IPR issues facing open-access

This article reports on some of the findings of the UK JISC-funded RoMEO Project (Rights MEtadata for Open archiving) with regard to the copyright issues faced by academic authors wishing to self-archive their research papers. It concludes that much depends on the rights distribution chain between university, academic and publisher, and that a redistribution of rights amongst the various stakeholders is required.


Introduction
The UK JISC-funded RoMEO Project is investigating the IPR issues related to the 'selfarchiving' of research papers by academics and the subsequent disclosure and harvesting of metadata about those research papers using the Open Archives Initiative's Protocol for Metadata Harvesting (OAI-PMH). During the first phase of the project we performed online questionnaire surveys of four key stakeholder groups: academic authors, journal publishers, OAI Data Providers and Service Providers. We have also performed a very interesting analysis of 80 journal publishers' Author Copyright Agreements. The principle outcomes of the project are, firstly, the development of some simple rights metadata by which academics might protect their research papers in an open access environment, and secondly, a means of protecting the rights in all that freely available metadata that may soon be available.
The self-archiving process A simplified model of the self-archiving process consists of eight main stages: 1. Academic writes a paper 2. Academic self-archives the preprint (e.g. on an Institutional Repository) 3. Academic sends to a publisher 4. Publisher referees and accepts 5. Academic self-archives the post-print 6. Repository (Data Provider) creates metadata 7. Service Provider harvests metadata 8. End-user accesses the paper Unfortunately, each stage raises IPR questions. Due to space constraints, the written version of this paper will only cover the issues faced at stages 1-5. To read more about stages 6-8, please see RoMEO Studies 5: the IPR issues faced by Data and Service Providers 1 .

Stage 1 -Academic writes a paper
The first question here is: who owns the copyright in that paper? Arguably, under law it is the employer, however, by custom & practice it's usually the academic 2 . The second question is, does the academic know who owns the rights? Our survey showed that 61% of authors thought that they owned the rights in research papers. Similarly, an ALPSP survey in 1999 showed that 79% of authors thought they should retain copyright. Of course the copyright ownership issue is complicated if there is more than one author and if academics include third-party material in their papers. Half our respondents said that between 71-100% of their papers were multi-authored, meaning they would all need to agree whether to self-archive, when and where. One-quarter of respondents included third-party materials in their paper, meaning the third-parties would have to agree to the self-archiving in addition to the formal journal publication.

Stage 2 -the academic self-archives the preprint
In some disciplines there is no preprint culture, but for the sake of this paper we will assume that there is. The questions here relate to how that preprint will be protected in an openaccess environment. Copyright law still applies to open-access works, but the law only allows limited use by third-parties. Of more interest is how academics actually want their work to be protected. Our author survey showed that the majority of academics are happy for their work to be displayed, printed, saved, given away and excerpted from, as long as they are attributed as the author and all copies are exact replicas of the original. That is, far more liberally than copyright law or publisher e-journal licences allow 3 .

Stage 3 -the academic sends the paper to the publisher
The question here is: will the publisher see self-archiving as 'prior publication' (the so-called Ingelfinger Rule 4 , 5 ) and refuse to even consider the paper? Seventy-five per cent of the copyright agreements we analysed asked the author to warrant that the work had not been "previously published". However, only two explicitly stated that self-archiving was considered prior publication. It might be assumed that the rest that prohibited self-archiving also held the same view, however, there were some publishers that specified "no prior publication" but allowed self-archiving.
It all comes down to your definition of "publication". To academics, it means formal publication in a recognised peer-reviewed vehicle. However, in most debates publishers' have adopted a broader, dictionary definition of 'publication', such as, "the act or an instance of making information public". I would suggest that most academics do not see self-archiving as an alternative to peer-reviewed publication. Indeed, I like to make the distinction that selfarchiving provides communication, whilst peer-reviewed journals offer publication.

Stage 4 -the publisher referees and accepts the paper
Once the publisher has accepted the paper, will they ask for copyright assignment -or an exclusive licence -or a non-exclusive licence? Our analysis showed, that 90% of agreements asked for copyright assignment, 6% required exclusive licences, and 4% nonexclusive licences. However, we found that exclusive licences were often equally as restrictive as copyright assignment. Thus, if I give you my coat -you have my coat and I don't. If I licence you my coat, on an exclusive, perpetual basis, the effect is the same: you have my coat and I don't. This is something for authors to be aware of.
The time of copyright assignment is also important to the self-archiving process. If you don't have to assign copyright until after the work has been refereed, an author could self-archive the refereed version before assigning copyright, and not break any laws. (The publisher could then refuse to publish it, but depending on the publication the risks vary). However, 69% of agreements required copyright assignment prior to the paper being refereed, meaning only the preprint could be legally posted in this way. Worryingly though, in 15% of these cases, copyright would not revert to the author if the paper was rejected, meaning those papers theoretically could never be published elsewhere.
So, what happens if authors don't own the copyright in their journal articles? There are two scenarios where this might be the case. Firstly, if the employer owns copyright, and secondly, if a government owns copyright. Interestingly, 57.5% of agreements acknowledged that the US Government may own copyright. However, US Government-owned works are by definition in the public domain. This illustrates that many publishers can and do cope with a parallel publication model where a work may both be in the public domain and also published in a peer-reviewed vehicle.
Assuming it is down to the individual academic to assign copyright, how will they respond to that? Our author survey showed that although 90% assigned copyright, 49% did so reluctantly. Over half of the remaining 10% signed "exclusive licence agreements" instead, which may not be a good sign.
The next question is, what does the academic get in return for copyright assignment in terms of any rights to use their own work? It was concerning that 28.5% of agreements did not grant authors the right to use their work in any way whatsoever. Of course, the right in which the RoMEO Project was most interested was the right to self-archive. 42.5% of agreements (representing 49.1% of journal titles covered by the survey) allowed academics to self-archive their works, however, there were no standard conditions under which self-archiving could take place. For example, there was also no real consensus as to whether the preprint or postprint (or both) could be posted. This makes life very difficult for authors wanting to self-archive and for Repositories wanting to facilitate it.
Stage 5 -the academic then self-archives the postprint If the publisher does allow self-archiving of the postprint, is the author allowed to use the publisher PDF? Only one publisher in our analysis explicitly stated that they would allow the self-archiving of the PDF, however, the author had to purchase it from the publisher in order to do so.
ConclusionsThe self-archiving process clearly raises a number of IPR questions, and much depends on the rights distribution chain between University, academic and publisher. This issue has been discussed in detail in RoMEO Studies 1: The impact of copyright assignment on academic author self-archiving 6 . However, it seems clear that a redistribution of rights amongst the various stakeholders is required. Firstly, Institutional Repositories need the right at the outset to archive employee's work -rather than requiring academics to licence them back the right to do so. Authors, meanwhile, want to be able to publish where they like, and to be able to re-use their own work how they like, and to be able to specify how others' may use their work. Publishers also need the right to publish the definitive version of the work (although not necessarily exclusively), and a means of generating income from the value they do add to the process. We hope the outcomes of the RoMEO project will contribute to these discussions.
A series of six studies based on the project's work, as well as other project outputs including a Directory of Journal Publisher's self-archiving policies, are all available from the Project's web site 7 .