Re: Elsevier's ChemWeb Preprint Archive

From: Stevan Harnad <>
Date: Tue, 3 Oct 2000 15:52:29 +0100

Dear James,

Some comments: as I've taken some time with these, I'd like you
permission to post them to some of the relevant lists on this topic
(such as my AmSci Forum).

On Mon, 2 Oct 2000, Weeks, James (ELSLON) wrote:

> First of all, I think it is important to note that we also believe that the
> preprint server should act as a permanent archive of submitted articles.

Yes, but my point was that it should be (1) a permanently (and freely)
ACCESSIBLE archive, (2) containing not just submitted articles
(unrefereed preprints) but also accepted articles (refereed

Your reason (below) for not going this far makes reference to (i)
embargo policies (the "Ingelfinger Rule") plus (ii) copyright
restrictions. But there is a simple (and legal) way to get around both
of these (see later):

> Chemists often communicate in different ways to many other scientists and to
> achieve our goals we are slowly having to change the mindset of chemists.

There is only one pertinent aspect of the relevant mindset, and
chemists do not differ in any way from all other scholars/scientists in
this sole relevant respect: ALL authors of refereed research papers
give them away (to the publisher), none receive or wish to receive
fees/royalties, and none benefit from or wish to have access barriers
to this give-away refereed research, because access-barriers are
impact-barriers, and no one benefits from those.

So don't look for preprint-related peculiarities among disciplines:
focus on the only relevant variable: Is there any discipline that would
NOT benefit from the elimination of all access-barriers to its refereed
papers on-line?

> Many chemistry publishers, however, are a potential barrier to the concept
> of preprints in chemistry. Many will not accept papers for submission which
> have already been preprinted elsewhere, and others will not accept the paper
> if it remains available in any other form.

You are referring here not to copyright policy, which is a legal matter,
but to "embargo" policy -- which is just that. A journal can have all kinds
of arbitrary policies about what papers it announces that it will and
won't referee. (It could decree that it will not referee a paper by an
author who as ever collaborated with any author who has ever submitted
a paper to any other journal.)

Such arbitrary policies fall into two categories, (1) the justifiable
ones (such as declining to referee a paper that has already been
published elsewhere, or is in the wrong area)) and the unjustifiable
ones (such as declining to referee a paper that has been previously
circulated to peers in any form, whether on-paper or on-line). I will
not discuss the justifiability of the "Ingelfinger Rule" here. I have
written on it before, e.g.:

    Harnad, S. (2000) E-Knowledge: Freeing the Refereed Journal Corpus
    Online. Computer Law & Security Report 16(2) 78-87. [Rebuttal to
    Bloom Editorial in Science and Relman Editorial in New England
    Journal of Medicine]

    Harnad, S. (2000) Ingelfinger Over-Ruled: The Role of the Web in
    the Future of Refereed Medical Journal Publishing. Lancet (in

There is no need to spend time on the justifiability of the rule (it is
not justifiable, and when examined, it is obvious that it has only one
purpose, to protect the vendor's current revenue-streams). It has
nothing to do with the interests of science, with which it is directly
at odds, being an impact-barrier.

But the relevant variable is not justifiability: It is enforceability.
The Ingelfinger Rule is not only not a legal matter, but it cannot be
enforced even as a matter of (unjustified and unnecessary) policy.
Authors have only to self-archive their preprints with some cosmetic
changes that prevent them from matching the submitted draft; this could
only draw journal editors into a slippery slope of defining how much of
a difference is difference enough! Moreover, journal editors would have
to consent to becoming 24-hour-a-day Web-sleuths, trawling for look-alikes
for every submission (and, being scientists, like the rest of us, I
doubt they would allow themselves to be drawn into that). And even that
could be foiled by encrypting the archived version and revealing the
key elsewhere, until the refereeing is underway. The needle in the
haystack would be impossibly small.

