Re: Central versus institutional self-archiving

From: David Goodman <David.Goodman_at_LIU.EDU>
Date: Mon, 4 Oct 2004 00:21:14 -0400

Dear Stevan,

In the Brody et al. studies, the effect of OA in enhancing visibility and use for many of the earlier papers studied, lasts for much longer than two years. I refer you to your own group's data and Brody's graphs.

The citation half life for almost all journals even in the fastest-moving fields is considerably more than 2 years. According to JCR, the average citation half life of the 10 highest impact factor journals in nuclear physics (excluding review journals) is 5.8 years; in neurosciences it is 4.64.
Some publishers do only charge for recent material. Some do just the opposite. An example is Elsevier Web Editions -- a paper subscriber gets free electronic access to the most recent 12 months only; if it wants more, it must purchase at considerable extra cost the Science Direct or Elsevier Electronic Subscription versions. This publisher clearly considers the data of value for an extended period.

The generalization that only the most recent year or two matters is thus proven false from three independent lines of data. It does not rule out the speculation that it might be correct in some instances for some users,


Dr. David Goodman
Associate Professor
Palmer School of Library and Information Science
Long Island University

-----Original Message-----
From: American Scientist Open Access Forum on behalf of Stevan Harnad
Sent: Sun 10/3/2004 1:03 PM
Subject: Re: Central versus institutional self-archiving

On Sat, 2 Oct 2004, David Goodman wrote:

> Doesn't it depend on the institution: in particular upon the
> institution's reliability, its commitment to self-archiving and OA in
> general, and its general orientation towards digital access and
> preservation?

What is the "it" in question? The desirability and benefits of self-archiving?
The greater probability of self-archiving propagating across disciplines and
institutions with an institutional self-archiving mandate rather than a central
self-archiving mandate?

Sure the institution has to cooperate, but that cooperation is
mandatory under a mandate (if the institution wants to receive the
research funding!) And a department or even an individual can set up an
institutional OAI Eprints Archive.

And "preservation" is the biggest and most persistent
(well-preserved!) red herring that besets the sluggish progress
of self-archiving. How many times does it need to be repeated that
self-archiving is a *supplement* to the publisher's version, provided in
order to provide immediate maximized access and maximized impact? The
preservation problem concerns the publisher's official version, not
the self-archived supplements (even though the latter manage to preserve
themselves quite well too, thank you very much!)

The growth region in many fields is the first 6 months to two years
from publication. That is when results are used, applied, built-upon,
cited. Publishers are well aware of this, and it is for this reason
that there is far more willingness to agree to provide access after an
embargo period of 6 months to two years.

    "Shulenburger on open access: so NEAR and yet so far"

That, however, is not OA, which must be immediate. OA self-archiving
provides that immediate OA. Now why are we beclouding that clear, huge
benefit by worrying needlessly about how long the self-archived version
will be preserved? To a first approximation, by the time the author and
his institution stop caring about it, the publisher will have long stopped
caring about it too. So focus preservation efforts on the publisher's
proprietary version, and by the time that preservation problem is solved,
that preserved version can be made OA too. Meanwhile, let the immediate
OA supplement, self-archived by the author, do the immediate work it was
designed to do, which is to guarantee that the research is accessible
online to all of its would-be users, regardless of whether or not their
institutions can afford to pay for the publisher's official version!

> I will certainly agree with you that every academic and research
> institution should have such responsibility and commitment; I also agree
> that if an institution does have it, then it is a satisfactory place for
> use self-archiving. However desirable, however urgent, this is not now
> the case.

No, we don't really agree, David. The commitment needs to be to
access-provision, not particularly to preservation. (Having said that,
there is no reason to worry that the self-archived supplements won't be
with us for many, many years to come anyway!)

> The proper approach is to upgrade the institutions; I
> suggest that an appropriate technique is for funding agencies to require
> that a university has such a capacity, not just for post-print
> self-archiving, but for all the other important uses of an institutional
> repository. However strong your argument for action is, it would be
> reliance upon speculation to count on it as a short term development.

I again completely disagree. What needs to be mandated is access-provision
to all peer-reviewed journal output, *not* all of the other conceivable
functions of an institutional digital repository (IR)! Please let us not
lose sight of the real goal. It is because the IR movement is headed off
in all directions that it is getting nowhere. Please let us not saddle
the focussed OA self-archiving mandate with that same indirection!

    "EPrints, DSpace or ESpace?"

> For independent scholars, for whom OA is a real boon as users, there is
> no alternative at all for them as authors but personal pages maintained
> upon personal or commercial servers--the extreme of instability.

There are plenty of friendly central OA archiving hosts such as Arxiv and
CogPrints. But independent scholars are the exception, not the rule. Let
us not take them as the pretext for adopting the suboptimal form of
self-archiving for the self-archiving mandate.

> While your arguments are correct for what should be the case, in the
> real world the institutional basis for them is lacking, at least in the
> US. (In the UK, the current proposal properly provides for the use of
> the British Library as an archive for instances not covered by an
> institutional archive; although I do not know the details, they may
> prove sufficient.) Any reliance upon institutional archives in the US
> would also need such a facility.
> A similar argument could be constructed for the use of institutional
> archives as a back up for centralized facilities; the NIH has proven of
> superb responsibility over several decades, but it too is subject to
> politics.
> What do I mean by permanent? the lifetime of papers in a self-archive
> must extend at least for the period of copyright, which is what permits
> toll access--life plus 75 years. After this period the maintenance of
> accessibility is the same problem of OA and for toll-access.

You are worried about the preservation of nonexistent chickens hatched
from unlaid eggs. And your worries, if acted upon, simply delay further
the long overdue egg-laying. Stop worrying about preserving and start worrying
about providing! That's the real problem. And once the OA eggs (supplements!)
are provided, their own existence will be the best driver to ensure preservation
of OA itself.


Stevan Harnad
Received on Mon Oct 04 2004 - 05:21:14 BST

This archive was generated by hypermail 2.3.0 : Fri Dec 10 2010 - 19:47:36 GMT