Stevan Harnad, Universite du Quebéc à Montréal & University of Southampton

Input for the 'First DRIVER Summit', Panel Discussion, 2008-01-16



COUNCIL OF THE EUROPEAN UNION  (comments: Stevan Harnad)


The EU Council's Conclusions show the tell-tale signs of penetration by the publisher anti-OA lobby; familiar slogans, decisively rebutted many, many times before, crop up verbatim in the EU Council's language, although the Council does not appear to realize that it has allowed itself to become the mouthpiece of these special interests, which are not those of the research community.


Council of the European Union: Conclusions on scientific information in the digital age: access, dissemination and preservation


What follows is intended as a constructive critique of this EU Council statement (all boldface quotes are from the Council's statement, the underscores have been added) so that EU Open Access policy can be brought in line with progressive policy in the rest of the world, including what is already the policy of the European Research Council (ERC), which has already mandated OA self-archiving), and which is also the policy recommended by European Research Advisory Board (EURAB), recently also unanimously recommended by the Council of the European Universities Association (EUA) -- as well as being the same policy originally recommended by the European Commission's own Study on the Economic and Technical Evolution of the Scientific Publication Markets of Europe, whose recommendation A1 (once again, to mandate OA self-archiving), and supported by a petition with over 25,000 individual signatories and nearly 1500 institutional signatories:


"the importance of scientific output resulting from publicly funded research being available on the Internet at no cost to the reader under economically viable circumstances, including delayed open access"


(1) The descriptor 'at no cost to the reader' flagrantly conflates paid site-licensing with Open Access (OA). This wording was undoubtedly urged on the EC by the publisher lobby. The focus should be on providing free online access webwide. That is OA, and that is what makes the real objective clear and coherent.


(2) The descriptor 'delayed open access' refers to publisher embargoes on author self-archiving. If embargoes are to be accommodated, it should be made clear that they apply only to the date at which the access to the embargoed document is made OA, not to the date at which the document is deposited, which should be immediately upon acceptance for publication. (The DRIVER network of Institutional Repositories (IRs) can then adopt the 'email eprint request' button that will allow individual users to request and receive individual copies of the document semi-automatically. )


(3) What should be deposited in the author's own institutional IR immediately upon acceptance for publication is the author's peer-reviewed, accepted final draft ('postprint'), not the publisher's PDF (or XML). There are far more publisher embargoes on the PDF/XML than on the postprint, and the postprint is all that is needed for research use and progress. The postprint is a supplementary version of the official publication, provided for OA purposes; it is not the version with the primary digital preservation problem.


(4) Digital preservation should not be conflated with OA provision: There is a (separate) problem of the digital preservation of the publisher's PDF/XML, but this is not the same as the problem of providing OA to the author's postprint. The postprint, though it can and should be preserved, is not the canonical copy of the publication, so the two preservation tasks should not be conflated.


(5) Self-archiving research data is also a different matter from self-archiving research publications. Data-archiving is not subject to a publisher embargo, and it needs independent preservation, but data-access and data-preservation should not be conflated with OA provision.


(6) Deposit should be directly in each author's own IR: Distributed institutional depositing and storage should not be conflated with central harvesting and indexing: Deposit Institutionally, Harvest Centrally.


(7) Direct central deposit should be avoided except in cases where the author is institutionally unaffiliated or the author's institution does not yet have an IR. For those cases, there should be at least one provisional default repository such as DEPOT or EurOpenScholar.


(8) Research (publications and data) should not be conflated with other forms of digital content. The problems of cultural heritage archiving, for example, are not the same as those of research publication archiving. Nor are the problems of archiving the same as the problem of access-provision (OA).


"ensure the long term preservation of scientific information -including publications and data"


This is an example of the complete conflation of OA-provision with digital preservation, including a conflation of authors' supplementary postprints with the publisher's original, as well as a conflation of research publications with research data.


DRIVER will not have a coherent programme unless it clearly and systematically de-conflates OA-provision from digital preservation, primary publications from authors' supplementary postprints, and publication-archiving from data-archiving, treating each of these separately, on its own respective terms.


"experiments on and wide deployment of scientific data infrastructures with cross-border, cross-institution and cross-discipline added-value for open access to and preservation of scientific information"


This again conflates OA provision with digital preservation and conflates publications with data. It also conflates both of these with IR interoperability, which is yet another matter. (And webwide OA is, by definition, cross-institution, cross-border and cross-discipline, so that is a non-issue.)


What is an issue, however, is institutional versus central depositing, and it is crucial that DRIVER have a clear, coherent policy (insofar as research archiving is concerned -- this does not necessarily apply to other forms of digital content): Deposit Institutionally: Harvest/Index/Search Centrally.


The emphasis of DRIVER should accordingly be on ensuring that the distributed IRs have the requisite interoperability for whatever central harvesting, indexing, search and analysis are needed and desired.


"promoting, through these policies, access through the internet to the results of publicly financed research, at no cost to the reader, taking into consideration economically sustainable ways of doing this, including delayed open access"


Economic sustainability is again a red herring introduced by the publishing lobby into language that should only concern the research community and research access. The economic sustainability of publishing is not DRIVER's concern.


DRIVER's concern should be interoperable OA-provision (plus whatever cultural-heritage and other forms of archiving DRIVER wishes to provide the infrastructure for).


Nor are publisher access-embargoes DRIVER's concern: DRIVER should merely help ensure immediate deposit in IRs, and it should facilitate research usage needs through IR interoperability as well as the IRs' email eprint request button.


"2008 working towards the interoperability of national repositories of scientific information in order to facilitate accessibility and searchability of scientific information beyond national borders"


Insofar as research is concerned, it is not the interoperability of national repositories that is crucial but the interoperability of all OA IRs.


"2009 contributing to an effective overview of progress at European level, informing the Commission of results and experiences with alternative models for the dissemination of scientific information."


This is again a red herring (for both the EU and for DRIVER) introduced by the publishing lobby: Research archiving and OA-provision are neither a matter of alternative publishing models nor a matter of alternatives to the generic peer-reviewed publication model. Publishing reform and peer review reform are not DRIVER matters. They can and will evolve too, but DRIVER should focus on the deposit of current published research as well as research data in IRs, and the interoperability of those IRs. That is the immediate problem. The rest is merely speculative for now.


"B. Invitation to the Commission to implement the measures announced in the Communication on "scientific information in the digital age: access, dissemination and preservation", and in particular to: 1. Experiment with open access to scientific publications resulting from projects funded by the EU Research Framework Programmes by: defining and implementing concrete experiments with open access to scientific publications resulting from Community funded research, including with open access."


This is a vague way of saying that the publishing lobby has persuaded the EU not to do the obvious, but to keep on 'experimenting' as if what needed to be done were not already evident, already tested, already demonstrated to work, and already being done, worldwide (including by RCUK, ERC, NIH, and over a dozen universities, and now recommended for all European universities by EUA):


The EU should mandate that all EU-funded research articles (postprints) are deposited in the fundee's IR immediately upon acceptance for publication. Access can be set in compliance with embargoes, if desired. And data-archiving should be strongly encouraged. DRIVER's concern should be with ensuring that the network of IRs has the requisite interoperability to make it maximally useful and useable for further research progress.


The Immediate-Deposit/Optional Access (ID/OA) Mandate: Rationale and Model


Optimizing OA Self-Archiving Mandates: What? Where? When? Why? How?