In most of the discussions about e-publications and open access (OA) in recent years, the focus of attention has tended to be on the interests and needs of researchers in the sciences, and of the libraries that seek to serve them. Significantly less attention has been paid to the needs and interests of researchers in the arts and humanities; and indeed e-publication and open access initiatives, and general awareness of the key issues and debates, are much less advanced in the arts and humanities than in the sciences.
This article reviews some of the discussions that have taken place, and the evidence that has been gathered, about e-publishing and open access and their impact and implications for researchers in the arts and humanities in the UK. Researchers in these disciplines do not, of course, constitute a homogeneous whole, and they vary as to how they conduct their research and publish their findings. Across a broad front, however, there are some significant differences between their interests and needs and those of their colleagues in the sciences.
We focus in this paper on the activities of the UK's Arts and Humanities Research Council (AHRC)  and the Research Information Network (RIN) , using them to highlight some of the specific issues that have arisen in discussions in the arts and humanities. But our approach is also a broad one, covering different kinds of publications as well as data and other information sources that may not be formally published. While the open access agenda has in general so far been primarily concerned with scientific journals, the relative importance of journal literature as an information source is generally less in the arts and humanities than in other domains; monographs and collections of essays in book format remain important as do prestigious outlets and sources of information in many areas, while more and more data resources are being published online. We therefore include a number of other kinds of online, open-access resource in our discussion, and discuss them alongside comparable subscription-based e-publications.
The AHRC was established in 2005 and is the principal funder of research in the arts and humanities in the UK. In 2002-4 its precursor body, the Arts and Humanities Research Board (AHRB) held a series of meetings, chaired originally by David Eastwood as its Chief Executive and then by his deputy, Michael Jubb, of an E-publishing Round Table. This brought together a group of librarians, publishers and academics, with the aim of establishing a dialogue about how best to develop and exploit new models and services for publishing research results and findings.
The Round Table has not met since November 2004; but it influenced the development of the AHRC's ICT in Arts and Humanities Research Programme, which has taken forward some of the issues raised by the Round Table, as we shall see. It also helped to spawn a larger-scale Research Strategy Seminar on E-publishing in the Arts and Humanities, held in May 2005, again bringing together librarians, publishers and academics, and with representation from funding bodies. The AHRB and AHRC also participated in the lengthy discussions across all of the UK Research Councils (RCUK) in developing a position statement on access to research outputs. Following the publication of that statement in June 2006, however, it was not until September 2007 that the AHRC followed other Research Councils in producing a policy statement of its own.
The RIN was also established in 2005. Its role is to gather and analyse evidence about the key developments in the research information landscape across all disciplines; and to provide advice and guidance on the strategic development of policies and services in the interests of UK researchers. It has sought to establish dialogue between the various players in the scholarly communications process, and has established a Research Communications Group for that purpose, including representatives of research funders, librarians and information specialists, researchers, and publishers.
Through the work of the Research Communications Group, the RIN has been able to establish a number of jointly-sponsored projects and initiatives, including a review of the current state of knowledge about scholarly journal publishing in the UK . A paper has been published , endorsed by all the key bodies representing librarians and publishers in the UK, as well as research funders, setting out the key principles and public policy goals we are seeking to achieve through the scholarly communications process.
The RIN has also produced various studies of researchers' behaviour and use of information services. A report on researchers' use of academic libraries and their services  shows that arts and humanities researchers differ in a number of ways from their colleagues in the sciences in their use of library services. It also shows that they are as yet much less aware, and make significantly less use, of e-publications and open access services of all kinds.
We draw on this kind of evidence and the discussions outlined above in the rest of this article.
Online publication of journals, whether open access or not, has not yet proceeded so far in the arts and humanities as it has in other disciplines. Surveys undertaken on behalf of the Association of Learned and Professional Society Publishers  clearly demonstrate that in the arts and humanities a lower proportion of scholarly journals are published online than in science, technology and medicine. Hence print is still an important part of the landscape; and a RIN survey in 2006-7  showed that three fifths of researchers in the arts and humanities (compared with one fifth in the life sciences and physical sciences) still rate print versions of current issues of journals as very useful for their research.
