Date: Thu, 27 Feb 2003 15:21:09 +0000 (GMT)
From: Stevan Harnad <harnad@ecs.soton.ac.uk>
Subject: Re: ePrint Repositories

On Thu, 27 Feb 2003, [identity deleted] wrote:

 > [T]he latest report from the RSLG http://www.rslg.ac.uk/final/final.pdf
 > makes interesting reading and it is clear that at least
 > some interest groups intend to use institutional repositories not just to
 > increase access but also to bypass publishers altogether, or have I
 > misinterpreted?

Based on the often contradictory and incoherent plans that are being
voiced about institutional repositories these days -- particularly the talk
about bypassing publishers -- I can entirely understand your confusion!
(It is not you who have misinterpreted, but some of the repository
enthusiasts who have got it a bit garbled, because they have not thought
it through carefully themselves.)

Fortunately, it can all be clarified.

(1) Please bear in mind that the idea of institutional repositories is
very new, so although there is a lot of talk these days about the
*software* for creating these archives, there has been far less talk
(and even less thought) about what sort of *content* these repositories
should house, and why, and how.

(2) Despite the lack of forethought (and often also a surprising lack
of information), universities have pushed ahead with the momentum for
institutional repositories (because they are basically a very good
thing). I think I know quite well by now what the five main notions
are that are churning around in administrators' and librarians' minds
in this connection, and not all of them make sense, nor are they all
compatible, or even desirable:

     (2a) Preservation of Digital Content: This is the most general,
     hence the vaguest mandate of all. (What *kinds* of digital
     content? Whose? Why?)

     (2b) Publication, Alternative Publication, and Alternatives to
     Publication: This is in some ways the wackiest notion of all,
     and a microcosm of the incoherence I spoke of. Universities have
     the simultaneous desire (i) to become online, in-house publishers
     themselves (if there is money to be made that way), (ii) to provide
     alternative "forms" of publication (alternative forms of peer review,
     for example -- invariably untested and speculative, if not contentious
     ones), and (iii) to provide means for making some of their own output
     public in forms other than formal publication.

     (2c) Remedying the Serials Crisis: The serials crisis is real,
     but the notion that since it is our institution that generates our
     published content, we should not need to buy it back from publishers
     is incoherent (since universities are mostly buying in *other*
     universities' published content, not their own). Self-archiving our
     *own* refereed journal publications does make sense, but in and of
     itself it has nothing directly to do with the buy-in problem.

     (2d) Courseware: There is also the sense that these repositories
     could house the university's growing online courseware content,
     either to make it open-access or to cash in on it (e.g.,
     through toll-access distance-education revenue).

     (2e) Maximizing the Research Impact of Institutional Research Output
     by Making it Open-Access (pre- and post peer review and publication).
     (This is really the only coherent, focused, motivated  agenda so far
     for Institutional Repositories. The rest is just a hodge-podge.)

The last of these (2e) is the only institutional repository agenda that is
pertinent to this Forum. It is irrelevant that some universities have
vague, wide-eyed notions of becoming in-house online publishers at the
moment: For the published papers in the institutional Eprint Archives
for their own refereed research output are *by definition* published
by *another publisher* and have nothing whatsoever to do with the
institution's inchoate yearning to become an online publisher! Each paper
in question is published by whichever of the 20,000 peer-reviewed journals
it appeared in. The only content in a university's Eprint Archive is its
*own* research output. This neither constitutes a university publication
(merely a means of providing open-access to its published output) nor
does it repackage the contents of a publisher's journal: Any University
Eprint Archive holds only its own vanishingly small contribution to the
contents of any of the 20,000 journals.

Stevan Harnad