Research Support Librarian, Open University
One of the primary tools of the Open Access (OA) movement is the institutional repository (IR) — a freely-available web-accessible database in which university faculty are able to deposit their research outputs, notably papers that they have published in scholarly journals, and also books and book chapters.
The genesis of the institutional repository can be traced back to a 1999 meeting held in Santa Fe New Mexico, where the so-called Open Archives Initiative Protocol for Metadata Harvesting (OAI-PMH) was formulated. The aim of the meeting was to create an infrastructure that could build on the success of the physics preprint repository arXiv.
Founded by Paul Ginsparg in 1991, arXiv had become an important resource for scholarly communication within the physics community, and there was a growing desire to replicate the model in other disciplines. During the Santa Fe meeting a strong case was also made for creating institutionally-based repositories that catered for all research areas in a single university, and over time the IR has become the dominant repository model. However, the primary aim of the Santa Fe meeting was to create a protocol to make repositories interoperable, regardless of whether they were central subject-based repositories or institutional repositories.
Why the need for repositories? After all, scholarly communication was outsourced to publishers long ago. The appeal of the model arXiv pioneered, however, was that it exploited the ability of the Internet to allow research results to be communicated much more rapidly than was possible with traditional publishing — where publishing a paper in a scholarly journal or book can take many months, or longer, a researcher can deposit a paper in an online repository the moment it is completed.
More importantly, traditional scholarly publishing was in crisis. Since the end of WWII an explosion of new scholarly journals, constantly rising subscriptions, and falling library budgets, had created a situation in which universities and other research institutions could no longer afford to buy all the journals their researchers needed. Moreover, even though publishers had begun migrating their journals to an online environment, subscription prices were not falling (as would have been expected, since traditional costs like printing and physical distribution go away on the Internet), but inexplicably continuing to rise. The suspicion was that the fundamental problem was publisher greed.
It should be noted that arXiv was intended to supplement the traditional model (by sharing preprints prior to publication), not to replace it. Nevertheless, its model was sufficiently compelling that some also viewed it as a solution to serial price inflation, and had begun to call on colleagues to make copies of all the papers they published in scholarly journals freely available on the Internet, by self-archiving them. If every researcher did so, it was reasoned, the research community's access problem would be resolved.
Again, the aim was still not to replace traditional publishing but to supplement it. In fact, the objective was quite simple: If researchers belonged to an institution that subscribed to the journal in which a particular paper they needed had been published they could access the publisher's version of the paper using their institution's subscription. If, on the other hand, their institution didn't have a subscription to the journal in question, they could use the author's self-archived version. In this way, it was assumed, all 2.5 million articles published in the world's 24,000 scholarly journal each year would be freely available to all — in one form or another.
It was apparent, however, that if researchers simply dumped their papers hither dither on the Internet it would be difficult for others to locate them. What was needed was a custom-built software platform to allow universities to create a dedicated repository in which faculty could archive them. And as the emphasis shifted from central subject-based repositories to smaller cross-disciplinary repositories, it was realised that a low-cost solution would be needed. In 2000, therefore, the UK's University of Southampton released EPrints. The first dedicated repository software, EPrints was made available as freely downloadable Open Source software.
Importantly, EPrints was OAI compliant — which meant that EPrints repositories could expose standardised metadata descriptions of their contents on the Internet. These could then be collected by specialist harvesters and aggregated into a virtual cross-searchable global archive offering a single search interface. When a search was conducted the hits would then link back to the source material in the host repository. To this end in 2002 the University of Michigan launched the first OAI harvester, OAIster.
But while the objective of the self-archiving movement may have been simple, implementation has proved enormously difficult, and nine years after the Santa Fe meeting only around 1,000 of the world's 22,000 research institutions have yet to create an institutional repository. Moreover, those who have done so generally discover that only about 15% of their researchers will spontaneously deposit their papers in them.
In short, the IR movement has been confronted by a number of unanticipated challenges. First, as indicated above, it has proved immensely challenging to persuade researchers to take on what most quickly conclude is a thankless and burdensome additional chore. Second, getting senior management to support self-archiving, or provide the necessary funds to create and manage an IR, has proved nearly as difficult. Third, many publishers have sought to obstruct self-archiving, fearful that if free copies of the papers they publish become widely available on the Web their subscription revenues will dry up.
Since scholarly publishers have historically made it a condition of publication that researchers assign copyright to them, they have had a strong hand to play. By insisting on copyright transfer they effectively acquire ownership of the papers, and many have either refused to permit self-archiving, or insist that authors only do so after an embargo period.
In recent years, however, effective lobbying by OA advocates has begun to make some headway. Increasingly conscious that they are generally ignorant about the research output of their own faculty, for instance, university managers have begun to warm to suggestions that institutional repositories are the natural tool to collect the kind of management data they need if they want to monitor the productivity of faculty.
OA advocates have also had some success in convincing research funders that in a digital environment publicly-funded research should be freely available to all, not locked behind financial firewalls whose only purpose is to protect incumbent business models. After all, point out OA advocates, most research is publicly-funded, and authors give their papers to publishers without charging them. Why then should the research community have to buy its papers back in the form of journal subscriptions? Would it not be better, they argue, if publishers found a business model more suited to the networked world?
In response, funders have started to introduce self-archiving mandates, making it a condition of funding that research outputs are made freely available on the Web. In other words, if researchers won't self-archive willingly, let's make them do so.
To date 21 funders have mandated public access to research findings, including the UK-based Wellcome Trust, the US National Institutes of Health (NIH), and all bar one (the EPSRC) of the UK research councils. In addition 19 universities have introduced institutional mandates, and major universities like Harvard and the University of California are currently considering doing so.
