Central versus institutional self-archiving
harnad at ecs.soton.ac.uk
Fri Jun 11 03:52:36 EST 2004
Whether they realize it or not, librarians are preoccupied with
their classical problems pricing, preservation and permissions;
their understanding of the researcher's problem of immediate access and
impact fades in and out, and tends to remain uncertain at best.
The main thing to keep in mind in weighing David Goodman's comments are
the following facts:
(1) The problem is *not* that institutional OAI-compliant Eprint
Archives are too few or hard-to-create.
http://archives.eprints.org/eprints.php?action=browse Those are
non-problems. The problem is that they are still too *empty*
-- because institutions create them but do not also create institutional
OA-provision policies so as to fill them:
(2) Hence it follows -- a fortiori -- that the problem is not the
preservation of these so-far-far-too-bare cupboard contents, but their
filling, through the adoption of institutional self-archiving policies.
(3) One of the (several) things holding up the realization that it is
institutional self-archiving policies that are immediately needed, and
not something else, is the diversion of attention to non-existent
preservation problems, and ungrounded preferences for central archives
(whose many immediate disadvantages I have enumerated already: they
really *are* too few, have no entity to support them, have central
instead of distributed burdens and costs, have been growing
too slowly, do not generalize across disciplines as institutions do, have
no common stake in impact with their authors, as institutions do,
cannot mandate or monitor OA provision as institutions can, and run
needlessly afoul of publishers' phobias about 3rd-party
(4) By far the best guarantor of OA's longevity will be its rapid
growth and prevalence. Till this happens, all talk of "preservation"
is whistling in the wind.
(5) My suggestion that institutional archives could be the *sole* means
of access and preservation in the special case of new start-up OA (gold)
journals -- so as to reduce start-up costs and get a leg up on what might
eventually be the nature of all journals -- was explicitly made as a
special, tiny-minority suggestion, for new OA journal start-ups only. New
journals come and go. Their real preservation problem is the long-term
viability of the journal itself, not the storage medium. But in any
case, OA will not prevail via the slow, narrow and uncertain golden road
of OA journal creation, but via the fast, wide and sure green road of
self-archiving toll-journal articles.
So here it is implicit pricing concerns that have side-tracked David again
onto the golden road, and made him erroneously project his conclusions
onto the green road:
(6) We are not talking about OA via new journal start-ups but via the
self-archiving of articles from the existing journals. For such articles,
the self-archived OA version is a supplement, to provide access for
those would-be users whose institutions cannot afford the toll-version.
It is the toll version that has the preservation problem (and that has
nothing to do with OA). Nor has it to do with central vs. institutional
On Wed, 9 Jun 2004, David Goodman wrote:
> Prior Threads:
> "Central versus institutional self-archiving"
> "Open Access Journal Start-Ups: A Cost-Cutting Proposal"
> "Elsevier Gives Authors Green Light for Open Access Self-Archiving"
> But it is nonetheless wrong to regard Elsevier's standard as
Elsevier is not the problem: authors, their universities and their
funders are! Elsevier's self-archiving policy is fine. If their authors
and institutions would have access to and impact from 200,000 more
articles per year! Instead, David is fixated on putative inadequacies in
Elsevier's policy. It's as if the cupboard's are bare, Mom has been
given the green light to fill them with cookies, and is instead fretting
about whether they may need to be irradiated cookies to increase their
The kids are hungry, now!
> Very few universities have BOAI compliant archives now
Incorrect (and that's OAI-compliant, not BOAI...). Plenty of cupboards,
too few cookies.
> It will only delay things if the job requires as a preliminary,
> to get one's university to establish a proper archive.
That's the easy part, and has been done over and over. It's the
archive-filling policy that's needed now.
(And "proper" is a weasel-word, here, trading on spurious preservation
considerations that the existing OAI-archives allegedly lack: They lack
*contents*. The rest is a matter of *policy*: preservation is too, but
preservation policy is hardly the problem when there is nothing yet to
> The policy of the publisher concerned, which prohibits
> the use of such extra-university archives, is thus a major hindrance.
That is *zero* hindrance. The only hindrance is the lack of institutional
OA provision policy.
> Do we want results in '04? we will get them much faster if we didn't
> first have to establish several hundred archives.
To test this, how about setting a good example by filling the several
hundred institutional archives that already exist, and stand near-empty,
for lack of an OA policy?
> 2. The use of archiving will be significantly facilitated by accurate
> archives. The differences between the author's final version and
> the journals may sometimes be trivial--perhaps they are for someone
> of excellent writing skills-- but they will sometimes be very
Yes, and when great, authors can and will incorporate them. So why dwell
on this non-issue. It is not *accurate* archives that are missing, but
> 3. But Stevan also has proposed in a recent posting the use of such
> archives as the primary publication medium for a new journal. And here
> the standards are much higher.
What goes in the archive even of a virtual journal is the final,
accepted draft. End of story. (Rest is about whether or not to wrap
a copy-editing service into the peer-review/certification service. All
hypothesis and speculation about a tiny minority of new start-up
journals. Meanwhile the 95% of the corpus lies waiting...)
> 4. For this purpose, I proposed that disciplinary archives are better than
Archives are archives, not journals, whether central or institutional.
OAI-compliance is OAI-compliance, whether central or institutional.
ArXiv and CogPrints do not provide editing services or quality control:
> University archives are also unacceptable for this use...: the lack of
> commitment to maintain archives for those who have left the university
> by change of position, retirement, or death.
The cupboards are bare, the kinds are starving: We will only consider
storing irradiated cookies...
Let OA grow and it will generate the culture of preservation. Don't put
the cart before the horse. (As I have said repeatedly before, if the
physicists had thought this way in 1991, over a decades worth of access
and impact would have been lost, waiting for irradiation, and we would
not still have those OA articles today, in 2004, as we do...)
> 5. Since clearly neither can be shown to be reliable,* is there a
> solution? Yes, exactly the same one as for conventional journals. A
> permanent archive must be guaranteed by at least several national
> libraries or their equivalents. We accept nothing less from a publisher:
Ah me! So focus on the publisher's primary version and don't load the
self-archiving supplementary version with further needless pre-emptive
baggage! It's slow enough coming as it is!
And don't project the situation of the tiny minority of start-up gold
virtual journals with the overwhelming majority in which the OA version
is *not* the primary and sole version!
> *when it is a question of reliability, if some of those knowledgeable
> think it reliable and some not, the only safe conclusion is that its
> reliability has not been demonstrated.
So don't self-archive till this is sorted out to the satisfaction of all
I suggest a moratorium on all mention of matters relating to pricing,
preservation or permissions, in favor of single-minded focus on immediate
OA provision via self-archiving, now! After fifteen years of all this,
surely it's time to give that a try too, for a change!
More information about the Jrnlnote