that shows the relative growth of serial expenditures at ARL Libraries versus the consumer price index over the past twenty-five years.
These are typical expenditure curves for all university libraries – and the University of Southampton was no exception. It was for this reason that the University Library sends out a questionnaire each year asking staff which journals they would least mind cancelling! Yet the serials crisis is a curious sort of crisis in that most research staff are simply unaware of any problem. They feel free to publish in whatever journal is most appropriate for their research and see no reason to restrict their choice to the journals that the University can afford to subscribe to.
The Research Assessment exercise in the UK is intended to measure ‘research impact’ and this is judged in a number of ways. One form of research impact that can easily be measured is the number of citations by other researchers to each paper. In order to garner citations, a research paper needs to be accessible and read by other researchers. Not all researchers – and certainly not the general public whose taxes have usually helped fund the research – have access to all research journals. Physicists have solved this accessibility problem by setting up arXiv – a repository for un-refereed, pre-publication ePrints. The US National Library of Medicine has solved the accessibility problem in a different fashion. The full text of all research papers produced from research funded by the National Institutes of Health are required to be deposited in the PubMedCentral (PMC) repository after publication in a journal, usually after some ‘reasonable’ embargo period from 6 to 12 months. Similar open access policies have now been adopted by other funders of biomedical research such as the Wellcome Trust and the Bill and Melinda Gates Foundation.
The repositories PMC and arXiv are examples of subject-specific, centralized research repositories. However, it is my firm belief that each research university needs to establish and maintain its own open access ‘institutional repository’ covering all the fields of research pursued by the university. At Southampton, in the Electronics and Computer Science Department, with colleagues Les Carr, Wendy Hall and Stevan Harnad, we established a Departmental Repository to capture the full text versions of all the research output of the Department to assist us in monitoring and assessing our research impact. A graduate student in the Department, Rob Tansley, worked with Les Carr and Stevan Harnad to develop, in 2000, the EPrints open source repository software. Robert went on to work for Hewlett-Packard Laboratories in the US and wrote the DSpace Repository software in collaboration with MIT. The EPrints and DSpace repository software are now used by many hundreds of universities around the world. For a list of repositories and software see: http://roar.eprints.org/
As Dean of Engineering, I tried to use the example of the EPrints repository in Electronics and Computer Science as a model for the entire Engineering Faculty. By the time I left Southampton, this had only partially been implemented, but I was enormously pleased to see that by 2006 the University had mandated that all research papers from all departments must be deposited in the ‘ePrints Soton’ repository. In 2008, this was extended to include PhD and MPhil theses. For more details of Southampton’s research repository, well managed by the University Library, see: http://www.southampton.ac.uk/library/research/eprints/
There is much more that can be said about this ‘Green’ route to Open Access via deposit of full text of research papers in Institutional Repositories. For a balanced account, I recommend Peter Suber’s recent book on ‘Open Access’ published by MIT Press, to be available under Open Access 12 months after publication. Peter describes the different varieties of Open Access – such as green/gold, gratis/libre – and also issues of assigning ‘permission to publish’ to publishers versus assigning copyright (https://mitpress.mit.edu/books/open-access). In addition, the Open Archive Initiative supports two community-supported repository standards: OAI-PMH for metadata and OAI-ORE for aggregating resources from different sites into compound digital objects (http://www.openarchives.org/). Also relevant is the Confederation of Open Access Repositories or COAR whose website states:
COAR, the Confederation of Open Access Repositories, is a young, fast growing association of repository initiatives launched in October 2009, uniting and representing 90 institutions worldwide (Europe, Latin America, Asia, and North America). Its mission is to enhance greater visibility and application of research outputs through global networks of Open Access digital repositories.
Why is all this important? It is important because the present scholarly communication model is no longer viable. While many journal publishers perform a valuable service in arranging peer review and in publishing high quality paper and online journals, the unfortunate truth is that universities can no longer afford the costs of the publishers’ present offerings. For example, it was not possible for me as Dean to establish a new research area in the Faculty and have the library purchase the relevant new journals. In such an unsustainable situation, it is obvious that we need to arrive at a more affordable scholarly publishing model. However, instead of just waiting for such a model to magically emerge, university librarians need to be proactive and take up their key role as the guardians of the intellectual output of their university researchers. It is the university library that has both the resources and the expertise to maintain the university’s institutional research repository. And this is not just an academic exercise. Managing the university’s research repository will surely become a major part of the university’s ‘reputation management’ strategy. Studies of arXiv have shown there to be a significant citation advantage for papers first posted in arXiv, and subsequently published in journals, compared to papers just published in journals (arXiv:0906.5418). Similarly, it is likely that versions of research papers that are made freely available through an institutional repository will also acquire a citation advantage – although this conclusion is currently controversial. Nevertheless, like it or not, universities will increasingly be evaluated and ranked on the published information they make available on the Web. For example, the Webometrics Ranking of World Universities takes account of the ‘visibility and impact’ of web publications and includes both an ‘openness’ and an ‘excellence’ measure for research repositories and citations (http://www.webometrics.info/). I am pleased to see that Southampton features in 32nd place in Europe and 119th in their World rankings J
To be continued …
Originally posted in 2013