Re: Central versus institutional self-archiving

From: Stevan Harnad <harnad_at_ecs.soton.ac.uk>
Date: Thu, 5 Oct 2006 23:44:00 +0100

On Thu, 5 Oct 2006, Andrew A. Adams wrote:

> The other thing that ArXiv does, as well as performing as a CR for
> peer-reviewed work, is to provide a swift feedback route for work in
> progress.

Distributed IRs, 100% full of their own institutional output,
and harvested by OAIster and other search engines, both global and
discipline-based (including Arxiv, which by then will be a harvester, not
a direct locus of deposit) can then do the same job for all disciplines.

Mandate self-archiving of postprints in each author's IR immediately
upon acceptance for publication (with the self-archiving of preprints
encouraged but optional) and you'll have it all.

> Speed: presently it can take years for a major peer-reviewed article to
> appear.

Mandate self-archiving of postprints in each author's IR immediately
upon acceptance for publication (with the self-archiving of preprints
encouraged but optional) and you'll have it all.

> So, what we have with ArXiv is a place where, in a parallel strand to the
> formal peer-review process, readers are willing to take the time to quickly
> skim through pre-print work... There are a number of
> stories (a number of them validated) of highly senior researchers, including
> some Nobel laureates even, who have been engaged enough in the community of
> ArXiv, to send comments, encouragement [on] work they have seen on ArXiv and been
> impressed by.

The two critical questions to ask here (and everyone always forgets to ask them, and
forgets that Arxiv today definitely does not and cannot answer them) are:

(1) Will this scale? Can all or most work expect to find qualified peers who
will find and take the time to provide feedback?

(2) More fundamentally: Is this meant as a *substitute* for peer review
(in which case question (1) becomes even more critical) or merely an
occasional *supplement* to it (in which case mandated self-archiving
of classically peer-reviewed postprints plus optional self-archiving of
unrefereed preprints are perfect complements)?

> To conclude. While we push our way towards universal OA via the IR route, it
> is in our best interests (both philosophically and pragmatically) to ensure
> that we find ways to retain the benefits of the communities such as the
> ArXiv, rather than ignoring these separate benefits.

Why would mandated institutional self-archiving of peer-reviewed postprints lose
any of Arxiv's benefits? Virtually every preprint in Arxiv is eventually
peer-reviewed and becomes a postprint. The extra layer of feedback (for the
preprints that actually manage to elicit feedback) is a bonus, but it is a bonus
that is just as available (or unavailable) for preprints self-archived in their
own institutional IR as for preprints self-archived in a CR like Arxiv.

So there is no either/or here. What is needed is 100% of OA's target
content (the annual 2.5 million peer-reviewed postprint published in the
planet's 24,000 peer-reviewed journals and conference proceedings). The
way to get that content is for its authors' institutions and funders
to mandate that it must be self-archived, to maximize its usage and
impact. And the natural place to self-archive it is in the author's
institution's own IR: Any disciplinary or multidisciplinary or national
collections can then be harvested therefrom.

But no one can mandate the self-archiving of unrefereed preprints; that must
remain the author's choice. The IRs will merely offer them that option.

Stevan Harnad
Received on Fri Oct 06 2006 - 00:40:13 BST