Re: Central versus institutional self-archiving

From: Stevan Harnad <>
Date: Thu, 21 Sep 2006 22:38:47 +0100

On Thu, 21 Sep 2006, David Goodman wrote:

> your idea of relying primarily on IRs requires that every institution
> mandate the deposit.

You have misunderstood. The idea is to always mandate institutional
self-archiving, never central self-archiving, whether you are the
researcher's institution or the researcher's funder. Princeton mandates
that Princeton researchers self-archive in Princeton's IR. NIH mandates
that Princeton fundees self-archive in Princeton's IR (and Harvard fundees
in Harvard's IR). PubMed Central can then harvest from the IRs of its fundees.

(Plus a few generic Central Repositories for those whose institutions don't yet have
IRs or those who don't have an institution.)

> There is no question that there are
> many IRs being established, but this is something much
> easier for an institution to do than to require deposit.

That is a non sequitur. We were talking about *where* to mandate deposit, whether you
are an institution or a funder.

If you want to talk about *whether* to mandate self-archiving, and *how*, we can talk
about that too. But it's all in the Immediate-Deposit/Optional-Access Mandate, which is
completely immune to all the putative difficulties (completely!):

> Especially in the US with the hundreds of research
> universities and other scientific research centers, it will take a lot
> of convincing, university by university.

It is easy to convince universities of the benefits of self-archiving,
but with all the provosts signing in support of the FRPAA, it sounds
as if most already know.

Now the only part left to do is to convince the provosts that if they
are convinced enough to support the FRPAA self-archiving mandate, there
is no need to wait for the FRPAA to be adopted: they can already mandate
self-archiving at their own university, using the ID/OA mandate.

> So far, not one US institution has been convinced, according to ROARMAP.

There are US universities with self-archiving *requests* in ROARMAP, but you are right
that there are none yet with self-archiving mandates (and not too many anywhere else

But I think the dominoes are poised to fall, since there is virtually
nothing separating a self-archiving policy request from an ID/OA
self-archiving mandate except realising it and going ahead and adopting
ID/OA. It's already been shown repeatedly (by Alma Swan) that it can
be done, and (by Arthur Sale) that once done, it works:

> Even world-wide, the number with a required policy stands at
> 7 (1 international laboratory, 4 universities, and 2 academic departments.)

That's correct.

> "Stands" is the word, because only 1 university has been added in 2005, and only 1 in 2006.

That's correct too. But meanwhile there have been 5 research funder
mandates, plus the FRPAA mandate proposal, the EC proposal, and now
all the provosts and university presidents endorsing the FRPAA mandate
proposal. So that looks like a lot of dominoes to me, waiting for
a breeze...

> In the 5 recent UK mandates, all established in 2006, 3 require deposit in a
> centralized repository, and 2 accept either that or an IR.

Yes, that is unfortunate, but readily remediable. And you'll notice that
several mandates and mandate proposals that were first formulated as
central ones, have since been modified to CR-or-IR. (Only the UK Select
Committee got it spot-on, with an IR-only mandate the first time:

with a little help:

> To say that "direct deposit in CRs is extremely counterproductive at
> a time when self-archiving has not yet been established as a systematic
> research imperative" is not responsive to the situation.

I'd say it was highly responsive: It's trying to reconcile institutional
IR mandates, which cover all content, and funder CR mandates, which
cover only their own subject matter. The reconciliation is that all
mandates, institutional and funder, should be IR mandates, and CRs should
harvest. That generates all OA content, across all disciplines and institutions, and
can also fill any CRs we feel we want.

> Since it has not become established as "a systematic research imperative," we
> had better rely not only on the individual institutions, but also on
> the funders, if we are to get most of the research articles to be OA. The funders
> are now the principal parties who see the advantages of requiring OA,
> and act accordingly.

And that is exactly why the UK Select Committee recommended that *funder*
mandates should be for *IR* self-archiving. And that is what several of
the RCUK funder mandates are recommending (either preferentially or as an
IR-or-CR choice); and FRPAA is recommending it (so far only as a choice)
too. But as it is optimal and scales, I have no doubt that the IR-choice
will prevail.

Stevan Harnad
Received on Thu Sep 21 2006 - 23:02:34 BST

This archive was generated by hypermail 2.3.0 : Fri Dec 10 2010 - 19:48:31 GMT