The field of social sciences and humanities (SSH) is the poorer cousin of the sharp-suited, well-funded science, technology and medicine (STM) sector. Yet it faces a similar crisis in publishing, with rising costs for books and journals. In STM, this crisis has been one of the drivers for open access, but this has not been the case in SSH so far.

The dearth of funding in the SSH sector has been one the main reasons it has lagged behind in getting research online and embracing open access. There is not a lot of money around to finance author-pays models of open access (OA), although there has also been an absence of drive on the part of researchers towards open access, backed by a cultural resistance in some disciplines to any sharing of research at all.

But recently, there has been a dawning of understanding among researchers that OA can bring benefits much broader than simple speed and ease of access to research.

At the same time, publishers facing demands for open access have started to make their concerns known, citing the long tail of access to research in this sector that would threaten their business model. Compared with the STM sector, there is a much higher proportion of journal articles accessed for the first time over a year after publication in SSH.

The SSH sector spans a huge number of disciplines with widely differing research models. A one-size-fits-all approach to OA is not an option. Economics, for example, relies heavily on reports and working papers. It has a culture of exchange of working papers, which lends itself to an OA model. Linguistics researchers are also interested in the OA debate, and psychologists who receive grants from the likes of the Wellcome Trust or the NIH in the US must deposit their papers.

The launch of the Open Humanities Press (OHP), an international OA publishing collective in critical and cultural theory, at the end of April is one sign of the growing realisation of the need for OA in humanities.

Gary Hall, one of the founders of OHP, says: “When other online movements and practices, such as Creative Commons, free software, open source and peer-to-peer file sharing, have been regarded as providing models for new regimes of culture, new kinds of networked institutions, even for the future organisation of society, why has the open access movement had comparatively little impact on humanities to date?”
poor penetration

Hall cites the difficulty of funding author-pays OA and the greater familiarity of the STM sector with electronic resources as possible explanations for the lack of penetration. He says: “There is greater emphasis on the dissemination of research findings as fast as possible in the sciences than in the humanities generally. After all, if you find a cure for cancer, there is a sense in which it is important to let people know about it, and have them test and verify your research as soon as possible – a sense that is not quite present with regard to a philosophical reading of Deleuze, no matter how interesting or original it may be.”

He adds that there are practical, historical reasons why SSH hasn’t embraced OA. “Sigi Jottkandt has pointed out bibliographic conventions that are slow to change. Many still only recognise the publisher version of an article as the official publication of record and do not yet accept pre- or postprints as acceptable citation alternatives.”

He believes a negative perception of the digital medium persists and needs to be addressed before OA can really take hold in the humanities.

“After looking at the various efforts under way, we concluded that an editorially driven international press, focused on building respect through its brand, is what is required to tackle the ‘credibility’ problem,” he says.

“We started Open Humanities Press to do just that. Once we have the core community and support structures in place, we see the next phase as a matter of achieving the critical mass necessary for creating a wider awareness of OHP and, by extens ion, open access in the humanities.”

The EU has put its weight behind moves to hasten OA in SSH through the so-called Action 32 of the STM-based COST (Co-operation in the field of Scientific and Technical Research) European programme.

Action 32 aims to create a digital infrastructure for collaborative humanities research on the web and to establish and foster the growth of scholarly communities that will provide feedback to the IT developers on the needs and expectations of humanities researchers and will serve as a core group of early adopters.

Expectation shift
Michael Jubb, director of the Research Information Network (RIN), points out that while it is true that there is less pressure and hurry to have the very latest research on your desk in SSH than in STM, things are changing. “Having said that the pace of research is different and the expectations very different, both these things are changing,” he says. “There is a growing expectation about the availability of source material in digital form.

“One of the most interesting things that has been happening is that researchers increasingly expect to find the source text and data. They are certainly expecting catalogues and other sorts of finding aids.”

Teaching and learning OA to research is of particular importance in making available to final-year undergraduates the kinds of source material to which they need access to do advanced-level work.

“The distinction between research sources and teaching sources is very fluid in humanities and social sciences, certainly at final year and master’s level,” Jubb says.

Jonathan Gray of the Open Knowledge Foundation (OKF) is part of the movement to get researchers and IT developers together at ground level to develop the systems needed for OA. “We want to encourage people to publish their own scholarship. We don’t get involved in policy-level debate. We really feel that technical developers and researchers should be working much more closely together. We are working with groups of academics to use software to build resources that are useful to them in terms of teaching, and we are forging links with researchers.

“We are analysing the source material being made available digitally and raising the issues of lack of testing and the practical restrictions in the ways in which source material can be used and analysed and data extracted.”

Gray believes the first step to OA in the SSH sector is to provide better access to research that is already in the public domain. “There is a huge corpus of material in the public domain that is not very visible at the moment,” he says. “One of the visions behind the OKF is a wealth of material that anyone is free to access, reuse and redistribute. While we don’t think that all knowledge should be open all the time – there are always pros and cons – as an organisation we focus on collating, archiving and increasing the visibility of material that is open already.”

As part of its public domain works project, OKF has set out to create an algorithm to determine whether a certain work is in the public domain when given details such as date of publication or date of death of author.

Much of the debate on OA in the SSH sector tends to be based on the assumption that, as with STM, researchers want access to journal articles. In fact, the majority of scholarly publishing in these sectors, particularly in humanities, comes in the form of book publishing. In the 2001 Research Assessment Exercise for funding, only 37% of the items submitted in the arts and humanities were in the form of journal articles, compared with 96% in the sciences.

