I don't think people who fear free access to and use of unrefereed
materials (as at LANL) in the belief that it will lead inevitably to a
degeneration of the professional literature are likely to be comforted
by reassurances about the role of "the invisible hand" in controlling
the quality of the unrefereed material. Why? Because they envision the
corruption as beginning with the use of unrefereed pubs but spreading to
the practices of peer review as well via those people who, corrupted by
such practices as those at LANL, pass this corruption on to their
students and to grad students generally, and so forth, so that peer
reviewers will inevitably degenerate in time to the supposedly dismal
level of the users of unrefereed materials. In short, the invisible hand
is prone to corruption by the LANL system, too. This sort of view needs
to be answered.
The problem with Stevan's answer is that the activity of the invisible
hand cannot by itself account for the quality of the preprints when the
members of the preprint culture know quite well that the paper
subsequently to be submitted for journal publication and peer review
need not be identical with the paper first made available through the
preprint server. What one would expect, given this awareness, is a
general tendency for the preprints to be distinctly inferior to the
later versions -- less finished both in reasoning and the use of
evidence and in carefulness in the detail of professional presentation
-- since the author seemingly has little incentive to strive for
perfection in the initial version submitted through the server because
he/she knows that plenty of time remains for imperfections in it to be
corrected.
If, however, this has not been occurring at LANL during the many years
of its operation, the question is: Why not? What accounts for this
seeming anomaly, given the supposedly inevitable degenerative tendencies
of unfettered human nature? I suggest that, in addition to the
causation exercised through the author's understanding of what lies down
the road when the paper is formally submitted, there is another causal
factor at work which is probably at least as efficient in causing the
authors to do the best they can in the preprint, namely, the fact that
they want the respect of the people they respect most, who are not the
reviewers whom a journal editor may someday assign but rather the people
who will be reading the preprint. Those peers who use the LANL system
are the people whom the author is primarily addressing, and there is
good reason for them to be in top professional form since their future
might be more influenced by the opinion of their peers there than by
the opinions that might be generated in the future by the formal
publication of the paper.
On the other hand, if those who use the preprint server do not include
the people in the field whose views are determining the direction and
content of the ongoing process of the science this sort of critical
control won't be operative, and one might expect quality to degenerate,
in that case. Actually, though, what is more likely is that people in
the field will only rarely post their material there, or do so from
other motives, since the server is not functioning as a place of primary
publication in the field if publishing there has no tendency to affect
it. As I explained in the previous message, it is a matter of fact
whether or not a given publication mechanism is in fact functioning as
an instrument of primary publication, and that fact concerns the actual
professional practices at a given time. There can be borderline and
undecidable cases, especially where the discipline is itself amorphous
or multi-directional in its tendencies, showing no clear tendencies of
development in one direction rather than another, so that it is
intrinsically unclear where primary publication is actually occurring,
if at all. A paper journal could be as questionable as a preprint
server as regards being a medium of primary publication in such cases,
and not even formal peer review at its most rigorous can establish it as
being a place of primary publication if the practices just aren't
there. There are many possibilities and variants. But the sciences
at LANL -- some of them, but not all that are listed there on the
website -- do not appear to be like that.
There could be other motives for using the server system to post papers
in an archive than the motives of those who use it as a place of primary
publication, and there is no justification for taking the a priori
denigrative stance that the other motives can all be captured by
referring to it as a kind of "vanity press" in that case. But what is
crucial here is the difference between a preprint server that provides
primary publication for a field and one that does not, and it seems
clear enough that the LANL server does provide a place of primary
publication for at least some fields, the best evidence of which is
perhaps its constant growth rate. For if the communication was
degenerating at LANL because of the sloppiness or shoddiness of the
material being made available there this would show itself in
increasingly serious misunderstandings, factionalism, manifestos,
manipulation of graduate students as pawns in the games being played by
professors, and so on, which would in turn result in a decline in use
and abandonment of it as a place of primary publication.
The question is whether we should call what happens there SUBSEQUENT to
publication via the server "peer review" or not, given that we usually
think of peer review as something that happens prior to publication
which functions as a filter for it. Note that I am not talking about
special post-publication "peer commentary" practices and the like. I am
just talking about the normal practices of those who use the LANL server
as their source of primary publication. The practice of primary
publication there as elsewhere certainly includes taking a critical
stance on what is published there, and may generate critical assessment
of it both of a private and a public nature along with subsequent
correction or defense, and of course the only kind of critical stance
that the users of the system are interested in is criticism from people
whom they regard as their peers. But is that normal practice of
assessment by one's peers enough to say that peer review happens there,
too, though with some differences from the way it works when it is a
review by someone appointed to the task later?
My own view is that a strong case can be made for saying Yes, peer
review is happening there as a matter of course. But there is a choice
point here because there are enough differences as well as enough
similarities for a reasonable case to be made for going either way on
it. I remarked in the earlier message that the success of the LANL
system raised some questions not hitherto raised, and this is one of
them. Should we say that peer review routinely does occur at LANL
itself? But whichever way that goes, we can see that the fear of the
LANL system as something inducing corruption that will inevitably spread
to corrupt peer review practices in general are really quite
groundless: if it doesn't incorporate peer review proper it certainly
incorporates peer critical control in quite as direct a way as does the
traditional system of peer review in connection with the journals.
This should make it all the more obvious that, as I was urging in the
prior message, the creation of server systems designed as mechanisms of
primary publication is one thing, and the creation of server-based open
discussion forums is another: the functional confusion of these can
only retard the development of both communicational forms and their
corresponding practices, and perhaps of other useful communicational
forms and practices as well, and, insofar as it does so, the future of
internet-based communication is severely and unnecessarily limited.
--
Joseph Ransdell <ransdell_at_door.net> or <bnjmr_at_ttu.edu>
Dept of Philosophy Texas Tech Univ. Lubbock TX 79409
(806) 742-3158 office 797-2592 home 742-0730 fax
ARISBE:Peirce Telecommunity http://www.door.net/arisbe
http://www.door.net/arisbe/homepage/ransdell.htm
Received on Wed Feb 10 1999 - 19:17:43 GMT