Re: Why hep-th has 40% red-links

From: Stevan Harnad <>
Date: Tue, 12 Sep 2000 17:36:47 +0100

On Tue, 12 Sep 2000, Tim Brody wrote:

> (Note: Red-links = citations to LANL pre-print reference number, e.g.
> hep-th/9906001, may or may not also contain published data)
> > Stevan:
> > Elite string theorists are a small, specialized group. Their numbers
> > and stature are about comparable with the scale of all of science in
> > the 17th/18th century, where the few practitioners world-wide (Newton,
> > Leibniz, etc.) at any time could communicate their research by simply
> > writing letters to one another.
> >
> > This is neither representative of research as a whole today, nor will
> > it scale (in my opinion).
> hep-th has 6000 authors, hep-ph has 7500 authors, with 13000 and 17000
> papers respectively. The difference in the number of "red-links"
> identified is respectively 40% and 20%.
> This represents a large and active group within LANL, although, as you
> say, the quoted article only relates to a small group this could be
> representative of the larger hep-th physicists who'se overall behaviour
> results in double the number of red-link citations.
> What other explanation(s) could there be for for a large difference in
> citation patterns?

Good point. And the question -- about whether or not the behavior is
representative of Physics as a whole, whether it scales to larger
numbers, and whether it generalizes to other disciplines -- is an
empirical question about which we have only hunches now.

The answer will come when more disciplines are self-archiving, and your
very revealing "digital embryological" techniques are applied to the
larger corpus!

> [cue argumentative as opposed to empirical]
> This behaviour does not need to scale, hep-th and hep-ph have been
> virtually static in the number of deposits since 1995 (growth has come
> from other areas), and the citation patterns have been relatively static
> since 1998. Although these "red-link" citations could also be citing
> published articles, it would appear to be the "settled" behaviour of 40%
> of citations being to LANL pre-prints, surely this must be a change away
> from citations in the peer-review world to citations in the e-print,
> pre-print world?

It might be; but don't forget the Invisible Hand! Virtually all those
preprints are written for, and submitted to refereed journals. The
analogy is that people are dressing with the anticipation of an eventual
inspection. That answerability tends to make people keep there socks up.
The fact that they are presentable even before inspection does not
predict how they would look if there were no inspection ahead.

    Harnad, S. (1998/2000) The invisible hand of peer review. Nature
    [online] (5 Nov. 1998)
    Longer version in Exploit Interactive 5 (2000):

> > Stevan:
> > Remember "Simon-says": We should definitely find out (but not
> > necessarily believe) what people SAY they are doing, and why.
> > We should also find out what they DO do, and what others do/say too.
> >
> > Then let's piece together the picture objectively.
> >
> > The string theorists are definitely a piece of the whole picture,
> > but equally definitely not a representative microcosm of it!
> hep-th and hep-ph are the most self-contained and long-standing areas of
> LANL, the behaviour of HEP authors may not represent medics or computer
> scientists but they may show the relative effect that instantly available,
> unrefereed articles could have on the research world.

You may be right; but again, this is preprint-user behavior in a world
WITH the Invisible Hand: It is not clear whether it is predictive of
what things would look like in a world WITHOUT the Invisible Hand.

> > Stevan:
> > [Nor is "theory" in general the dividing line, I think, for there are
> > more and less populated, more and less elite areas of theory too --
> > in my (Stevan-says) opinion...]
> But (feel free to correct me), hep-th is the primary digital source for
> theoretical physicists, and that is where theoretical physics research is
> being done.

But don't you agree that Theoretical Physics would need to be looked at
area by area, taking account the size of each area. One could get an
effect that looks like the behavior of a population of 5000 theorists,
when in reality it is the behavior of 50 sub-populations of 100
theorists each! (You may be in a position to answer this and related
questions with your OpCit analyses.)

Stevan Harnad
Professor of Cognitive Science
Department of Electronics and phone: +44 23-80 592-582
             Computer Science fax: +44 23-80 592-865
University of Southampton
Highfield, Southampton

NOTE: A complete archive of this ongoing discussion of providing free
access to the refereed journal literature is available at the American
Scientist September Forum (98 & 99 & 00):

You may join the list at the site above.

Discussion can be posted to:
Received on Mon Jan 24 2000 - 19:17:43 GMT

This archive was generated by hypermail 2.3.0 : Fri Dec 10 2010 - 19:45:51 GMT