Library Out Loud

To the consumer of Open Access (OA) scholarly publishing it’s really a very simple thing. You find a link to a journal article, a thesis or a book chapter and you click on it. The document is there, you can read it online or download it straight away, you don’t have to buy it or rent it, it doesn’t matter who you are or where you are as long as you are connected to the Internet. Actually the “official” definition of OA as contained in the Berlin Declaration on Open Access to Knowledge in the Sciences and Humanities goes beyond this simple right to consume and includes the right to redistribute and adapt that is now enshrined within Creative Commons licenses, but the term Open Access is still generally applied to documents that are made publicly available under the more restrictive provisions of copyright. The Berlin Declaration saw the Internet as a game-changer in the dissemination of research which had “fundamentally changed the practical and economic realities of distributing scientific knowledge and cultural heritage” at the same time as maintaining scholarly standards by creating “a comprehensive source of human knowledge and cultural heritage that has been approved by the scientific community.”

Money makes the world go around

For over a decade the prospect of widespread open access to scholarly publishing has been of considerable interest to libraries faced with what seems to be an endless progression of journal subscription price increases. The financial model for traditional academic journal publishing is a strange beast. Researchers, generally located in universities and other research institutions, submit articles to journal editors who either reject them outright or send them to reviewers. The reviewers read the articles and accept them, reject them or suggest changes. This may take some time particularly where changes are suggested which will involve at least one re-examination of the work before it is either accepted or finally rejected. Finally those articles that have been accepted by the reviewers are sent for copy editing and are published in crisp printed volumes, as highly-produced online PDFs or probably (in the case of the big publishers) both. In order to have the work published, authors are required to sign over copyright to the publishers – they retain the moral right to be identified as the authors of the work, and in the days of print they were sometimes granted a small number of reprints to distribute to colleagues, but no right to make and distribute further copies of their work.

This process involves the work of three parties – authors, reviewers and editorial staff – but only the last group receive payment for their work. In the meantime the journals are sold back to the universities and research institutions – that employ the authors and reviewers – at exorbitant cost. (Anyone who has been anywhere near the staff of the big publishers will recognise the aroma of serious money they give off.) This system predates the Internet by many years and at the time access to the web became ubiquitous among the research community about twenty years ago there was considerable hope that authors would use their ability to self-publish, or to publish cooperatively, and that the financial leverage of the big publishers would be massively impeded. This way not only would they avoid giving away their time for free in order to feed the fat cat publishers, but the power of the Internet would take their work free of charge to a whole new audience who had previously never been able to access it in print. Indeed the 1990s saw the first round of what we would call OA journals in the form of a number of Internet Journal of type titles, but many of them had faded away within a few years. Why was this?

What’s in a name?

To work as an academic researcher is generally to forego financial reward in order to enjoy a certain amount of freedom of inquiry and the ability to ask “real” questions. However, scholars are not entirely unworldly and universities run a nicely-graded system of prestige rewards based on honorifics – Doctor, Associate Professor, Professor, Distinguished Professor and so on. (These things matter, even when we pretend that they don’t.) Well, this prestige factor applies to journal publishing as well, and it was pretty soon obvious that researchers wanted to publish their work in the “best” journals, and that these were also the journals that they read and also that they cited. The real value of journals was reputational and as long as a publisher owned the particular quantum of reputation that the journal represented then they could charge a very considerable amount for access to its contents on the understanding that no library was going to cancel the subscription. (Libraries have reputations and prestige to look after as well.)