The OHP recognises this. “When launching OHP, we took the decision to focus on the situation with academic journals in the humanities because it is acad emics who in the main run those,” says Hall. “That said, we are hoping in the not too distant future to move toward a situation where OHP can publish academic books too.”

Unfamiliarity
Many researchers in this sector simply do not know how to go about making their research open access. A survey by RIN showed that only 14% of arts and humanities researchers (compared with 30% in the physical sciences and 36% in the life sciences) think they are familiar with the options for making their research outputs open access.

Jubb believes they are unlikely to take their lead from publisher efforts. Publishers really don’t have much of a handle on the way forward, he says. Humanities and social sciences are very much small beer for publishers compared with the STM sector.

T&F is one publisher that has dipped a toe in the water with open access in SSH. David Green, T&F’s global journals publishing director, says: “We have rolled out iOpenAccess, a hybrid OA pilot, to psychology journals, and have some education and other journals asking, normally where there is medical funding.”

However, Green is conscious of the experiences of the OUP with OA: “The OUP found very little take-up in SSH. There was also no interest in OA in our researcher focus groups.”

He believes it is too early to tell what the true impact of OA would be on the SSH sector.

“One of the big American medical journals found a one-time drop of around 5-10% of subscriptions when it made its back archive free to access after a couple of years. We saw something similar, if less marked, with two of our journals when they introduced their 12- and 24-month embargo postprint policies.

Renewals since have been good. This seems a common experience: a small loss in the first year after introducing some form of OA, followed by a large increase in usage.”

So would it hold more widely in SSH? “Hard to say, but we would remain concerned that SSH material has a much longer half-life and much longer usage tail than STM.

“There have been requests for the odd paper or issue to be made OA, and we’re dealing on a case by case basis with those queries, but most baulk pretty quickly when we quote them the going rate of $3,000-plus per paper (times the number of papers for an entire issue to be made OA).

“Strangely, these seem to be coming from disciplines like education, which are traditionally very paper-oriented and underfunded, hence the quick loss of interest and/or intense negotiating when they hear the OA rate. I don’t think that this necessarily indicates a growing interest in OA in those communities at all, rather an interest in wider dissemination of these particular papers/issues and a hope that we’ll just make them free without any cost to the authors, editors, etc.”

Sage business development manager Dave Ross says Sage is interested in including any title that has funding. “For the author-pays models to work there has to be money for the author to pay, so anywhere where there are reasonable levels of direct funding, open access could work at the moment,” he says.

“Biomedicine is the proven model but going forward we are very keen to explore other areas: engineering, then beyond that any social sciences disciplines where there is a high level of funding – for example, psychology.”

Outside the major publishers, most of the OA projects in this area are cottage industries. OHP, for example, is a volunteer effort by OA journal editors and OA advocates which has so far operated without any cash income at all.

The idea of institutional support for author payment fees has been bandied about for some time, with librarians mounting a robust challenge to the assumption that the money would come from subscriptions budgets.

Circuitous arrangements
While institutional or centralised funding could be a way forward, Ross points out that the diversity of finance arrangements would get in the way of developing a model. “No library or university finance department is administered in the same way as another, and library and research budgets are allocated by circuitous routes,” he says.

Jubb is part of a concerted effort to guide institutions towards centralised arrangements to pay publishing fees. He says: “I see no sign at all that research councils have much enthusiasm for meeting the costs of publishing. I am chairing a meeting on payment of publication fees and the practicalities of how institutions might take a more strategic approach to payment for publication.

“Pay-to-publish is now part of the scene and could be stifled unless there is a more strategic approach from institutions. We want to provide guidelines as to how we might set about doing it rather than leaving individual researchers to find a bit of money to publish.”

As the SSH sector embraces open access, visibility of material becomes key. Quality is important too; not just scholarly quality, but quality of usability. As these goals start to be achieved, open access could begin to drive changes in research patterns in these sectors.

In the STM sector, speed of access to findings is essential, whereas in the SSH sector, quality of access will allow broad pictures to be built of concurrent events, for example, or the geographical path taken by the development of ideas.

The development of tools to take the legwork out of this kind of research will leave the researcher free to think and develop ideas in a way that has not been possible before.

RIOJA’s quality assurance
The idea of providing a quality assurance layer to open access articles deposited in institutional repositories may be of especial interest to the fragmented and cash-strapped social sciences and humanities communities.

The Rioja (Repository Interface for Overlaid Journal Archives) project being run at University College London is looking into the overlay of quality assurance onto papers deposited to and stored in e-print repositories, based on the established physics repository ArXiv in the first instance.

The pilot project is building a toolkit to enable interoperability between journal software and public repositories supporting the overlay of quality certification.

Project manager Martin Moyle, of UCL Library Services, says: “The toolkit is deliberately discipline-neutral. In that sense, where there’s a will, there’s a way. Lots of money is going into repositories, which are increasingly a part of academic life across many disciplines.

“Overlay journals are just one way of making repositories work a bit harder for us and adding value to all those repository records.

“In disciplines like humanities and social sciences, which tend not to have the same levels of funding as the sciences, a repository overlay model could well keep down the costs of journal management. The costs to a journal team of content hosting and archiving, for instance, could be reduced significantly through an overlay approach.”

However he adds: “We felt we should test the premise that the overlay model might be significantly cheaper than traditional models. In a very basic sense, given a serviceable repository, that is obviously true; but much then depends on what features, service levels, functions and presentation standards the target community will settle for, which will obviously affect the costings, not forgetting the apparatus for certification, the volume of submissions, rejections and acceptances, supplementary archival arrangements, if required, and so on.”