Nevertheless, online publication, usually accompanied by a print version, is becoming increasingly the norm, and the evidence suggests that researchers in the arts and humanities, like their colleagues in other disciplines, welcome the benefits it brings in ease of access. The popularity of JSTOR's provision of backruns of journals, which are particularly strong in the humanities, is witness to the speed with which online access has been taken up.
Familiarity and use of online journals is not matched, however, by awareness and take-up of the potential for open access to journal literature. The RIN survey showed that as authors, only 14% of arts and humanities researchers (compared with 30% in the physical sciences and 36% in the life sciences) say that they are familiar with the options for making their research outputs open access. Hence only very small proportions of researchers have as yet published in open access journals or made their articles available in open access repositories. Under 3% (compared with 15% in the life sciences and 9% in the physical sciences) say they normally  publish in an open access journal; and 6% (compared with 9% in the life sciences and 21% in the physical sciences) place copies of their articles in an open access repository.
As readers of journal articles, 18% of arts and humanities researchers (compared with 44% in the life sciences and 32% in the physical sciences) say they are familiar with methods for finding open access material. Just 6% say they use open access journals frequently  (compared with the 35% of life science researchers who use BioMed Central journals alone). The proportion of arts and humanities researchers who say that they visit their own institution's repository frequently is in line with the average for all disciplines, however, at 7%.
The low levels of awareness and take-up of open access journals (we consider open access repositories below) stem in part from the nature of research in the arts and humanities, and the way it is funded. Most research in this group of disciplines is produced by individual researchers without the support of a specific project grant. There are some 12,000 research-active staff in the arts and humanities in UK universities and colleges; but the AHRC awarded in 2006-7 fewer than 300 research and related grants, along with just over 300 grants to support - in conjunction with the relevant university or college - a short period of research leave . There are relatively few other sources of project grants, and the great majority of researchers are thus supported not by such grants, but rather through their university or college's block grant from one of the UK Higher Education Funding Councils. That block grant is delivered mainly through the so-called QR funding stream which is based on the results of the Research Assessment Exercise (RAE). This enables institutions to provide researchers with the precious resource of time to conduct their research, with perhaps some small amounts of money for travel, consumables and so on.
Along with the other Research Councils, the AHRC now allows for the inclusion in its grants of provision for the payment of publication fees as indirect costs associated with the project. But since the AHRC funds only a small proportion of arts and humanities research in the UK, this provides for only a minority strand of such research. It is unlikely that universities and colleges will be willing to cover a significant amount of authors' publication costs unless a related funding stream is made available—though it is worth observing that it is not unknown for university departments to contribute to the publishing costs of conventional monographs. These funding restrictions limit the scope for open access journals in the arts and humanities funded through a publication fee model; and thus it is not surprising that there is no equivalent in these disciplines to BioMedCentral and the Public Library of Science.
Publication fees are not the only model for open access online journals. Their costs may be subsidised by direct financial support from a university or department, or from membership subscriptions to a learned society, or from sales of the printed edition. In many cases the subsidy is indirect, in the form of the unpaid time provided by their editors. Scholarly publication has long been reliant on unpaid effort, such as the input of the referees who undertake peer review for commercially published journals. But many other demands on researchers' time make the extension of this practice into the editorial and production process difficult to sustain. For this reason, it tends to be most suitable for quite specialised 'niche' journals. There were some 500 open access journals in the arts and humanities listed in the Directory of Open Access Journals (DOAJ)  in summer 2007, and many of them take this form (see, for example, from among the journals published from UK universities, Consciousness, Literature and the Arts, published by the Department of Theatre, Film and Television at the University of Wales Aberystwyth ; and 49th Parallel: An Interdisciplinary Journal of North American Studies published by the University of Birmingham [12). Such journals can play an important role in their subject communities, and although they do not all operate rigorous peer-review systems, many can and do.
