Open access journals – wheat, chaff and hopeful monsters

The Fifth Annual Conference on Open Access Scholarly Publishing, organized by OASPA, the Open Access Scholarly Publishers Association [1], was held in Riga, the capital of Latvia, on 18th-20th September 2013.  I was invited to attend to discuss the Open Citations Corpus, a repository of open bibliographic citation data initially harvested from the reference lists of open access journal articles.  This post contains my belated reflections on the major themes raised by that interesting meeting.

The conference, which was well organized by Claire Redhead, OASPA’s Membership & Communications Manager, was attended by 85 people, including representatives of some 37 publishers and of 22 other organizations including major funders, policy organizations and libraries, and a handful of other interested parties such as myself.  Both ‘pure’ Open Access publishers and more traditional publishers who publish a mixture of open-access, hybrid and subscription-access journals were represented.  The venue, the Radisson Blu Hotel, provided excellent accommodation and conference facilities, enabling us to work effectively, in five themed sessions of talks and discussion over two and a half days.

The lead keynote speaker for the Riga meeting was Lars Bjørnshauge, Director of DOAJ, the Directory of Open Access Journals, a marvelously useful information source that he founded in 2003 at the University of Lund, which since December 2012 has been under the supportive wing of IS4OA [2].

Lars started the meeting by exposing the first major theme, namely quality control in OA journals.  Given the ease of on-line publishing and the explosive growth of OA journals over the last few years, there is great current concern both in DOAJ and in OASPA to determine which of these open access journals are legitimate scholarly publications, and which are simply attempts on the part of fake or non-legitimate publishers to separate authors from their cash in the form of substantial article processing charges (APCs, aka author publication charges) in return for a mediocre service.  Lars suggested a number of criteria by which OA journals should be judged, include the following:

• That journals should provide the names and contact details for the editorial board.
• That journal articles should have a clear and transparent peer reviewing policy.
• That journal articles should be published under a Creative Commons CC-By attribution license, so that readers’ rights to re-use the articles are clear.
• That journal articles should have DOIs (Digital Object Identifiers), typically, for scholarly articles, DOIs issued to registered publishers by CrossRef.
• That journal articles should be accompanied by machine-readable metadata.
• That publishers should archive digital journal articles for future reference, and that this should be done in machine-readable formats such as an XML DTD such as JATS (the NISO Journal Article Tag Suite), rather than as a PDF document.

To this end, DOAJ and OASPA, together with the Committee on Publication Ethics and the World Association of Medical Editors, have recently jointly published a set of Principles of Transparency and Best Practice in Scholarly Publishing against which to evaluate all OA journals.

The second major theme of the Riga meeting, opened by Michael Jubb of the Research Information Network in the following talk, concerned the exact nature of scholarly OA publishing and its financial implications.  Michael outlined the striking progress being made in the UK to implement recent OA mandates for publication of research work funded by public purse, discussing both the Finch Report and the Research Councils UK’s Policy on Open Access.  This requires that articles be published under Creative Commons CC-By licenses, so that the content may be freely mined and otherwise re-used – the ideal situation for the reader.

This highlighted the distinction, often veiled for nominally open access publications from commercial publishers, between the two types of open access, characterised as Gratis versus Libre:

• Gratis Open Access signifies removal of the price barrier alone, giving a right to read the article off the screen.  However, the publication is still restricted by licenses that do not permit it to be downloaded or reused in any way.
• Libre Open Access signifies removal both of the price barrier and at least some of the permission barriers limiting reuse, giving rights to text-mine and re-use the article.

Thus, while both imply ‘free’ (a potentially ambiguous word), Gratis Open Access equates to ‘free as in beer’, while Libre Open Access equates to ‘free as in speech’, an analogy for which Peter Murray-Rust is to be thanked. Gratis Open Access is thus a necessary but not a sufficient condition for true Libre Open Access.  Many ‘open access’ publications by commercial scholarly publishers are only Gratis Open Access, while almost all publications by ‘pure’ Open Access publishers are Libre Open Access.

Michael also discussed the tension between Gold Open Access (OA article published on publisher’s web site), which typically involves payment of APCs, and has the benefit that the article is findable where you expect it to be, and Green Open Access, where the publisher permits the author to publish, in an institutional repository or similar third-party site, a preprint (the version of the article as first submitted for publication) or a postprint (the pre-publication version of the article after incorporation of authors’ responses to peer reviewers’ comments), often with an embargo restriction that prevents the article from being freely available until some substantial time after publication of the subscription-access journal issue containing that article.  The latter (Green OA) is the cheaper way for UK universities to comply with the RCUK policy, but means that potential readers may have difficulty finding the article, the content of which (if it is a preprint) may not be the same as the published Version of Record.

Representatives of other funding agencies and policy organizations similarly set forth their OA agendas.

The third theme to emerge from the conference was the contentious issue of hybrid journals, initially set forth by Liz Ferguson of Wiley-Blackwell.  Hybrid journals are scholarly journals in which one has to pay a subscription to view the majority of articles, but which may also contain OA articles for which the authors have had to pay substantial APCs.  The bone of contention is the degree to which commercial publishers should or do lower their journal subscription charges as the number of OA articles within them rises, to avoid what is politely termed “double dipping”, i.e. getting academics to pay publishers not just once but twice for the privilege of reading their own works of scholarship.  On this topic, several publishers claimed that they were doing the honourable thing.  Someone commented “The only way for hybrid is to take the publisher’s word”!