But progress remains slow, and many hurdles remain. There have also been some unintended consequences: After convincing university managers to adopt IRs, for instance, OA advocates have discovered that in many cases it is only bibliographic data that is being deposited, not the full-text. Since metadata is all that is required for information management purposes, and limiting deposits in this way avoids any risk of getting into copyright disputes with publishers, it is not perhaps surprising. From the point of view of Open Access, however, it is a most unsatisfactory development.
But how do things look on the ground? To find out I sat down recently with Bill Mortimer, Research Support Librarian at the UK's Open University. An advocate for Open Access, Mortimer has played a key role in the development of the OU's repository, Open Research Online (ORO).
What became evident during our conversation is that the OU's experience maps neatly onto the history of the self-archiving movement. As in many universities, the library created an institutional repository some years ago, but struggled to persuade researchers to deposit their papers in it. Without the necessary funds to continue supporting ORO, the library was then forced to put it on the back burner for a while. After Mortimer was appointed, however, the OU's Pro-Vice Chancellor for research was persuaded to adopt ORO as a central resource for the upcoming Research Assessment Exercise.
As a result ORO experienced a rapid growth in the number of deposits and today, says Mortimer, it is the fourth largest repository in the UK. But like so many repositories the bulk of the content in ORO today is metadata, and just 15% of its records consist of full-text.
As I talked to Mortimer it occurred to me that the OU's commitment to Open Access is interesting for a number of other reasons. Its philosophy, for instance, is well matched with the values of Open Access. Founded in 1969 by the then Prime Minister Harold Wilson, the OU's mission is to be "open to people, places, methods and ideas." As Mortimer puts it, OA "fits the aims and ethos of the OU like a glove."
The OU might also seem to have more to gain from Open Access than most universities — for while it is widely recognised (and globally respected) as a distance-learning institution, the OU is not well-known as a research institution. What better way of demonstrating its credentials, and showcasing its work, than by making its research freely available on the Web?
Much larger revolution
It also struck me that as a provider of remote education the OU is at the forefront of a much larger revolution, a revolution of which OA is but a component part — that is, the gradual "virtualisation" of education and research. From the outset, for instance, the bulk of the OU's teaching has been delivered virtually — initially via television and radio, later online. And today some of its courses are run entirely on the Web.
As research papers make the transition from physical objects to electronic files located in disparate online databases, scholarly communication is going through a similar process of virtualisation. And here too Open Access is just one piece of it. Today journals are not only accessed electronically, but the entire publication process is becoming virtual, with many parts of it now automated. Papers are submitted online, peer review is managed and undertaken online, and many journals no longer have any print equivalent.
Indeed, some argue that, as a result, the traditional gate keeping role played by publishers is fast becoming redundant, much in the way that the lecture hall is redundant at the OU. Within this larger revolution Open Access starts to seem like an alternative to traditional publishing, not a supplement.
It was no surprise to me, therefore, to learn that the OU has begun to develop its own Open Access ejournals. Nor was I surprised when I asked Mortimer if he thought that ORO might in the future be viewed as more of a publishing platform than a database of research outputs that he so readily agreed with me.
Publishers, however, are not the only gatekeepers in the scholarly communication process. So too are librarians. The question inevitably arises: Could the bricks and mortar library eventually go the way of the paper journal?
As one might expect, Mortimer has views on such issues, and it was interesting to discuss them with him. In the process, however, he revealed himself to be a serious-minded but modest interlocutor. And while he has undoubtedly given a great deal of thought to Open Access, and to the implications for librarians of the brave new electronic world, he was keen to stress that he is a practitioner, not a theoretician.
A pragmatic issue
In short, Open Access is for Mortimer primarily a pragmatic issue. As he pointed out, his job is to ensure that OU faculty have access to all the research they need, not to promote causes. It just so happens that Open Access currently offers him the best hope of achieving this. Indeed, I formed the impression that, while he is happy to discuss radical future scenarios, Mortimer is in many ways a traditionalist, and were journal subscription costs to suddenly plummet, to a level where the OU could provide faculty with all the research it needs, Mortimer would be more than happy to re-embrace the traditional subscription model.
Given this pragmatism it is interesting to note that, with his colleagues, Mortimer has recently suggested to OU management that the University introduce an "Immediate-Deposit/Optional Access" (ID/OA) mandate. The ID/OA mandate is a compromise strategy intended to force researchers to self-archive, but in a way that circumnavigates publisher opposition, and avoids any potential copyright disputes with them. And it is able to facilitate Open Access even where the full-text is not accessible on the Web. In short, it promises a very practical solution to the many hurdles currently besetting Open Access. It is no surprise, therefore, that it should appeal to a pragmatist like Mortimer.
It would be the next logical step for the OU to adopt such a mandate. But will it happen? When I spoke to Mortimer no decision had been made, but he stressed that senior management is "actively considering" the idea.
If you would like to learn more about the Open University and the development of Open Research Online, or are curious as to how an ID/OA mandate might work, please read the attached interview.
If you wish to read the interview please click on the link below. I am publishing it under a Creative Commons licence, so you are free to copy and distribute it as you wish, so long as you credit me as the author, do not alter or transform the text, and do not use it for any commercial purpose.
If after reading it you feel it is well done you might like to consider making a small contribution to my PayPal account.
I have in mind a figure of $8, but whatever anyone felt inspired to contribute would be fine by me. Payment can be made quite simply by quoting the e-mail account: email@example.com. It is not necessary to have a PayPal account to make a payment.
What I would ask is that if you point anyone else to the article then you consider directing them to this post, rather than directly to the PDF file itself.
If you would like to republish the interview on a commercial basis, or have any comments on it, please email me at firstname.lastname@example.org.
To read the interview (as a PDF file) click here.
If, however, you would prefer a more digestible article try the Computer Weekly website.