Re: Poynder Again on Point on Institutional Repositories

From: Stevan Harnad <>
Date: Thu, 9 Mar 2006 02:27:17 +0000

On Wed, 8 Mar 2006, Andy Powell wrote:

> I may be missing the point here, but if Richard is correct in saying
> that the immediate goal of 'open access' is being missed because of a
> diversion into broader 'institutional repository' issues and, further,
> that the OAI protocol itself is part of the 'problem' by allowing
> repositories to expose only metadata rather than full-text (and that
> support for the OAI-PMH hasn't resulted in the development of compelling
> 'eprint' discovery services anyway), then perhaps his conclusions don't
> go far enough? Why talk about repositories at all? Why not simply say
> that all scholarly output should be made available on the Web?

No, Richard is not faulting the OAI protocol for the low level of OA.
The fault of course is with the only ones who can provide OA, namely,
authors and their institutions. The OAI protocol is merely a way -- an
excellent way -- to increase the functionality of those articles, once
they are made OA.

The problem Richard was pointing out was that Institutional Repositories
(IRs) are being created for reasons other than to be filled with OA
content; instead of working to get the IRs filled with OA content, the
focus of many IRs is the preservation and curation of generic digital
contents of all kinds. While the urgent and primary problem of drawing
in OA content is neglected, IRs, hovering at the 15% spontaneous
self-archiving baseline, are certainly not going to be "compelling
'eprint' discovery services": 85% of their target content is missing!

But once Institutions adopt policies that mandate OA self-archiving,
their IRs will fill (as those few that have already mandated it have),
and then OAI services *will* become "compelling 'eprint' discovery
services" -- especially with their OAI interoperability and
functionality augmented by google's inverted full-text indexing, boolean
search, and citation ranking.

And, no, what's missing isn't a backward step from self-archiving in
OAI-compliant IRs to self-archiving on arbitrary websites; what's missing
is institutional and research funder policy mandating self-archiving. As
95% of researchers in two JISC international surveys have indicated,
if/when it is mandated, they will comply; and the four institutions that
have already mandated it have confirmed that this is the case.

That's why the long overdue RCUK self-archiving mandate is awaited so
avidly by the OA movement.

> That would leave institutions free to decide for themselves whether they
> reach the OA goal by developing an 'open access repository', by using a
> mainstream content management system or by simply encouraging academics
> to do whatever they do now to make their material available on the Web.

All of these things are irrelevant, since the problem is the 85% missing
OA content, not the IR system. Nor is there any reason to throw away OA
functionality. What is needed is to throw *in* a mandate. What will come
out of it is the missing OA content. The parting of ways Richard
suggested was between those that want IRs to be used for generic digital
preservation/curation and those who feel that the urgent immediate
priority is not that at all, but generating the missing 85% OA content.

> Note that I'm not necessarily endorsing this conclusion, just noting
> that it seems (to me) to follow more logically from the arguments he
> puts forward. To my mind, the language we use ("repository") is just as
> much a part of the problem as anything else.

It's nothing to do with how we use language; it's how we use IRs, and
institutional and funder policies, and our fingers.

Stevan Harnad
Received on Thu Mar 09 2006 - 13:12:59 GMT

This archive was generated by hypermail 2.3.0 : Fri Dec 10 2010 - 19:48:14 GMT