Falk Reckling of FWF (the Austrian Science Fund) told the conference of the Austrian experiment to avoid increased library payments to publishers while achieving open access publication, in which publishers were directly reimbursed for all papers funded by FWF, and the amount paid was then deducted from the subscriptions paid to them by the Austrian Library Consortium the following year.  He reported that most big publishers, including Elsevier, Wiley and Taylor & Francis, rejected that proposal.  Falk concluded by saying that only by working together internationally could funding agencies succeed in forcing change on the publishers, a sentiment echoed particularly by Cameron Neylon of the Public Library of Science.

Victoria Gardner of Taylor & Francis surprised me by pointed out that the average cost of publishing an article in the humanities and social sciences (HSS) was more than three times that in science, technology and medicine (STM), largely because of the papers were larger and had higher rejection rates.  She stated that the author-pays APC model was not financially viable for HSS, where many researchers, particularly early in their careers, lack funding income that could be used for that purpose.  There was thus tension in the debate “OA is unworkable for the Humanities” versus “OA is the future” that was not resolved.

Apart from the costs of OA, and particularly of hybrid journals, the other discussion issue concerned their future.  Are hybrid journals stepping stones to a fully open-access world, or are they hopeful monsters, doomed to extinction?  Liz Ferguson reported Wiley to have ~1,200 hybrid journals that between them had published only about 3,500 articles – on average just three articles per journal – with many journals having no OA uptake at all.  Several other publishers reported their experiences with hybrid journals.  But Georg Botz of Science Europe told the meeting flatly that “the hybrid model is not a working and viable pathway to OA.”

Apart from these main themes, there were talks from a number of leading OA publishers about their progress and achievements, and an interesting discussion about the growing trend of Open Access book publication led by Eelco Ferwerda of OAPEN, an open access library of books in the humanities and social sciences, and Cecy Marden of the Wellcome Trust, who discussed funding of OA monographs and the importance of Europe PubMed Central as a repository for OA publications.

The final morning of the conference saw presentations on a number of separate additional themes, including a “meet the OASPA members” session containing inspiring presentations by Brian Hole of Ubiquity Press, who described how quality OA publishing could be achieved at a fraction of the conventional costs discussed above, and by Lyubo Penev of Pensoft Publishers, an innovative OA publisher of journals relating to biological taxonomy, who described the Pensoft Writing Tool, an integrated online collaborative authoring, editing, reviewing and publishing platform that facilitates semantic enhancements and the publication of datasets accompanying articles (see also http://www.pensoft.net/page.php?P=31&SESID=auzykddds), that I have mentioned in a previous blog post.

It was during that session that I gave my invited talk entitled Open Citations Corpus – freeing scholarly citation data, described in a separate Open Citations Blog post.

In summary, OASPA’s Fifth Annual Conference on Open Access Scholarly Publishing revealed the world of open access journal publishing to be one of rapid growth and innovation, and of clear tensions regarding quality control, cost, licensing, and applicability to HSS.  It was clear that commercial academic publishers were attempting to hang on to established modes of publishing, funding and tight licensing while making limited concessions towards openness in the form of hybrid journals and a smaller number of OA journals under terms that benefit them financially.  The elephant in the room, never fully articulated, was the extent to which, and for how long, the commercial publishers’ traditional business model of high profit margins can withstand the competition of the newer Open Access publishers, particularly those offering radically cheaper publishing avenues for scholars.  Cameron Neylon reminded us that disruptive technologies achieve market dominance not by being disruptive but by addressing a recognised and current need, so that their adoption then drives the disruption.

Clearly for the foreseeable future we will have a transitional mixed publishing economy in which ‘pure’ OA journals will exist alongside subscription-access and hybrid journals, with a plurality of business models.  Improvements in mechanisms for micropayments and for splitting payments between multiple authors’ institutions are required.  Funders’ mandates are driving expansions of Gold and Green OA publishing and minimizing restrictions to reuse by adoption of more permissive CC-By licensing, but universities are struggling to find funds to pay OA APCs while maintaining conventional journal subscription payments.  Embargos to access and lack of interoperability between institutional repositories continue to restrict the usefulness of Green OA.  So there is real momentum towards OA, but mixed progress and great disparity between nations.  Further change is inevitable and to be welcomed.

[1]     OASPA was established in 2008 to represent the interests of Open Access (OA) journal publishers globally in all scientific, technical and scholarly disciplines.  Paul Peters from the Hindawi Publishing Corporation is currently Chair of the OASPA Board, and is editor of the OASPA Blog.   The presentations from the Riga meeting are currently available from http://oaspa.org/conference/presentations-coasp-2013/.  The 6^th Conference on Open Access Scholarly Publishing (COASP) will be hosted by UNESCO at their Paris Headquarters from September 17th – 19th, 2014.  Further information will be posted on the OASPA conference page (http://oaspa.org/conference/) as it becomes available. Details of the inaugural OASPA Asia conference, to be held in Bangkok in June 2014, are presently given at http://oaspa.org/conference/. 

[2]     IS4OA (Infrastructure Services for Open Access) is a Community Interest Company based in the United Kingdom, led by Caroline Sutton, the immediate past-president of OASPA and by Alma Swan, a consultant working in the field of scholarly communication who has a long history of supporting open access initiatives.  IS4OA acts as an umbrella organisation for open access activities and services that are of value to the academic community and beyond, providing business structure and expertise and a means of obtaining and channelling financial support for these activities and services.