The following query was received Tuesday:

> on Thursday the Wellcome Trust will announce that as of October this year,
> all new grant recipients "must" post any papers arising from Wellcome funded
> work on either PubMedCentral or the yet to be established UKPMC, "within six
> months of publication". From October next year, the same will apply to
> existing grant holders. (This information embargoed by the Wellcome until
> midnight Wednesday)
>
> First, I wonder what your view is on the announcement in general,
> particularly in relation to the NIH position?

Wellcome's policy of *requiring* self-archiving is a great improvement
over NIH's *requesting* it.

However, requiring it to be deposited in PMC or UKPMC is a big and
unnecessary strategic mistake.

What Wellcome should have required is that the articles be deposited in
each researcher's own Institutional Repository, from which it could then
be *harvested* by PMC or UKPMC.

That would have greatly increased the influence of the Wellcome policy,
touching on all disciplines and all institutions, not just the biomedical
research that Wellcome funds. It would have helped propagate a standard,
universal practice in all researchers' institutions, one that could be
followed by all researchers in all fields at all institutions.

And it would have sacrificed nothing of what the present Wellcome policy
seeks (which is that the research should also be accessible centrally
via PMC/UKPMC).

This flaw the Wellcome policy shares with the NIH policy. The second
flaw it shares is to collaborate in a 6-month embargo (the NIH embargo
is up to a year).  Both should have required *immediate* deposit upon
acceptance for publication. (Research progress is not based on 6-
or 12-months delay in access to research findings.)

In contrast to the NIH/Wellcome embargo policy -- which is not Open
Access, but merely Back Access -- the RCUK looks as if it might adopt
the optimal policy (the one recommended by the UK Select Committee last
year and already quasi-adopted by Scotland), which is to require immediate
institutional self-archiving (with central self-archiving as an option
only if the institution does not yet have an Institutional Repository).

*That* would be the policy that could serve as the take-home message
for emulation by the rest of the world research community.

> Second, I'd be interested to hear your views on the Wellcome's championing
> of a UK PubMedCentral. My understanding is that they're more positive about
> it than RCUK, which has concerns about the cost of a UK version of PMC
> versus other archive options.

The problem is not only the (needless) cost of a central archive
(although the JISC study on institutional vs. central archiving by
Swan et al., which strongly recommended institutional rather than
central archiving, followed by central *harvesting* if desired,
did cite the cost as one of the many reasons for recommending
this).

          Swan, Alma and Needham, Paul and Probets, Steve and Muir,
          Adrienne and O'Brien, Ann and Oppenheim, Charles and Hardy,
          Rachel and Rowland, Fytton (2005) Delivery, Management and
          Access Model for E-prints and Open Access Journals within
          Further and Higher Education. JISC Report.
          http://cogprints.org/4122/

          Swan, Alma and Needham, Paul and Probets, Steve and Muir,
          Adrienne and Oppenheim, Charles and O'Brien, Ann and Hardy,
          Rachel and Rowland, Fytton and Brown, Sheridan (2005) Developing
          a model for e-prints and open access journal content in UK
          further and higher education. Learned Publishing.
          http://cogprints.org/4120/

The main reason for institutional rather than central self-archiving
is generality: Virtually all researchers have institutions;
and each institutions is just a 2000-dollar server plus some
free software away from having an institutional repository
http://www.arl.org/sparc/pubs/enews/aug01.html#6
with the UK already 3rd in the world, with over 50 such repositories
(and about 190 universities and IHEs, about 70 of them research-active):
http://archives.eprints.org/eprints.php?action=browse

Even more important than the fact that institutional self-archiving
distributes the archiving load across institutions is the fact that it
covers all disciplines and it is (as Swan et al. stress) a natural part of
institutional "culture," since the institution (not some central entity)
is the research-provider: Researchers and their own institutions are
the ones that have the joint interest in -- and share the benefits of --
maximizing the usage and impact of their own research output. Research
impact is already rewarded by institutions in their hiring, promotion
and salary evaluations. It is also rewarded in the UK by the RAE
(Research Assessment Exercise) -- and, indirectly, by the Research
Councils. Hence institutions wield the carrot and stick that can require
and reward self-archiving of their own research output by all of their
own researchers across all of their disciplines.

The RCUK policy helps researchers' institutions go in that direction. The
Wellcome/NIH policy does not, or does so far less than it could --
and this, for no substantive reason whatsoever. Nothing is gained by
exclusively requiring central archiving, but a lot is lost.