Of course reputation is not manufactured out of fresh air and, just as a Distinguished Professor can point to a considerable body of work of high importance, so journal editors and publishers were able to show through the use of citation counts and Impact Factors that they did indeed publish work of quality. What is more many of those who had promoted the first round of free journals had found that there was quite an amount of work involved in handling submissions, arranging reviewers, publishing online and retaining online archives if they wished to maintain the quality of their publications, and many of these titles had faded within a few years unless there was substantial institutional or scholarly society support to underwrite their costs. So journal prices continued to rise at well above the rate of inflation and access to their contents remained restricted to staff and students of academic and research institutions, but in the early 2000s there was a big payoff from the Internet to these groups in the form of the publishers’ “big deal” agreements which gave them unprecedented access to far more titles than they had ever believed possible. Libraries’ costs went up but so did their reputations so it looked like a win-win-win. (Publishers, scholars/students, libraries.) Only the general public, whose taxes underwrote the whole business, were excluded from the feast.

What to do about this?

There have been a number of responses to this frustrating situation and together these constitute the Open Access movement. The most direct method of making something OA is to simply place it on the Internet, but this was generally in breach of the copyright agreement that the authors had signed with the publisher not to distribute it themselves. (In the field of physics there was a longstanding tradition of giving fast access to new research by making preprints – versions of the article prior to peer review – available through the Arxiv repository, but this was practice did not spread to other disciplines.) It was sometimes argued that copyright was not an absolute barrier to open distribution as long as the open version contained enough points of difference from the copyrighted version, but this was never a very satisfactory solution as no one would ever have been exactly sure what the differences were and in the end good scholarship depends on the ability of authors and readers to come to a shared understanding of exactly what is being said. Another method of making works OA is to flip the financial model and make the authors pay for the privilege of being published in the journal while distributing the articles for free. This solution recognises that even with the extensive involvement of non-paid authors and reviewers there are still substantial organisational and production costs involved in quality publishing, and also that researchers and research organisations receive considerable benefit from having their work appear in properly credentialed publications. The third method of making peer-reviewed works OA is that neither the reader nor the author pays the journal’s costs but that someone else does, typically a university or research institution (which is often the editor’s employer). As mentioned above this fully open system began shortly after the Internet made it possible, and although it proved difficult to sustain in many cases it has never entirely disappeared. A fourth method through which peer-reviewed articles enter into publicly accessible webspace is through delayed OA, when access restrictions are removed from articles after a defined length of time.

There are other possible solutions of course, including open peer review, but that is a little beyond our present scope, so I’ll just concentrate on these four and then we’ll have a look at mandates.

Green Open Access

In a 2001 article in Ariadne Six proposals for freeing the refereed literature online Stevan Harnad wrote that “Self-archiving all preprints and postprints can be done immediately and will free the refereed literature overnight. The only things holding authors back are (groundless and easily answered) worries about peer review and copyright.” Harnad’s argument was that authors had a basic ownership of their words and a right to distribute them as they wished that copyright assignation did not deprive them of, despite the claims of publishers that peer review gave them a stake in the finished product, and his forthright statement of the case provides the core justification for the inclusion of copyrighted works in digital repositories. An important aspect of the argument was that this right lay with authors rather than their employers or funders, and while employers or funders could require their researchers to archive works (more on this later) the act of archiving had to be done by the authors, hence the term “self-archiving”. While journal publishers (who often refer to themselves as “copyright owners”) might not have considered their claims to the right to limit distribution to their subscribers as groundless, they nevertheless came to the party and within a few years most of them did allow for versions of the articles from their journals to be placed online under fairly tightly-defined conditions. Typically these are

• That the officially branded and paged PDF as published in the journal is not used, although usually they allow the “author postprint” (the accepted version of the article after peer reviewing is completed) to be used. A minority of publishers are more restrictive and insist that only the preprint (the article prior to peer review) can be used, while others allow use of the official PDF or even insist on its use.
• That the article must be placed in a recognised digital repository or, at the very least, on an institutional website. Usually publishers insist that the repository be an institutional one, like Massey Research Online, rather than a subject-based one like E-LIS (Eprints in Library and Information Science).
• That a period of embargo, ranging from a few months to two years after publication, is observed before the article is archived.

