Search   |   Back Issues   |   Author Index   |   Title Index   |   Contents

Conference Report


D-Lib Magazine
March 2005

Volume 11 Number 3

ISSN 1082-9873

The Implementation of the Berlin Declaration on Open Access

Report on the Berlin 3 Meeting Held 28 February - 1 March 2005, Southampton, UK


Stevan Harnad
Chaire de Recherche du Canada
Centre de Neuroscience de la Cognition
Université du Québec à Montréal
Montréal, Québec, Canada H3C 3P8

Red Line


Open Access (OA) means immediate, permanent, free online access to the full text of all refereed research journal articles (2.5 million articles a year, published in 24,000 refereed journals, across all disciplines, languages and nations). The Berlin Declaration on Open Access <>, launched in October 2003, has so far been signed by 55 institutions globally, including large national research institutions such as France's CNRS and Germany's Max-Planck Institutes; national Academies of Science such as those of China, India and the Netherlands; international research institutions such as CERN; and individual universities and research funding agencies around the world.

To sign the Berlin Declaration is to declare one's institution to be in favor of the abstract principle of OA, but until now it did not indicate a commitment to actually providing OA, nor did the Declaration specify the concrete steps an institution should take in order to do so. The purpose of the subsequent series of Berlin meetings has been to agree on a practical policy of OA provision that the institutions who have already signed, as well as the other research-provider institutions worldwide, could commit themselves to adopting and implementing in practice in their own institutions.

Attending the most recent meeting (Berlin 3;, hosted by the University of Southampton, UK, were delegates from institutions that have already signed the Berlin Declaration as well as delegates from institutions that have not yet signed. Among the nations represented were India, Japan, the USA, Canada, and many of the countries of Europe. Each institution, large and small, reported on its progress to date in providing or planning to provide OA. Among the most advanced institutions were the large national ones such as France's CNRS and Germany's Max-Planck Gesellschaft, and international ones such as CERN in Switzerland; but there were also many small, individual universities, some of them among the most systematic and successful in adopting and implementing OA provision policies.

CERN, one of the largest research institutions in Europe, has a self-archiving policy that has already resulted in 60% of its research being made available OA via self-archiving, and is committed to reaching 100% soon. But individual universities, or university consortia, are also making dramatic progress: Minho University in Portugal, for example, has designed a determined policy to self-archive its research output, as has the SURF/DARE consortium of the 12 principal Dutch universities. Scandinavian universities have also been very active (Sweden, Denmark, Norway, and particularly Finland), and so have a number of other universities and research institutions in Italy Germany and France.

India, after some initial momentum, has been idling somewhat, but the unflagging efforts of OA's champion in India, Professor Subbiah Arunachalam, and some incipient government-level support, may manage to re-invigorate this before long. (Arun also gave an inspiring post-conference talk on Indian Knowledge Villages, which bring Internet resources to the rural poor.) Japan, too, is poised for OA progress (as is Australia, which sent no delegate this time but has many institutions very active with regard to providing OA to research results). The delegate from Pakistan, unfortunately, had to cancel at the last minute, but we will be hearing more from that country too, thanks in part to the efforts of Dr. Alma Swan.

Dr. Swan also gave a splendid talk in the special UK session (all videos and powerpoints are now accessible at the Berlin 3 website), reporting her now-famous finding (from 2 international surveys) that although researchers are still extremely uninformed about OA, with only about 25% of them providing OA today, the vast majority (79%) state that they will self-archive their work, and do so willingly—but only if and when their institutions or their research funders require them to do it.

And that, of course, became the specific concrete policy recommendation of Berlin 3: Universities, research institutions and research funding agencies should require—as a matter of institutional policy—that their employees/fundees deposit a supplementary copy of each of their published research journal articles into their own institutional OAI-compliant repository <>. (OAI-compliant means that each article has the metadata tags—author, title, journal, year, etc.—that will make all the institutional repositories interoperable, so that their contents can be harvested, rebundled into subject collections, and seamlessly searched and accessed, as if they were all in one global virtual archive, without users having to know or concern themselves with where any article is actually located.)

This clear, simple and practical policy recommendation by Berlin 3 also turned out to be almost identical to the recommendation that had been made by the UK Parliamentary Select Committee last year. That recommendation was ultimately declined by the UK Government, but the Committee's call was obviously heard, for the British Funding Councils (RCUK) are currently formulating an OA policy of their own).

The special UK session at the Berlin 3 meeting—consisting of 3 selected speakers from the UK Workshop on Institutional Repositories that had taken place one month earler—was chaired by Catherine Jones of CCLRC, which has already gone ahead and implemented a self-archiving policy of its own, ahead of the rest of RCUK. The session made it clear that British universities were in tune with the UK Select Committee recommendation on OA, even if Lord Sainsbury was not. For example, the 20 universities in the SHERPA consortium, reported on by Bill Hubbard—not to mention Southampton University itself, the first in the UK to adopt first a departmental and then a University-wide self-archiving policy—have all established institutional archives and are working on filling them. And Derek Law of Scotland was quick to remind us that his country—which traditionally pulls more than its weight both in research output and research funding—was already ahead of the rest of the UK at a national level, having adopted an OA policy of its own while the UK Government dithered.

The US was under-represented at Berlin 3. Andy Gass attended on behalf of the Public Library of Science; Jim Pringle of the Institute for Scientific Information was also present, as was Christine Borgman of UCLA (and Oxford), but there were no reports of progress on implementing institutional OA policy in the US. However, this is somewhat deceptive, historically at least, because the 'Bethesda Statement' (June 2003) had been a direct precursor of the Berlin Declaration, and influenced the form it took [1]. In addition, there is another US development today that is OA-related, though it is not OA in the strictest sense: it is the National Institute of Health (NIH) 'public access' policy of inviting fundees to deposit their articles in PubMed Central within 12 months of publication. This is better termed 'Back Access' rather than Open Access, which can and should mean immediate access for the sake of maximizing research usage, impact and progress [2].