> Also, where do you think this requirement leaves authors if the journal
> they publish in does not permit archiving, or not within 6 months?

There is an extremely simple and universal solution for the 8% of articles
that are published in journals that do not yet give their green light
to immediate self-archiving:
http://romeo.eprints.org/stats.php

but again, the solution only works with an immediate institutional
self-archiving requirement: For the 8% articles from non-green journals,
the author is still required to deposit the metadata (author, title,
date, journal-name, etc.) plus the full-text in the author's Institutional
Repository -- but it is up to the author whether to set access to the
full-text as "Open Access" or only "Institutional-Internal Access." The
metadata are still accessible to any would-be user webwide, and if they want
an eprint, they need merely email the author, who can email it to them.

Hence one size already fits all, 100%. And no call for any collaboration
in a 6-month delay. The 8% will shrink to 0% soon enough, once the
optimal institutional self-archiving policy is in place worldwide.

> Finally, I'm interested in your view of this decision in the context of the
> expected RCUK announcement. [It is rumoured] that it requires author
> archiving in repositories where they exist, but does not require the
> establishment of such archives.

I hope RCUK will have the good sense not to stipulate any more than they
need to: They need only require immediate self-archiving, in institutional
or central archives. They need not say anything about *not* requiring
self-archiving if a suitable archive does not exist! We're talking about
the investment of a few thousand dollars by each university for a server
and a bit of sysad time for set-up and maintenance. The OA returns on
that tiny investment, in terms of enhanced institutional research impact
and income, would vastly outweigh the cost:

    "The dollar value (in salary and grant income) of one citation
    varies from field to field, depending on the average number of
    authors, papers and citations in the field; the marginal value of one
    citation also varies with the citation range (0 to 1 being a bigger
    increment than 30 to 31, since 60% of articles are not cited at all,
    90% have 0-5 citations, and very few have more than 30 citations:
    http://www.crsc.uqam.ca/lab/chawki/classement_citations.htm ).
    A much-cited study estimated the "worth" of one citation
    (depending on field and range) in 1986 at $50-$1300:
    http://www.garfield.library.upenn.edu/essays/v11p354y1988.pdf"

    http://www.ecs.soton.ac.uk/~harnad/Temp/researchmoney.htm

See also today's posting by the CNRS's Institut Jean Nicod, the first
Institution to adopt and register a self-archiving policy. This four-year
report provides statistics on how this policy has enhanced the
Institut's visibility and impact:

    "Rapport sur la visibilite electronique de l'Institut Jean Nicod"
    http://www.ecs.soton.ac.uk/%7Eharnad/Hypermail/Amsci/4547.html
    http://www.institutnicod.org/impact/ijn_2001_2005.pdf
http://www.eprints.org/signup/fullinfo.php?inst=Institut%20Jean%20Nicod
http://archives.eprints.org/eprints.php?url=http%3A%2F%2Fjeannicod.ccsd.cnrs.fr%2F

Institutional archive-creation will take care of itself. The RCUK should
be careful not to unwittingly insert a gratuitous and self-fulfilling
opt-out clause, effectively nullifying the force of its requirement,
by essentially saying "You are 'required' to self-archive if you already
have an IR, but if not, not!".

Stevan Harnad

AMERICAN SCIENTIST OPEN ACCESS FORUM:
A complete Hypermail archive of the ongoing discussion of providing
open access to the peer-reviewed research literature online (1998-2005)
is available at:
http://www.ecs.soton.ac.uk/~harnad/Hypermail/Amsci/
        To join or leave the Forum or change your subscription address:
http://amsci-forum.amsci.org/archives/American-Scientist-Open-Access-Forum.html
        Post discussion to:
        [log in to unmask]

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:
        http://www.eprints.org/signup/sign.php

UNIFIED DUAL OPEN-ACCESS-PROVISION POLICY:
    BOAI-1 ("green"): Publish your article in a suitable toll-access journal
            http://romeo.eprints.org/
OR
    BOAI-2 ("gold"): Publish your article in a open-access journal if/when
            a suitable one exists.
            http://www.doaj.org/
AND
    in BOTH cases self-archive a supplementary version of your article
            in your institutional repository.
            http://www.eprints.org/self-faq/
            http://archives.eprints.org/