Re: Alternative publishing models - was: Scholar's Forum: A New Model...

From: Stevan Harnad <harnad_at_COGLIT.ECS.SOTON.AC.UK>
Date: Fri, 7 May 1999 13:27:27 +0100

On Thu, 6 May 1999, Arthur Smith wrote:

> Journals will relinquish some of their current-content and distribution
> roles to things like the preprint archives, but will be taking on new
> responsibilities also in areas traditionally dominated by abstracting
> and indexing services (through interlinking, subject-focused
> searching and browsing, and the like).

My guess is that these add-ons will be better and more cheaply provided
by the community itself -- once the entire corpus is available online,
and for free.

> Critical features are the archiving itself, and searching/indexing -
> these are a lot easier to implement for a centralized site.

Or a Virtual centralized site, drawn together from distributed
interoperable sites by a Gateway like NCSTRL.

    Davis, J. R. and Lagoze, C. (1999) "NCSTRL: Design and Deployment of a
    Globally Distributed Digital Library," to appear in Journal of the
    American Society for Information Science (JASIS)

  sh> Multiple submission
  sh> (whether parallel or serial) is already the bane of the current
  sh> overloaded referee system.
  sh> Articles rejected by one journal are certainly submitted to another
  sh> (and just about everything is eventually published somewhere), but
  sh> surely once a paper is accepted ONCE by a journal, no further
  sh> refereeing is called for.
> Stevan - a very good point, and one that I believe has been neglected
> by a lot of the proposals for new kinds of peer review (like some of the
> stuff in the Caltech Scholarly Forum proposal). However, if
> it is funded by the author, then I think John's idea is quite workable -
> most authors will probably not feel the need to pay for multiple
> evaluations, and those with very limited resources could perhaps pay for
> only the barest evaluation. If we are to go to a system where the author
> pays for peer review (as you have been advocating) I think multiple
> evaluations are going to be inevitable one way or another.

Not the way I envision it: Journals, with their known brand-names, will
continue to be the quality controllers. They will continue to implement
classical peer review, to their respective standards of rigour.
Referees (i.e., us, continuing to referee for free) will continue to report
to the editors of journals, and the authors will be answerable to the
editors, as before. If a paper cannot be revised to the standards of a
given journal's referees, it will be rejected. Then, as usual, it can
be submitted to another journal for refereeing, as it is now, usually a
journal lower in the prestige hierarchy.

Here is an important point you may be misunderstanding: Because (as you
of course know), peer review is not just a yes/no stamp, but an
interactive, feedback-based corrective process that involves,
recommendations, revisions, re-refereeings, it is not something whose
OUTCOME an author can "buy." [That would not be publication
page-charges, it would be bribery and a vanity press!]

Hence page-charges will only be assessed for ACCEPTED articles.

[There could also be a smaller submission charge, for refereeing
papers, going toward page charges if the paper was accepted, but that
would first have to be tested carefully, for it is an empirical matter;
there is no need or justification for implementing untested,
speculative proposals for "improving" peer review. One could argue that
submission charges would be a good deterrent against nuisance
submissions that simply spam the overburdened referee pool on whose
services all journals will have to keep drawing. But submission charges
might also deter worthy submissions, and thereby affect overall
submission rates and quality. Paying referees is likewise speculative
(and in my opinion unrealistic).]

So the only way to implement page charges that does not tamper in any
way with classical peer review is to assess them only for accepted
papers (factoring in the costs of processing the rejected papers with
the overall cost per accepted article).

This seems to rule out multiple submissions (if the finite refereeing
pool was not reason enough). One might imagine papers being accepted in
a lower level journal, and then being submitted to a higher level
journal, but assuming that referees continue to be referees, and
standards and impact factors continue to be maintained, it may well be a
time and money saver for journals to continue to decline to referee
papers that have already been published elsewhere, as they do now.
Referees are not known for being happy to be invited by the editor of
Journal X to referee a paper that turns out to be one that they have
already refereed (and rejected) for Journal Y.

The point is that referees will continue to contribute their services
for free in order to maintain the quality of the literature. They are
not a resource to be abused or squandered, whether in the era of S/L/P
or in the era of page-charges. And journals will continue to depend on
peer review to maintain their quality, prestige and impact factor. It
will not become a market in which one wants to accept as many articles
as possible in order to increase revenues. A journal's REJECTION RATE
will continue to be one of the measures of its quality; and lowering
standards will continue to be felt through drops in impact factors and
author quality.

Stevan Harnad
Professor of Cognitive Science
Department of Electronics and phone: +44 1703 592-582
Computer Science fax: +44 1703 592-865
University of Southampton
Highfield, Southampton
Received on Wed Feb 10 1999 - 19:17:43 GMT

This archive was generated by hypermail 2.3.0 : Fri Dec 10 2010 - 19:45:30 GMT