Re: OA IRs are not peer-reviewed publications: they are access-providers

From: Stevan Harnad <>
Date: Tue, 21 Feb 2006 14:56:43 -0500

On 21-Feb-06, at 2:01 PM, JJ Jacobson wrote:

>> SH: (1) The right way to make the distinction between published,
>> peer-reviewed
>> material and unpublished material is the classical way: by
>> *tagging* it as such.
> We're having a conversation about just such "tagging" for the
> Georgia Tech IR. We have a whole category of phenomena (articles in
> peer-reviewed journals, invited participation in various kinds of
> symposia & so on, submissions to juried events) and are trying to
> decide how to identify it. Would someone be so kind as to send me
> a link to an example of this kind of tagging in another repository,
> preferably one that uses qualified DC? Extra points if the
> repository uses dspace.

You may wish to look at the default tagging options for the Eprints

But let me add a suggestion: There is no point over-tagging. An IR is
an access-provider. It is not a publication and it is not a library
catalogue either. (The publications to which the IR provides access
are already published and catalogued, elsewhere, where they belong!)

The three tags that are genuinely functional, and that you definitely
should use are:

(1) Published vs. In Press vs. Submitted vs. Unpublished
(2) Refereed (peer-reviewed) vs. Unrefereed
(3) Journal article vs. Conference Proceeding article vs. Book
Chapter vs. Book vs. Unpublished draft (preprint) or OTHER

Optional unpacking for OTHER (add what you feel you need):

(4) Oral presentation, conference poster, magazine article, PPT-like
presentation, multimedia etc. etc.

The thing to avoid is either to imagine that the IR is a journal,
doing quality control (leave that to the peer-reviewers!) or that the
IR is trying to do an exhaustive classification of publishing and
nonpublishing document categories (there is no need).

What there is need for is a clear, simple, useful tagging scheme --
one that does not leave the depositor needlessly agonizing over
superfluous options. If your institution or research funder does have
a standardized taxonomy of publication types -- for CV-creation,
grant-application and performance review -- by all means use that,
because then the IR can be used to help generate CVs and grant
applications automatically, and to do institution-internal audits and
evaluations. But don't let it become an obstacle or a source of
needless complexity.

And definitely do not bother with a subject taxonomy! Users will be
searching externally, and they will be searching the contents of your
IR, harvested along with the contents of many other IRs. Search will
be boolean full-text (as in google). If you wish, let depositors
insert free-text keywords in the keyword box on deposit but
definitely do not make them jump through the useless hoop of
classifying their document in a prefabricated subject taxonomy! They
should merely specify their department (and possibly other
institutional descriptors, such as research group, lab, center,
institute) but no subject terms. It is a complete waste of time and a
deterrent to attracting and keeping depositors.

Stevan Harnad
American Scientist Open Access Forum
Received on Tue Feb 21 2006 - 21:04:35 GMT

This archive was generated by hypermail 2.3.0 : Fri Dec 10 2010 - 19:48:13 GMT