---------- Forwarded message ----------
Date: Thu, 7 Feb 2002 03:55:55 +0000
From: Stevan Harnad <harnad@COGPRINTS.SOTON.AC.UK>
Subject: Clarification of "parasitism" and copyright

I forward here the correction of a very important error
made by one of the supporters of free access:

> Date: Wed, 6 Feb 2002 18:18:04 +0000 (GMT)
> From: [Identity Removed]
>
> 3) So, I do believe that publishing will make a transition from publishers
> owning copyright -and restricting the people who can read your paper - to a
> model where they are funded not for the paper copy but for providing a
> refereeing service and a permanent, curated electronic journal archive with
> a permanent URL. (The difference between this model (largely due to Stevan
> Harnad)and Paul Ginsparg's e-Print physics archive is that Ginsparg puts up
> everything on a non-refereed basis and 'parasites' off the journal peer
> reviewing system by putting up the final refereed version of those papers
> that are accepted - in violation of the copyright agreement between the
> author and the publisher. I much favour Stevan's vision in which peer
> review is paid for to enhance signal to noise!) The OAI is I think a
> significant step in this direction (and the OAI Metatdata Harvesting
> Protocol is one way of constructing a universal search engine - but will be
> only as good as the metadata associated with the paper ...)

It is EXTREMELY important to understand clearly the following point:

Author self-archiving -- whether in a central Archive like Ginsparg's,
or distributed institutional ones like the ones I advocate -- is
"parasitic" (my term!) on journal-implemented peer review in two senses:

    (1) The pre-refereeing preprints that are self-archived are of much
    better quality than they would be if they were not all destined for
    submission to (and answerability to) journal peer review. (This is
    the "invisible hand" effect of knowledge of eventual answerability
    to peer review.)
    http://www.cogsci.soton.ac.uk/~harnad/nature2.html

    (2) The post-refereeing postprints, having been peer reviewed, are
    parasitic on the extant journal system even more directly: It is
    the JOURNALS that have paid the costs of implementing the peer
    review (although of course the peers themselves review for free).

It is because of this fundamental "parasitism" that Eprint Archives
(whether central like Ginsparg's or distributed institutionally like
ours) cannot be described as a SUBSTITUTE for the peer-reviewed journal
system. They are not substitutes; they depend on it fundamentally. For
if the peer review were removed, the quality standards and the quality
would collapse.

However, there is no copyright violation issue here! (And that is why
the Ginsparg archive, extant for over 10 years and now containing over
150,000 articles, has never been the subject of a single copyright
violation suit.) In fact, the growth of the practice of self-archiving
by authors has driven a change in journal policy, which was vague about
this previously. More and more journals (and this lately includes even
the journals of publishers who have been criticized for overpricing,
such as Reed-Elsevier) have now explicitly changed their copyright
transfer policies to accommodate author self-archiving: The Elsevier
journals, for example, and Nature, no longer object to the author
self-archiving the pre-refereeing PREPRINT (this was not, in fact,
a copyright matter, because preprints predate the submission, but it
was helpful to have the "permission" made explicit in the copyright
transfer agreement anyway, for psychological reasons).
http://www.cogsci.soton.ac.uk/~harnad/Papers/Harnad/harnad00.lancet.htm

In addition, the American Physical Society journals, the highest
quality physics journals of all, explicitly allow the author to
self-archive the peer-reviewed POSTPRINT as well (and this is
indeed a copyright matter).
http://www.cogsci.soton.ac.uk/~harnad/Temp/APS/copy_trnsfr.pdf

Many other journals will also agree to this if asked. And for those who
do not, there is always the completely legal option of self-archiving a
"corrigenda" file, linked to the self-archived preprint, specifying
what changes need to be made to upgrade it to the peer-reviewed draft.
http://www.cogsci.soton.ac.uk/~harnad/Tp/resolution.htm#Harnad/Oppenheim

So the "parasitism" is not a copyright issue. It is another issue, and
a double one: (a) How to pay the essential costs of peer review? and
(b) How NOT to pay for any MORE than the essential costs of peer review,
if that is all researchers want and need?

And here the growth in the practice of author/institutional
self-archiving can perform two functions: (i) it immediately frees
access to the entire refereed literature and (ii) it puts pressure on
journals (subscription cancellation pressure, because of competition
from the author's self-archived free version) to cut costs and downsize
to the essentials (peer review) while at the same time creating the
institutional revenues (the windfall savings from cancellations) to pay
for those essential costs, as a SERVICE, on the institution's OUTGOING
research papers, instead of as a PRODUCT: the institution's INCOMING
library serials subscriptions.

Finally, the reason I now favor institutional self-archiving over
central self-archiving is that the university is the natural entity to
drive, mediate, reward, and benefit from the transition: It is the
university and its researchers and research output that benefit from
maximising their research impact by making it freely accessible to all
would-be users by self-archiving it. It is the university and its
researchers and research that benefit from having all refereed research
from other universities freely accessible to its researchers (something
its library serials budget could never have afforded) and it is the
university that stands to gain from the annual windfall savings from
serials cancellations, only a portion of which (~10-30%, or $200-$500
per paper) will need to be re-directed to cover peer review costs per
outgoing paper, once the journals have downsized to the essentials.

To bring this about, the universities have to become very activist in
institutional self-archiving, helping their researchers to self-archive,
and providing and maintaining the Eprint archives for them to
self-archive in. As the above commentator indicates,
OAI-interoperability will then take care of the rest.

With this clarification, we are in complete agreement. (It is an
empirical question, on which not much depends, whether it will prove
optimal for journals to retain a role in archiving and preservation, or
whether distributed archiving, with proper mirroring and backup
aggreements, together with OAI interoperability agreements, can take
care of it best.)

Stevan