Central versus institutional self-archiving
harnad at ecs.soton.ac.uk
Mon Aug 9 08:29:24 EST 2004
To minimize multiple postings, I have combined three postings from
(1) Jan Velterop, (2) Thomas Krichel, (3) David Goodman on
the same topic thread, plus my own replies to the first two of them. -- SH
(1) Date: Mon, 9 Aug 2004 09:47:26 +0100
From: Jan Velterop <velterop at biomedcentral.com>
> On Sun, 8 Aug 2004, Stevan Harnad wrote:
> Institutional self-archiving is the more powerful and general
> strategy [than central archiving].
Why would this be so? (I'm sure you explained this before, quite possibly
many times, but given the volume of your output it takes me rather too long
to find it.) And what about funder-central archiving?
SH: Given the volume of my output, it would induce chronometric
explosion if I tried to recap it all every time I was asked!
Please see the recent threads starting:
"Central versus institutional self-archiving" (Aug 8)
"AAU misinterprets House Appropriations
Committee Recommendation" (Aug 3)
"Re: Mandating OA around the corner?"
SUMMARY: The US Congress self-archiving mandate need to be amended
so as not to stipulate central self-archiving (e.g. PMC) as now:
It should either not stipulate where to archive at all (except that
the archive must be OAI-complaint) or it should preferentially
recommend institutional self-archiving (as the UK Committee's
recommended mandate did) for full-text, with PMC harvesting
the metadata and only a backup (where needed) for the full-text.
The 3 main reasons for preferring institutional self-archiving are:
(1) A mandate to self-archive specifically in PMC will only affect
biomedical research funded by NIH whereas a mandate to self-archive in
the author's own institutional archive will generalize and propagate
the practise and policy of self-archiving across institutions and
their disciplines. http://www.eprints.org/signup/sign.php
(2) OAI-compliance makes all OAI-compliant OA Archives completely
equivalent and interoperable, so PMC can harvest all the metadata
anyway; there is no reason whatsoever why it also needs to house all
the full-texts too (except as backup). http://www.openarchives.org/
(3) Many of the 8000+ journals surveyed, of which 84% have given
their green light to author-institution self-archiving, have balked at
giving it also to self-archiving in 3rd-party archives. So mandating
3rd-party archiving raises further needless obstacles (even though
the distinction is silly, and this is probably only a minor obstacle).
(2) Date: Mon, 9 Aug 2004 10:47:14 +0700
From: Thomas Krichel <krichel at openlib.org>
Richard Durbin writes
>rd> I believe that central open-access archiving is far superior to
>rd> distributed open access archiving.
This is what Stevan said years ago. I even had a shouting match with
him in front of our funders about this at the time. He has changed his
The truth remains: Some communities (Economics, Computing) have
distributed archiving, and some of the features of these distributed
systems are superior to what is found in central systems. In some disputed
subject matters, such as the ones in the humanities, I can not see how
central archiving will ever work. Some will say Derida is a philospher,
others will say he is a charlatan. Who will decide if his work is admitted
to a central archive?
>rd> The biological community is well on the way towards central archiving.
Who will be doing the central archiving? How will it be funded?
My answer to this debate is that there is no answer. Different scholarly
communities will come up with different ways to publish. Some will
publish in open access, some in closed access, other in central system
(say Physics) others in distributed system.
What is probably least likely to work is the "the provost will beat
the shit out of the academic if (s)he does not upload her work in the
University archive" approach, that seems to be favoured by some. But even
this could work in certain countries with a national research assessment,
with financial penalites for non-performers.
Thomas Krichel mailto:krichel at openlib.org
visiting CO PAH, Novosibirsk http://openlib.org/home/krichel
SH: Thomas is quite right that I have changed my mind -- in response
to new developments, new evidence, new arguments -- on several
I originally advocated distributed, institutional
then recommended instead central self-archiving -- in ArXiv
itself, for all disciplines -- because of the growth of ArXiv.
I even founded CogPrints, a central archive in the cognitive sciences,
explicitly intended for eventual centralized subsumption by ArXiv (1997).
But people still hardly self-archived, and ArXiv kept growing only
linearly, and in its own areas, never accelerating, never spreading
Then came the OAI metadata-harvesting protocol, which made all
OAI-compliant archives equivalent and interoperable (1999).
It was only then (and certainly not because of the shouting match
with Thomas Krichel, which predated OAI and was not based on any
substantive argument or evidence against central self-archiving!) that
I realized that distributed institutional self-archiving had now,
because of OAI interoperability, become the more powerful
and general route to (what was yet to be called) OA (2001).
It remained only to note that spontaneous self-archiving was still not
happening anywhere near fast enough, hence that (1) free institutional
OAI Archive-creating software (2002),
(2) the systematic gathering of empirical evidence to
demonstrate the impact-enhancing power of OA (2004),
and (3) an institutional and research-funder mandate
(in the form of a generalization of the existing "publish or
would be necessary in order to accelerate
self-archiving toward 100% OA within polynomial time!.
(3) Date: Sun, 8 Aug 2004 16:32:00 -0400
From: David Goodman <David.Goodman at liu.edu>
I agree with Stevan that the best overall position about choosing at the
present time between central or institutional self-archive locations
should be "Both are needed, helpful and welcome," and that the main
reason should be which variant "has far greater potential to grow and
generalize across disciplines and institutions to generate 100% OA."
I am less sure that the balance for that is *always* institutional
archives: some institutions will be (and are) ready adopters, and some
will not. Some will be trustworthy adopters, and some will not. Of
course, the same is true about each discipline: some have progressed
very much further than others. Perhaps the potential reach of a decision
for a single discipline in *some* cases may be greater that of many
I have no interest in persuading those who favor one form of archives
of my perhaps different view, and see no reason for such arguments in
either direction. Such disputes are negative in their effect on the goal.
I have, and many have, various disagreement in details for both the US
and the UK schemes. I have no intention of discussing them now: once the
principal issue has been achieved, we can return to more detailed planning
and improvement. If we engage in long arguments over the definitive
solution (as one academic organization has recently suggested), it will be
many years lost. This does give one good reason for favoring regulations
which--while they achieve the principal end--are not excessively specific
about the details.
For the alternate "gold" OA Journal strategy, my analysis is the same as
Stevan's: It will be so much slower that, however potentially excellent,
it may not be wise to be emphasize it at the present. Perhaps its time
will come once the basic idea of OA gets established. But here again,
since there are organizations ready to adopt this now--and even perhaps
some publishers--there are sufficient people of different talents working
on OA that those who think they can work best on this track should do so.
All the results will complement each other. They will also serve to
answer those critics who find one or another path objectionable: there
is also another way.
It sounds silly but is apparently necessary to restate the obvious:
each individual person should work with what that person can do best.
Some can work better to form and improve institutional archives, some
disciplinary. Some can work on establishing archives; some on improving
access to them; some can work even on OA journals. Some work better with
authors, some with publishers, some with librarians, some with those
running and developing the growing archives, and even perhaps some with
legislators and regulatory agencies. .
I shall do what I feel I can do best. I intend to argue about it with
those who do not yet see the merits of OA, rather than with other friends
of OA who may see things somewhat differently. At the present time,
I think it especially necessary for those interested in OA to act as
a united front to ensure the proposed regulations are adopted in their
Dr. David Goodman
Palmer School of Library and Information Science
Long Island University
dgoodman at liu.edu
Executive Committee, COUNTER
(formerly, Princeton Univ. Library)
"Unfortunately, I do not know when we will have 100% OA.
Fortunately, I shall soon find out."
More information about the Jrnlnote