The University of Nottingham has codified these allowances at publisher and individual journal level on its Sherpa/Romeo website and it is this classification system that the Symplectic system uses in determining the eligibility of articles to be deposited into Massey Research Online.

Gold Open Access

A Gold OA article is one that is openly available to all Internet users in its published form free of charge. It may appear in a fully Gold OA journal like PLOS One, BMC Public Health or Language Learning and Technology or it may appear in a so-called hybrid title like Molecular Biology and Evolution which is a subscription journal that will make an article OA if the authors pay a fee. Journals from Public Library of Science (PLOS One, PLOS Biology etc) and BioMed Central (BMC journals) are fully commercial operations but they make their income from charging their authors rather than their readers. Article Processing Charges (APCs) can vary from a few hundred dollars up to a few thousand and are intended to cover the costs of peer review and publishing. Both Public Library of Science and BioMed Central have strict peer review policies but PLOS do not exclude articles on the basis of “significance” – instead they leave it up to their readers to decide whether the article was worth publishing once their reviewers have determined that it follows sound research protocols. This approach seems to have been affirmed by the research community with their journals attracting work from important authors and healthy rates of citation.

Language Learning and Technology and similar titles (Journal on Smart Sensing and Intelligent Systems is a Massey-based example) follow a more precarious financial model in that they charge neither their authors nor their readers, relying instead on the time and energy of their editors and/or institutional or organisational support. Over time the number of these titles has declined, but for clearly-defined research communities they can still offer a real alternative to the commercial models.

Delayed Open Access

Some publishers recognise that their journals have a high value during the first few months after publication and that readers in subscription-paying research institutions would not tolerate a delay in access to their contents, so they continue to make a good income from selling to libraries but make their contents OA after a period of six months or a year. Typically these are scholarly society journals rather than “commercial” titles in areas of science with a high turnover of new ideas and my own research suggests that they make a measurable impact on the total amount of OA publishing. There is some reluctance to regard this as a form of Open Access and it falls short of the Gold OA definition of being “born free” but for a title like Molecular Biology of the Cell which makes its contents free after a month it seems like OA in all but name. There is a good list of Delayed OA journals at Highwire Press.

Document Sharing

So far we’ve looked at OA very much as an author-to-publisher-to-reader linear system, but there are other ways in which documents can circulate. The most obvious alternative to getting something from a vendor is to get it from any person or institution willing to share it with you. Libraries have always been in the document sharing business and most would still claim that “even if we haven’t got it we can still get it for you” through inter-library loan. The Internet has made this possible on a huge scale and emailing a colleague whose library subscribes to a journal to which yours doesn’t is a widely used method of accessing information. Publishers have had to turn a blind eye to this as it was not only impossible to police but confined to the academic community, but they still prohibit placing articles on open websites which offers a real alternative to their business of placing articles on closed websites.

Sharing, however, is very much part of Internet culture and there have been a number of large-scale document-sharing utilities that challenge the copyright model. These include Mendeley (whose sharing features have been severely limited after being bought by Elsevier), Academia.edu and Researchgate. I did a quick survey of the availability on Researchgate of a sample of over 300 articles published in 2012 and found that over 70% were represented with a record and over 25% were available full-text. So far Researchgate hasn’t received a serious challenge from the publishers and as it requires one of the authors to upload the full-text it is not strictly speaking a document sharing service in the way that Mendeley used to be. It offers another intriguing feature (which I haven’t tried) whereby you can request the full document from whichever of the authors has deposited the record. You have to be logged in to use this feature, and you have to be a member of an academic or research community (or invited by one) to register so it is not strictly OA, but it is an intriguing development. Ease of uploading documents and lack of obvious restriction or vetting make Researchgate a tempting alternative to repositories for time-poor researchers but whether it acts for the good of OA and the non-affiliated researcher over the longer term remains to be seen. Watch this space.

Bruce White

Next week: Mandates, Upsides, Downsides, Open Research and Who Benefits?