One interesting possibility for the development and support of these journals in the future may be the concept of the 'overlay' journal. On this model, the journal becomes a specialist peer review service. The journal is not a publisher of content, but provides an interface to papers deposited in institutional or subject repositories, guaranteeing that they have passed the journal's quality standards. Such a redefinition of the journal's role would make the publication of OA journals much easier and more efficient. A project under the Institutional Repositories programme being supported by the Joint Information Systems Committee (JISC), is addressing the relevant technical issues . It is worth noting also that the Social Sciences and Humanities Research Council in Canada announced in the spring of 2007 a scheme of aid to open access research journals, with the objective of assisting journals offering barrier-free access to peer-reviewed scholarship in the social sciences and humanities .
There are advantages for the arts and humanities community as for others in making e-prints of journal articles available in open access repositories as widely and rapidly as possible. But again the arts and humanities position is in some respects different from that of other disciplines. Timely access to the latest literature is important in the arts and humanities as elsewhere; but since the pace of the advance of knowledge is typically slower, arts and humanities researchers are more likely to be interested in the final versions of articles, or post-prints, rather than pre-prints. Although there is a significant amount of informal sharing of drafts, the pre-print culture is much less developed than in many other disciplines.
A closely related issue concerns the very long shelf-life of humanities journals: if researchers are conscientious when they begin a new research project, they may well need to review the relevant literature for a century or so back. Thus the half-life of the use and citation of journal articles is typically much longer in the arts and humanities than in other disciplines. One potential consequence is that arts and humanities journal publishers may be less willing to allow open-access posting of e-prints even after an embargo period. However, it is worth pointing to the success of JSTOR with its provision of backruns of journals behind a 'moving wall' (in effect an embargo period) of typically three or five years. A majority of the JSTOR journals are in the arts and humanities, and they are heavily used by researchers. It is not clear that JSTOR's success has had any damaging impact on subscriptions.
In similar vein, it is notable that in practice many arts and humanities journal publishers do allow authors to make their articles available through repositories, albeit in many (but not all) cases after an embargo period; and that they are thus prepared to accept the risk of a fall in subscriptions. Indeed, the willingness of publishers to allow self-archiving outruns the inclination of researchers in arts and humanities to take advantage of the opportunity. It is not yet clear whether this is a lag due to the relatively late growth of online publishing in this subject area, or whether there are distinctive features of arts and humanities research and publication practices which make it more difficult in principle to persuade researchers to place their articles in open access repositories. The longer, less intense pace of scholarly exchange may mean that arts and humanities researchers feel less need of immediate and toll-free online access to the latest journal articles in their subjects, just as they are less interested in pre-prints than in the final published versions. Some may be willing to save up consultations, if necessary, for occasional trips to major research libraries. Others, knowing that many items will not in the end have been worth the effort of tracking them down, may make the pragmatic decision to do without them, accepting the risk of missing something important. Despite these qualifications, anecdotal evidence, as well as access logs and the correspondence received by authors, suggests that arts and humanities material in OA repositories is consulted, and that its accessibility is valued beyond the specialist subject community - by academics in countries without well-resourced libraries or without a tradition of study in a given discipline, and also by independent researchers and members of the non-academic public.
A final distinctive feature of the arts and humanities relates to revenues and costs. As the RIN's UK Scholarly Journals 2006 Baseline Report  notes, it is difficult to get authoritative and consistent data on journal prices. But the evidence shows clearly [16 that in general, the cost to libraries of subscriptions for arts and humanities journals is lower than in other domains. Many journals are also available at moderate cost to private subscribers. On the other hand, it must be recognised that the resources available to support research in the arts and humanities are, as noted above, much more limited than in the sciences; and there is no evidence to suggest that the ratio of the cost of subscriptions to the resources available to support them is more favourable. Indeed, there is evidence to suggest that library subscriptions to arts and humanities journals have been reduced in some institutions in order to meet the rising costs of science journals.
Resource constraints thus bring particular problems for the arts and humanities in relation both to the established subscription-based model for journals, and to open access models: there is less resource than in other disciplines to support either model. It is especially important, therefore, that careful thought should be given to the costs, benefits and implications of moving to new models. The issues include how to sustain the positive role that the established subscription model has played in journal development, the valuable support it has provided for the work of learned societies, and the UK's strong position as a publisher, and exporter, of learned journals in the arts and humanities.