So this is all nonsense. The Ingelfinger Rule is just a Wizard-of-Oz
phenomenon: It can only make authors refrain from self-archiving their
preprints if they are foolish enough to take it at face value, and
believe it is justified and enforceable. (Physicists -- who I believe were
faster off the mark than the other disciplines only because they are
smarter and more serious than the rest of us -- were not deterred for
one microsecond by such absurd policies -- even though some of their
journals, e.g., Science, have them.)

> What we had initially planned to
> do, and in fact what we are doing, is only allowing the author to remove the
> full text of his/her paper if they provide details for us to create a
> hyperlink to an online version of the published article. This is done via
> software created by our sister company, MDL, called LitLink.

Now what on earth is the benefit of that? I go to get the paper, and I
am redirected to a vendor's financial firewall, with tolls blocking
access/impact in the form of Subscription/Site-License/Pay-Pare-View
(S/L/P). Vendors should be delighted with that, as all it represents is
free advertising for their product, instead of freeing access to it.
But what benefit is it to the researchers who wrote those give-away
papers, and to research itself, to countenance those obsolete
impact-barriers any longer?

> All of the
> other details concerning the original submission to CPS, including the
> author name, affiliation, article title, abstract, and also the discussion
> about that article remain accessible on the CPS. Furthermore, the original
> articles are never erased from the server and can be used to contest issues
> of prior art.

Meta-data are welcome, but they are "cheap": Primary journal publishers
already give them away for free to secondaries (abstracting, indexing,
citation services) because as long as they hold the full-text itself
hostage behind their S/L/P firewall, it is again just a form of free
advertising for what they are selling.

But why should the authors, who are not selling anything, have to
subsidize those needless tolls with lost impact for their ideas and

(The only ESSENTIAL cost to cover is much smaller than the S/L/P tolls:
It's just the cost of implementing peer review -- 10% of the total
S/L/P outlay. The rest -- the on-paper version, the publisher's PDF,
deluxe on-line enhancements, etc. -- are, and should be sold as, add-on
OPTIONS only, not held hostage to obligate tolls under the self-serving
rhetoric of "value-added" (take-it-all or leave-it-all). If those
add-ons really have a market, they should be able to find it with no
coercion. The self-archived "vanilla" refereed draft MUST be freed.

(A permanently archived, time-stamped, authenticated, encrypted draft of
the preprint, for the purposes of establishing priority claims, is all
well and good -- but what is the justification for keeping it

> There is also nothing to stop an author posting an article on the CPS which
> has been submitted, refereed, accepted or even rejected by a journal
> publisher. The CPS policy is simply not to accept an article where the
> copyright has already been transferred to a publisher, unless that publisher
> supports the CPS.

The Harnad/Oppenheim strategy can get around this competely, and
completely legally:

First self-archive the preprint (no copyright issue there). And then
simply, in the worst case, after unsuccessfully asking the publisher to
allow the refereed postprint to be self-archived, having transferred
all other rights, on-paper and on-line, for-free and for-fee -- 70% of
publishers will agree if asked, 10% already do, but for the 20% who
won't agree: simply self-archive the "corrigenda" in a file linked to
the already-archived preprint.

The strategy is also described in:

    Harnad, S. (1999) Free at Last: The Future of Peer-Reviewed
    Journals. D-Lib Magazine 5(12) December 1999

Completely legal, and always elicits a surprised chuckle from audiences
when I describe it, so deeply entrenched is our habit, so far, of
conflating the author NON-give-away literature (royalty/fee-based) --
which happens to be the lion's share of the literature, and the only
possible model in the Gutenberg Era, with it's associated costs with
the author-give-away literature, in the PostGutenberg Galaxy of
Scholarly Skywriting.

> I hope these comments help to explain our policies. We would of course be
> most happy for you to comment further on the CPS, and maybe discuss it with
> you further. Bryan Vickery, Community Development Manager of
> will be attending the International Chemical Information Conference in
> Annecy this month, where I believe you are speaking.

Looking forward to seeing Bryan there:

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):

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 - 19:45:53 GMT