Re: DASER 2 IR Meeting and NIH Public Access Policy

[Stevan Harnad <harnad@ecs.soton.ac.uk>]

On Tue, 6 Dec 2005, David Stern wrote:

> If you had stayed through the final presentation you would have heard 
> one other suggestion that directly addresses your highest priority: 
> immediately increasing the percentage of OA material in the 
> repositories.
>
> My suggestion was to place OA materials immediately in centralized 
> repositories rather than waiting for each researcher organization to 
> mount its own Institutional Repository (IR).

David, the option of depositing in an existing OAI archive if/when the 
author has no other immediate place to deposit has been treated repeatedly 
in this Forum. And of course the commonsense strategy, highly recommended, 
is not to wait for an IR if there is not one available, but to deposit 
immediately in an existing OAI archive (and then later encourage your 
institution to create one). I have been offering offshore archiving in 
CogPrints to authors with no IR for years, for example, and Peter Suber 
has been negotiating with Brewster Kahle to make the Internet Archive 
OAI-compliant and available for overflow archiving for unaffiliated 
authors.

That all goes without saying; but lack of archives are not the problem. 
Lack of archiving is the problem. Its main symptom is a plenitude of IRs 
with a penury of content.

So whereas your suggestion -- that authors who are desperate to 
self-archive today and have no IR to do it in should do it in a central 
archive -- is welcome, it does not address the real problem today, which 
is that authors are not spontaneously desperate to self-archive. And it is 
institutional self-archiving mandates that will induce authors to do it; 
authors' own institutions are in a position to monitor and reward 
compliance; and authors' institutional IRs are also the natural place for 
research-funders to mandate that their fundees self-archive (though 
funders too could offer a central OAI archive to accommodate authors whose 
institutions don't have an IR: it can all be harvested back in due 
course).

> arXiv was a success because it had an immediate critical threshold of 
> materials in a discipline.  This would not have happened if we had 
> waited for the majority of authors to have IRs.

Arxiv was a success because the physicists were already desperate to share 
their articles, and were already doing it in paper. Arxiv happened to 
start with a central archive. This was before the days of OAI 
interoperability. If OAI had already been there, we would have offered 
Eprints as a free OAI-compliant archive-making software then, and OA IRs 
would have started in 1991. As it is, computer scientists, who have been 
self-archiving even longer than physicists, and have self-archived a 
larger total number of articles, happen to have done it on their own 
websites (and before that, their ftp sites), and all of that was then 
harvested into a "central" virtual archive, Citeseer.

That's all history today. Institutional archives and central archives are 
equivalent and interoperable, and they are not the problem: Getting 
authors to self-archive (for their own good, and the good of their 
institutions, funders, and research itself) is the problem. Central 
archiving is certainly not the solution; in fact, in the OAI-interopreable 
harvesting age, central archiving is the more primitive option. 
(Librarians tend to be more comfortable with collections located in one 
physical space, but that is not way it is done in the digital era.)

That's what I would have said at DASER if my plane's departure time had 
not prevented me from missing your talk...

> Many important research organizations still do not have IRs, and will 
> not have fully functional ones for some time for many reasons which must 
> be accepted as reality.

They can and will, if they wish; and there are no substantive reasons, 
functional or otherwise, for not having them (assuming the institutions 
have research output at all, that they are publishing, and hence should be 
self-archiving: if not, nolo contendere).

And in the meanwhile, as noted, the existing central archives can take up 
any immediate slack. But immediate slack is not the problem: the slackness 
of 85% of researchers in doing the optimal and inevitable is. And 
mandates, not central archives, are the solution.

> Yes, we can harvest the information centrally for those with IRs, but we 
> can quickly increase the possibility of mass contributions through 
> providing and emphasizing shared repositories for those without IRs.

David, you are dreaming if you think the non-archiving of 85% of the OA 
target corpus today is due to the lack of an archive to archive it in: It 
is due in (smaller) part to insufficient knowledge of the usage/impact 
benefits of self-archiving and in (larger) part to insufficient 
inclination to self-archive spontaneously (because of competing 
priorities, unfounded worries that it's time-consuming, unfounded that 
worries it's illegal, etc. etc.).

A clear-cut institutional and research-funder mandate to self-archive 
(preferably in the author's own IR) is the solution, as described in my 
proposal for fixing the flawed, failed NIH public access policy.

> We really don't need to do anything technical, as arXiv could 
> immediately add additional discipline archives.  We only need to 
> redirect authors to existing infrastructures.

You sound like me, 8 years ago: We've been there, done that, and gotten 
nowhere. The problem isn't a pressing tide of would-be self-archivers with 
no place to self-archive. The problem is to press the tide, generate the 
flow...

> Might this be a proactive and significant change in policy resulting in 
> immediate positive impact?

In a word: No. It is a symptom of being too far removed from the action to 
have gotten a clear sense of what is and is not happening, and why, and 
hence what needs to be done to get it happening.

Cheers, Stevan