Moves towards open access to journal literature have the potential, we believe, to bring clear benefits to the arts and humanities; at the very least, they will promote a healthy pluralism, and provide other options to traditional business models and modes of access. Some of the benefits may be indirect: if one of the longer-term consequences of open access is to reduce the cost of science journals, this could (but of course may not) help to reduce the pressure on library budgets in the arts and humanities. But overall the benefits may be more limited than in some other domains, and the pace of change is likely to be slower: neither open access journals nor open access archives are likely to have a transforming impact in the arts and humanities for the foreseeable future.
Monographs probably play a larger role in research in the arts and humanities than in any other domain. Increases in the number and the prices of monographs, as for journals, put pressure on library budgets. A market in e-monographs is developing, for instance through Oxford University Press's Oxford Scholarship Online initiative  which makes over 1,650 monographs available on a subscription system, in subjects including classics, history, linguistics, literature, philosophy and religion, and through Cambridge University Press's eBookstore . There is also some self-publishing of e-monographs.
The pace of development remains relatively slow at present, partly because at present people find e-monographs difficult to read and therefore like to have the traditional printed form. It is notable, for example, that the Oxford University Press experience to date is that online availability had not so far affected print sales. In fact, the continuing advantage of printed media for sustained or close study may mean that online availability promotes print sales: if readers have the opportunity to browse a book online, that may help them to decide that the item is one they need to acquire in printed form. The advantages of printed over electronic media for readers also make print-on-demand a significant development. It is already used by established publishers to keep older titles available, and may provide a more cost-effective way of publishing monographs of a specialist nature—although to judge from some existing print-on-demand services, it is not guaranteed to lead to reductions in cost to the consumer.
There are other kinds of book in the arts and humanities which are facing increasing difficulty in finding conventional publishers, and may benefit from e-publication: conference proceedings, for instance, or the publications of historical record societies. So far, however, the possibilities of electronic publication do not seem to have been much exploited in these respects. The points made above mostly concern 'secondary' research literature, rather than 'primary' source texts. It is worth emphasising that e-publication has been immensely important for research in the arts and humanities through the creation of primary text archives (though there are often issues of quality assurance, metadata, and so on). On the one hand we have well-established resources produced commercially, e.g. Patrologia Latina  and Early English Books On-line  or with charitable and other funding, e.g. Thesaurus Linguae Graecae . These are subscription-based, though often free at the point of use for scholars whose universities have paid the (sometimes substantial) subscription. On the other hand there are also mass-digitisation projects such as those sponsored by Google, Microsoft, and the Open Content Alliance. Overall there has been an extraordinary transformation in the accessibility of textual material in the arts and humanities. A particularly powerful new development is that of new forms of electronic critical edition. These are especially interesting for their systematic use of hyperlinking and the association of text with images, thus creating a new kind of publication impossible through conventional means. The reader can move freely backwards and forwards between the edited text, the apparatus of notes and variants, and transcriptions and images of source texts: a well-known example is the series published by the Canterbury Tales Project .
We should also keep in mind the prospect of developing other new kinds of multi-media e-book, combining text with still or moving images and with sounds. This may be particularly important for the creative and performing arts, but is not much developed yet; see for example Douglas Tallack's 3 Cities Project , or the e-books and online writing of Jorn Ebner [224 and Sue Thomas .
In the arts and humanities, as in the sciences, the UK lags behind a number of other countries in making theses available in digital form. The EThOS initiative  promises to address this issue and to make UK theses available on open access for global use. The case for this is as strong in the arts and humanities as in other domains. Most theses do not deserve to be formally published, but they may still contain material that would be useful to a limited number of researchers working in the same area. At present they are likely to be unknown unless someone cites them - and even then, the difficulty of getting hold of them means that people generally will not bother unless there is really strong evidence that they will be essential. Open access online publication of theses will make it much easier to find out that they exist, and to get access to them. Routine thesis deposit would also be a way of getting new researchers into the open access repository habit. Knowing that a thesis will be publicly accessible might even provide an incentive to improve its quality - to the examiners, as well as the candidate.
