Re: Central versus institutional self-archiving

From: Stevan Harnad <>
Date: Mon, 9 May 2005 17:10:45 +0100

In the British Medical Journal (BMJ) Tony Delamothe has written a
useful and welcome report about important progress in the Open Access
(OA) self-archiving of the UK medical research literature.

    "Initiative could give free access to UK medical research"
    BMJ 2005;330:1043 (7 May), doi:10.1136/bmj.330.7499.1043-a

The report contains one ambiguity, however, that in fact turns on what
is (in my own view) an extremely important strategic error about *where*
articles should be self-archived. (Optimal policy: in either the author's
own institutional repository or a central repository like PubMed Central,
but *preferably* the former, subsequently harvested by the latter.)

> the consortium [led by the Wellcome Trust] will... set up a UK "mirror"
> of PubMed Central, the free online archive of life science literature
> administered by the US National Library of Medicine... to allow the
> ingestion of [UK] peer reviewed articles arising from research funded
> by the consortium partners.

It is fine to set up more OA archives to "ingest" OA articles, but where
should authors, from all disciplines, self-archive? They don't all do
medical research, they don't all have central funders, and they don't all
have central archives. What they all do have is their own universities
(or research institutions), which employ them, and are in a position to
mandate, monitor, reward and co-benefit (along with their researchers)
from the self-archiving of their own institutional research output,
in their own Institutional Repository (IR). And an increasing number
of these universities and research institutions are currently setting
up their own Institutional Repositories (IRs), which cover their own
research output across all their disciplines (and, distributing the
self-archiving load across all institutions, are an incomparably less
expensive proposition than central archives).

The most general and natural way to self-archive all research output,
in all disciplines, is to self-archive it in the researcher's own
Institutional Repository (IR). All the IRs are OAI-compliant, hence
completely interoperable. Their metadata can be harvested centrally so
the contents of all IRs can be seamlessly searched as if they were all
in just one global archive:

Central, subject-specific archives are of course OAI-compliant too,
hence also part of the global OAI virtual archive. This means that in
principle it doesn't matter where an article is self-archived -- in the
author's own IR, or a central archive like PMC, or both.

But whereas it does not matter where articles are self-archived, it
does matter that most articles (85%) published annually today are *not*
self-archived at all:

Hence for a self-archiving *requirement* by a research funder to have
maximum capacity to generate self-archiving, it should be preferentially
aimed at the most general means of self-archiving, the one that is
applicable to all disciplines at all institutions, rather than only to
the central archive of one research funder's subject-matter.

Research funders can mandate self-archiving. So can research
institutions. But to maximally encourage institutions to mandate
self-archiving of *all* their research output, what better way do
research funders have than to mandate that their fundees self-archive
their research preferentially in their own institutional IRs?

For this reason (and many others), the specific recommendation of the
JISC report on UK self-archiving was to self-archive institutionally
and then harvest centrally:

    Swan, Alma; Needham, Paul; Probets, Steve; Muir, Adrienne;
    Oppenheim, Charles; O'Brien, Ann; Hardy, Rachel; Rowland,
    Fytton; & Brown, Sheridan (2005) Developing a model for e-prints
    and open access journal content in UK further and higher education.
    Learned Publishing 18(1): 25-40.

    Swan, Alma; Needham, Paul; Probets, Steve; Muir, Adrienne;
    O'Brien, Ann; Oppenheim, Charles; Hardy, Rachel; & Rowland,
    Fytton (2005) Delivery, Management and Access Model for E-prints and
    Open Access Journals within Further and Higher Education. JISC Report.

> The Wellcome Trust has already announced that it is making deposition of the
> author's final accepted (peer reviewed) manuscript in an open access archive a
> condition of funding, and the Research Councils UK looks set to follow their lead
> (BMJ 2005;330: 923[Free Full Text], 23 Apr). A study commissioned by a committee
> of the UK's further and higher education funding bodies found that only 3% of
> authors would not comply with such a request from their funders.

But the Wellcome Trust (like NIH) specified that the deposit should be in
a central archive (PubMed Central or a European counterpart).

In contrast, the respondents' very high levels of willing compliance in
that (JISC) commissioned study (conducted and written by the very same
primary authors as the other JISC study on central versus institutional
self-archiving) was part of the rationale for recommending institutional
rather than central self-archiving.

Moreover, the researchers surveyed had reported that they would willingly
comply with a self-archiving requirement from their funders *or their
[institutional] employers*. (The 2004 study actually projected 69% willing
compliance, 27% reluctant compliance and 4% non-compliance, but has now
been updated with a much bigger sample to 81% willing compliance, 14%
reluctant compliance and 5% non-compliance.)

    Swan, Alma and Brown, Sheridan (2004) Authors and open
    access publishing. Learned Publishing 17(3):pp. 219-224.

    Swan, Alma and Brown, Sheridan (2004) JISC/OSI JOURNAL AUTHORS SURVEY
    Report. In JISC Report=20

    Latest version:

Moreover, the forthcoming UK Research Council policy recommendation is
following the lead of the UK Select Committee -- which recommended
institutional self-archiving -- not the lead of NIH/Wellcome, which
recommended central self-archiving.

> By making deposition of the final manuscript a condition of funding, UK funders
> are going beyond the situation in the United States. In a climbdown from its
> initial proposals, the US National Institutes of Health is requesting, rather than
> mandating, its grantees to make the final version of their papers available for
> public display in PubMed Central within a year of publication.

Yes, the UK medical consortium funder policy is preferable to the US NIH
policy because it requires rather than merely requests self-archiving, but
it is far from optimal unless it requires *immediate* self-archiving (rather
than 6-12-month embargoed self-archiving) and unless it preferentially recommends
institutional rather than central self-archiving (with subsequent central
harvesting if desired).

Stevan Harnad

 Prior AmSci Topic Threads:

    "Central vs. Distributed Archives" (1999)

    "PubMed and self-archiving" (2003)

    "Central versus institutional self-archiving" (2003)

    "A Simple Way to Optimize the NIH Public Access Policy" (2004)

A complete Hypermail archive of the ongoing discussion of providing
open access to the peer-reviewed research literature online (1998-2005)
is available at:
        To join or leave the Forum or change your subscription address:
        Post discussion to:

UNIVERSITIES: If you have adopted or plan to adopt an institutional
policy of providing Open Access to your own research article output,
please describe your policy at:

    BOAI-1 ("green"): Publish your article in a suitable toll-access journal
    BOAI-2 ("gold"): Publish your article in a open-access journal if/when
            a suitable one exists.
    in BOTH cases self-archive a supplementary version of your article
            in your institutional repository.
Received on Mon May 09 2005 - 17:10:45 BST

This archive was generated by hypermail 2.3.0 : Fri Dec 10 2010 - 19:47:52 GMT