Re: Self-Archiving Refereed Research vs. Self-Publishing Unrefereed Research

From: Stevan Harnad <>
Date: Fri, 10 Aug 2001 12:48:12 +0100

On Thu, 9 Aug 2001, Arthur Smith wrote:

> > On Tue, 7 Aug 2001, George Lundberg wrote:
> > > [...] I would be greatly averse to having
> > > my doctor treating me based on some ??published self-archived article.
> > > Therein lies one of the principal rubs in this discussion...
> sh> Or is this still just the usual conflation of the self-archiving of
> sh> refereed research with the self-publishing of unrefereed research?
> Who does the un-conflating? The authors? How do readers trust them? Or
> do readers have to un-conflate themselves?

Precisely. And I strongly advise all participants in this discussion,
and all thinkers on this topic, to pause for a moment and reflect on
this, because unreflectively continuing to conflate self-archiving with
publication (and hence also putting them into false opposition)
completely obscures what is really going on here:

(1) There is a refereed research literature which is currently behind a
fee-based, access-blocking firewall (S/L/P), the average
refereed-journal paper (out of 2 million+ annually) being inaccessible
to many if not most of its potential users worldwide. Only those
researchers lucky enough to be at institutions that happen to be able
to afford to subscribe to the particular journal (out of 20K+ journals
in all) that that paper happens to be published in will have access.
Every case of lost access is a loss of potential impact for the author
and a loss of potential research productivity for the potential user,
and for research itself.

(2) To try to remedy this, authors, even in the on-paper era, used to
distribute their refereed research as much as they could on their own
initiative, sometimes by mailing publisher-purchased offprints to
reprint-requesters, sometimes by mailing photocopies of the final
accepted manuscript. Every otherwise access-denied user who gains
access through author-initiatives like this represents a net gain in
impact for the author and in potential progress for research.

(3) The very same author-initiative to maximize the impact of their
research is the motivation behind self-archiving. Authors self-archive
so that the access-denied subset of the potential usership of their
paper -- all those researchers, present and future, who do not happen
to be lucky enough to be at an institution that can afford paid access
to that particular journal -- can access that particular article (out
of 2M+ annually).

(4) Yes, the authors are "self-certifying" that their archived version
is indeed their refereed final draft (just as they did with their
photo-copied typescripts). Yes, some authors may lie (though it's not
at all obvious why they would want to! whereas it is abundantly obvious
why they would want to unblock access and impact by self-archiving...).

(5) So what? As a special case, all the lucky users who had enjoyed
access to the publisher's official version still have access, and are
"safe." Those with no prior access now have access to the author's
unofficial version (some subset of which might conceivably be bogus, or
careless). Is there ANYTHING sensible to say about this, other than: Go
ahead, self-archive, but (as usual), Caveat Emptor when you use a
self-archived draft rather than the "official" one?

The only reason people imagine otherwise is that they are conflating
publication with self-archiving, not realizing that the latter is a
SUPPLEMENT to the former, not a SUBSTITUTE for it.

    Harnad, S. (2001) The Self-Archiving Initiative. Nature 410:

> And I think George Lundberg's
> point is more than just the mixing of refereed and un-refereed research
> - even among that which is refereed, is it responsible to present
> essentially in the same context work which has passed only the minimal
> review appropriate to high-level research communication, and work that
> has been editorially selected and repeatedly reviewed as suitable for
> presentation and recommendation to physicians for use in practice?

I cannot follow this, about "responsibility" and "context," at all. We
are talking about authors self-archiving their final, refereed draft,
the one that has been peer-reviewed, revised, finalized and accepted.

In the on-paper "context", the journal-name tag was sufficient to
identify this refereed primary research literature. Are you suggesting
that that same tag can't do exactly the same work in the on-line
"context" with the ("irresponsible"?) author's own self-archived
version of the very same final refereed drafts? Or are you referring to
other kinds of (refereed) articles, such as reviews? (But can't those
be self-archived too?)

I've heard this kind of worry expressed many times, but it sounds
equally arbitrary and trivial every time. (And, I must add, it also
sounds like special pleading, to defend the indefensible: It is not,
and never has been, the fee-based toll-gates that are protecting public
health! It is and always has been, the gate-keeping, namely, peer
review, which has nothing whatsoever to do with 70-90% of the
access-blocking gate-tolls that must currently be paid for access to
the refereed paper).

To free this literature is to free access to it, not to tranform or
reduce it into something else.

    Harnad, S. (2000) E-Knowledge: Freeing the Refereed Journal Corpus
    Online. Computer Law & Security Report 16(2) 78-87.

> That's not to say that author self-archiving is inherently bad - it's
> just that there is real value in the "un-conflating" or more accurately
> "selection" of "filtering" provided by journals, beyond "refereeing".

This is not un-conflating, it's yet another conflation, this time of
the toll-gating (S/L/P access fees) with the gate-keeping (peer
review). Minus the optional add-ons (on-paper versions, publisher's
PDF, publisher's online enhancements), a published, refereed journal
article is merely a refereed, revised, accepted final draft, certified
as such by the journal-label.

Now exactly where and what is the "selection" and "filtering" over and
above this?

> And that value is inherently linked to the control and distribution of
> the material involved, so S/L/P is a natural funding mechanism - though
> we among others hope we'll be able to find new ones.

The only "control" that is ESSENTIAL is: peer review. Distribution is
merely an optional add-on now that we are in the PostGutenberg Galaxy,
where author/institutions can publicly self-archive their (refereed)
"material" online for free for all forever.

Yes, the true cost of the essentials will still have to be covered, but
let us not forget that peers review for free, so all we are talking
about is the peer-review implementation costs. On your (high-end)
estimate that these amount to 30% of current S/L/P tolls, it is still
obvious that they could be covered out of less than 1/3 of the 100%
S/L/P savings, as outgoing peer-review service charges to
author/institutions, instead of incoming access charges to
reader/institutions -- with the added bonus that the entire refereed
literature will be free.

But while the S/L/P market is going strong, and covering all costs,
author self-archiving can proceed apace. There is nothing whatsoever to
wait for: Self-archiving is a supplement, not a substitute, for
refereed-journal publication. Go ahead and self-archive it all, now!

    Harnad, S. (2001) For Whom the Gate Tolls? How and Why to Free the
    Refereed Research Literature Online Through Author/Institution
    Self-Archiving, Now.

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 the ongoing discussion of providing free
access to the refereed journal literature online is available at the
American Scientist September Forum (98 & 99 & 00 & 01):

You may join the list at the site above.

Discussion can be posted to:
Received on Wed Jan 03 2001 - 19:17:43 GMT

This archive was generated by hypermail 2.3.0 : Fri Dec 10 2010 - 19:46:12 GMT