As the EThOS programme develops, however, an issue that will require consideration is the effect it will have on the practice, probably more common in the arts and humanities than in other domains, of turning a thesis into the author's first book: will publishers be less willing to support this if the original thesis is freely available? On the other hand the pressure to turn a thesis into a 'first book' might be reduced if the thesis itself were publicly accessible, and were regarded (e.g. for the purposes of research assessment) as a research output. At present, someone who has just completed a major doctoral research project must, in order to qualify as research-active, either rush it into print prematurely, or produce some new research, at just the time when the treadmill of temporary teaching posts makes it impossible to get any research done.
Electronic data production is an important element of arts and humanities research today as it is in other domains. Almost half of AHRB/AHRC-funded research projects have produced electronic data of some kind. With the Economic and Social Research Council (ESRC), the AHRC has until recently led the other Research Councils in operating, through the Arts and Humanities Data Service (AHDS), a system for the preservation of data output from council-funded projects; now that funding for the AHDS has been withdrawn, however, there are concerns for the sustainability of the data resources that it has, or would have, preserved. Sustainability raises much broader questions than preservation, especially in the arts and humanities. The typical AHRC-funded data output is a complex online multi-media database, and this presents major sustainability problems in both academic and technical terms: as regards the need, that is, both to keep the academic content up to date, and to ensure that the resource remains fully functional through the various changes that take place in the surrounding digital environment. Dealing with the issue of sustainability should also require due attention to issues of usability and reusability, including harmonisation and interoperability, and to quality assurance. It is fair to say that current and past AHRC-funded data-creation projects could probably benefit, or could have benefited, from more attention to these issues; we need to put structures and procedures in place to ensure that future projects are informed by the right kind of expertise in this respect.
The distinction between publication and mere dissemination should be as important in respect of arts and humanities research data as it is of research literature. To serve the needs of scholarly research, data outputs need rigorous quality assurance procedures, including peer review, comparable, but not identical, to those operated by reputable research literature publishers. The AHRC's ICT Programme has recently funded a project on this issue led by the Institute of Historical Research and the Royal Historical Society . Based on an extensive online survey, focus groups, interviews, and benchmarking reviews by specialists of a range of Web-based digital resources for history, archaeology and classics, it proposes a multi-staged assessment process after the completion of a resource but before its 'publication'. The results of the process, which may involve learned societies and/or subject associations, would be made publicly available and, where appropriate, incorporated in the resource.
The integration of research literature and research data is as important an agenda for research in the arts and humanities as in other domains. The AHRC ICT Programme is currently funding a trial project at the University of York on linking e-archives and e-publications . Based on the on-line journal Internet Archaeology , the aim of the project is 'to investigate novel ways in which electronic publication over the Internet can provide broad access to research findings in the arts and humanities, and can also make underlying data available in such a way so that readers are enabled to 'drill down' seamlessly into online archives to test interpretations and develop their own conclusions'. This sort of linking is particularly powerful in archaeology, which leads other humanities disciplines in its use of electronic resources, particularly for the preservation of excavation data. But it is potentially just as important in other disciplines, for instance theatre and design studies, where multi-media data is particularly prominent, as well as in other areas of the humanities. There is also considerable scope for linking print monographs to (updatable) supporting online data.
Finally, a number of broader issues that came up in the course of the AHRC discussions, though none of them are specific to the arts and humanities:
This survey may create a frustrating impression of limited penetration and slow progress. But if we step back and compare the current situation with the situation a decade ago, we may instead find the changes in the arts and humanities research landscape astonishing. At the very least, we perceive a bridgehead: new concepts and resources have become firmly established; it is possible that developments will accelerate from this point. A risk is that discussions of e-publishing and open access are dominated by the sciences. Given the much greater volume and financial weight of scientific research, this is inevitable. But it is important that the distinctive needs of the arts and humanities are not simply overlooked. Advocates of Open Access should be more aware of, and receptive to, the perspectives of the arts and humanities disciplines, and extend their advocacy to the arts and humanities community. For this to be possible, that community itself needs to develop a broader and better-informed internal dialogue about its e-publication needs, and the access issues associated with them.