Re: Journal Publishing and Author Self-Archiving: Complementary Or Competitive?

From: Stevan Harnad <harnad_at_ecs.soton.ac.uk>
Date: Fri, 26 Aug 2005 14:59:45 +0100 (BST)

> Don Sannella (DS): "a significant fraction of citations in published papers
> [were]... not read [by the author citing them]."

No doubt. And there may even exist a significant number of citations of papers
that don't exist. But the point being made here is not about scholarliness or
unscholarliness. There is an empirical effect to be accounted for: Papers that are
self-archived are cited 50-250% more than papers (published in the same journal
and year) that are not self-archived.

http://citebase.eprints.org/isi_study/
http://www.crsc.uqam.ca/lab/chawki/graphes/EtudeImpact.htm

That is not only a (statistically) significant percentage, it is a (numerically)
substantial percentage. By what causal chain does DS imagine that providing free
online access to a paper leads to more researchers citing that paper without their
actually reading it -- compared to equivalent papers (same journal/year) to which
free online access is not provided?

(It's not impossible; it no doubt happens sometimes: The title and author may look
relevant to a paper in progress, so they get added into some parenthesis. But
please, let's be realistic about what proportion of the 50-250% advantage is
likely to arise from that sort of behaviour, rather than making too much of a
minority practice. Surely the less far-fetched inference is that the free
full-text access is used to read papers that the user could not otherwise afford
to access, rather than just to cite them, unread. One can, after all, access the
author/title citation metadata in a free research index like Scirus
http://www.scirus.com/srsapp/ (which does not offer full-text access to items that
are not self-archived) or even Google Scholar http://scholar.google.com/ if one is
merely bent on citing papers sight unseen. That leaves the 50-250% advantage
unexplained.)

>> SH: "Is DS seriously proposing that in the Internet age... the researcher
>> whose institution cannot afford a subscription should walk to a library
>> (somewhere) for every click performed by the researcher whose institution can
>> afford a subscription?"
>
> DS: "I agree that it is more convenient to have access without walking
> anywhere... [but] Online access was not the topic."

On the contrary, online access was very much the topic: Open access is defined as
free online access: http://www.earlham.edu/~peters/fos/boaifaq.htm#openaccess It
does not concern access to the print edition (although users can of course print
off the free online version for themselves if they wish). It was the advent of the
online era that opened up the radical new possibility for researchers to maximise
the access to (and thereby the impact of) their research by supplementing the
publisher's paid versions (both print and online) with a free online version of
their own final drafts for those would-be users whose institutions cannot afford
access to the paid version.

Hence the issue is not merely one of "convenience," since research access is the
researcher's daily bread, time is money, and it would already be absurd to imagine
relegating the have-not researchers to slogging over to each paper with their feet
while their colleagues and competitors are doing it instantly with their fingers.
But even that is not the point, for the target constituency here is those
researchers whose institutions cannot afford paid access to either version, print
or online. (That was what earned DS the not-undeserved quip about Marie
Antoinette, to which I will shortly return.)

> DS: "I would guess that the vast majority of researchers already have
> access."

To test your quess, please look at the Ulrich's data on the number of journals
that exist, across disciplines, worldwide:
http://www.ulrichsweb.com/ulrichsweb/analysis/
and compare it with the number of journals that institutions can actually afford:
http://fisher.lib.virginia.edu/arl/index.html

Again, the 50-250% impact advantage of the articles that have author self-archived
supplements would seem to imply that the "minority" that do not already have paid
access (because their institutions cannot afford it) may be a pretty vast one too.

(See also the "Sitting Pretty" FAQ:
http://www.eprints.org/openaccess/self-faq/#29.Sitting )

>> SH: ... "(and what about the rest)?"

> DS: That's what Interlibrary Loan was designed for.

(If that does not call to mind "Qu'ils bouffent de la brioche," I don't know what
will!)

The point is that paid access refers to all forms of payment: subscription,
site-license, and pay-per-view (which includes InterLibrary Loan). And "can't
afford paid access" means: "can't afford access by any of these three means."
(Please consult the ARL statistics again.)

>>> DS: "How can we be sure scholars will cite the right version, and
> that statistics will be pooled and credited to the published version?"
>
>> SH: "Leave it to the research community. They stand to lose even more
>> than publishers if wrong versions are cited and citations are not credited to
>> their publications."
>
> DS: "I'll believe it when I see it. "They stand to lose" refers to authors
> whose work is being cited. "Wrong versions are cited" refers to different people,
> namely, authors of papers that cite this work. "Citations are not credited" refers
> to the people who control institutional repositories."

Correct, and in the online age, with the full-texts of all articles freely
accessible online, versions and citations will all be monitored,m coordinated and
optimised. Not to worry. And certainly not to use this worry as a pretext for not
making all those articles freely accessible. Clearly this is a research community
matter, between authors and authors (and their employing institutions and
funders), not the ALPSP.

> DS: "The rebuttal argues one thing ["Publishers giving away the full contents
> of the journal version (A) is not the same as authors self-archiving their final
> drafts (B)"] ...and then asserts a stronger thing (A has nothing whatsoever to do
> with B) as if these two things were the same, and they aren't. But I think this is
> a minor issue..."

Yes it is a minor issue -- except if publishers try to use evidence from the one
(full-journal give-aways) as grounds for delaying or diverting the other (single
author-draft give-aways), as the ALPSP did. It is the evidence from the one that
has nothing to do with the other.

Stevan Harnad
Received on Tue Aug 30 2005 - 11:59:57 BST

This archive was generated by hypermail 2.3.0 : Fri Dec 10 2010 - 19:47:59 GMT