The following new recommendation accordingly emerged from the Berlin 3 meeting [bold emphasis added] [3]:

In order to implement the Berlin Declaration institutions should:
(1) Implement a policy to require their researchers to deposit a copy of all their published articles in an open access repository.
(2) encourage their researchers to publish their research articles in open access journals where a suitable journal exists and provide the support to enable that to happen.

This wording is still a bit rough around the edges (the word 'implement' occurs rather too often, for example), but it does finally make the Berlin Declaration into a concrete practical action plan, striking a realistic balance between green and gold [4], and remedying the vagueness inherent in the Declaration's original wording [5].

A distillation of the original Berlin Declaration follows (all italics and bracketed interpolations are mine) :

"Open access [means]:
"1. free... [online, full-text] access [to what?]
"2. A complete version of the [open-access] work [ = what?] ... is deposited... in at least one online repository... to enable open access, unrestricted distribution, [OAI] interoperability, and long-term archiving.
"We intend to... encourage... our researchers/grant recipients to publish [?] their work [?] according to the principles [?]... of the open access paradigm [?]."

If the Berlin 3 recommendations are adopted, this could then be revised as follows (italicized passages are all my inferred revisions):

"Open access [means]:
"1. immediate free... [online, full-text] access to published research articles
"2. A complete version of every research article... is deposited... in at least one online repository... to enable open access, unrestricted distribution, [OAI] interoperability, and long-term archiving.
"We intend to... (1) require... our researchers/grant recipients to self-archive all their research articles in our own institutional repository and to (2) encourage them to make them... open access."

The distinction between requiring the depositing of the metadata and full-text in the institutional repository (for institutional record-keeping and performance evaluation purposes) and encouraging the researcher to make the full-text OA is the essence of the institutional policy that many of the institutions reporting successful self-archiving practices kept reiterating at the Berlin 3 meeting: The requirement must be an administrative function, rather like requiring that annual report submissions must be in a particular format in order to be processed; for the only real obstacle to 100% OA has been getting researchers (or their proxies) to do the requisite keystrokes to get their metadata and full-texts online [6]. There is a Registry of Open Access Institutional Archives as well as a Registry of Institutions Who Have Adopted a Self-Archiving Policy.

A desire was also expressed at Berlin 3 to generalize the Berlin series, so as to make it more reflective of the international nature of the OA movement whose milestones—Budapest, Bethesda, Berlin (and now Britain) all seemed to be voiced bilabial plosives (but no new name was settled upon). The next conference (Berlin 4) will take place at the Albert-Einstein-Institute in Golm, Germany, 5-7 October 2005.

Berlin 3 closed with a brilliant demonstration on OA data-archiving in chemistry by the Crystallography Unit at the University of Southampton (eBank) building on the Keynote address on data-archiving and interoperability by Tony Hey, Director of E-science. Describing the eBank effort would require an entire report of its own and would go beyond research article self-archiving to research data self-archiving.


On behalf of all the participants, the author expresses gratitude to the DELOS Network of Excellence on Digital Libraries, the UK Joint Information Systems Committee (JISC) and the Max-Planck-Gesellschaft for funding the Berlin 3 meeting.


[1] Although that legacy was not entirely helpful, in my opinion, because it helped redirect the OA movement toward its 3-year one-sided 'gold rush' toward OA journals [author-institution pays]—the 'golden' road to OA—to the near-complete neglect of the 'green' road of OA self-archiving).

[2] In the US, unlike the UK, Congress actually adopted a stronger version of a Public Access recommendation (though, with a 6-month delay, it too was not true OA), but at the implementation stage, the funder, NIH, weakened it. Conversely, in the UK, the House of Commons Select Committee made a true OA recommendation—to require immediate self-archiving by all UK institutions in their own institutional repositories—and then the Government declined to adopt it. Fortunately, the UK research funders, the counterparts of NIH, look as if they may well go ahead and adopt it anyway. Yet there is still some lingering risk that other institutions, funders or governments might clone and adopt the flawed NIH Back-Access Policy (as the Wellcome Trust has—with the best of intentions—already done) But now that the Berlin Declaration is putting its weight behind a policy of immediate institutional self-archiving, that risk may perhaps have diminished. There is accordingly a very good chance that other US funders, universities and research institutions will instead adopt the Berlin 3 recommendation itself now.

[3] The Berlin 3 recommendation was keyed in by the collective hand of the delegates, on Andy Gass's computer, via the fingers of Robert Terry of the Wellcome Trust.

[4] Green means publishing an article in a toll-access journal but also concurrently self-archiving it in an institutional open access repository. Gold means publishing an article in an open access journal. (See Harnad, S., Brody, T., Vallieres, F., Carr, L., Hitchcock, S., Gingras, Y, Oppenheim, C., Stamerjohanns, H., & Hilf, E. (2004) The Access/Impact Problem and the Green and Gold Roads to Open Access. Serials Review 30. <>; Shorter version: The green and the gold roads to Open Access. Nature Web Focus. <>.)

[5] The original Berlin Declaration is located at <>.

[6] I have accordingly dubbed this the "Southampton Bureaucratic Keystroke Policy," and it can be understood to be an alternative to giving up on OA altogether, as the NIH policy does, with the risk of locking us into mere Back Access instead of Open Access for many years to come.


Copyright © 2005 Stevan Harnad

Top | Contents
Search | Author Index | Title Index | Back Issues
Previous Conference Report | In Brief
Home | E-mail the Editor


D-Lib Magazine Access Terms and Conditions