Re: Open Access and ISI-indexed journals and articles

From: Stevan Harnad <>
Date: Wed, 3 Nov 2004 00:06:43 +0000

David Goodman writes:

> Before following Stevan's excellent advice...

but then he goes on to give advice that I not only could not describe as
excellent, but is rather the precise opposite of what I would advise on every
single point:

David advises that:

> ...the authors would be prudent
> if they checked the details of the publishers' current policy (which
> will probably be considerably more liberal than it was when they submitted
> the article).

I counter-advise that for the 92% of journals that are already green,
the author should not go to check the details and see if they have become
more "liberal": The author should just go ahead and self-archive!
(The ISI article on which this is based wisely ignored the yellow and
blue distinctions that David is urging us instead to pay more attention

For the 8% that are not yet green, I suggest self-archiving and writing to
the editor (not the permissions department!) of the journal to indicate
that you have done so, and why (to make sure all would-be users of your
article can access, use and cite your findings). Wait to hear whether
and what you hear in response. In most cases it will be nothing, and
that will be the end of it. If you hear something, then you can decide
whether you want to remove the article or negotiate further. Don't
negotiate in advance of self-archiving.

> The phrase "some form of author access" is, as appropriate, very
> inclusive. Only some of these publishers permit posting to an external
> server, such as arXiv or D-Lis, and limit the author to a server at his
> own institution. Some may even limit it to the author's personal home
> page. As unfortunately most institutions do not yet have IRs, the home
> page may be the only immediate legal choice.

I counter-advise that the mostly trivial, arbitrary or incoherent distinctions
along these lines should be completely ignored. The only ones to hew to are
whether or not you can self-archive your publisher's PDF. (If not, don't: just
self-archive your own refereed, corrected final draft. Better still,
make that your default option.) Ignore all distinctions about types of
servers at your own institution. They are all nonsense. And self-archive at
your own institution. All you need to do at a central server is deposit the
metadata, if you wish. Then link to the full-text self-archived in your
own institutional archive. End of story. No further details are relevant.

> They should also check exactly what form of the manuscript they are
> posting. Many of these publishers limi the posting to the author's
> manuscript as it was accepted by the editor after peer-review, but not
> including the changes introduced by further editing and production.
> (Depending on the author, they may be insignificant, or they may be
> extensive.) This limitation may, of course, be worded in varied and
> confusing fashion. I note that the best place to confirm the most current
> policy is the posted information for authors on the publisher's site.
> The information on printed forms, and, especially, printed in the journals
> is often out of date, sometimes by many years.

Counter-advice: Just self-archive your own refereed, revised, corrected final
draft as the default option. Don't bother even inquiring about the publisher's PDF
unless you really want to. It's an extra hassle for minimum gain. If you love PDF
so much, generate your own. All these other details are best ignored: Just go
ahead and self-archive.

> I and others have found that an excellent method to gain more liberal
> terms is simply to ask for them. As publishers receive more requests,
> they will probably eventually improve their policy to attract and retain
> their authors. This is one of the strategies that has helped bring about
> some of the progress that has been already attained. The Romeo site
> is helpful, but as the ISI found, it does not include all publishers.
> It has also sometimes not been altogether up to date, until corrections
> have been verified and incorporated.

Counter-advice: For the green 92% just go ahead and self-archive. Don't
waste time asking about details. For the remaining 8% self-archive and
notify (as above). Don't negotiate in advance, and only negotiate
afterward if there is any need for it.

> There do exist articles posted violating the publisher's copyright. There
> is already enough doubt among publishers that self-archiving will be
> done in a responsible and legal maner, that would make it advisable to
> be especially careful. We all look forward to when these restrictions
> will no longer need to be a concern.

Authors have been self-archiving their own final drafts since at least
1991, and the default strategy has been "don't ask, don't tell." Out of
300,000 articles self-archived in Arxiv for example, only 4 articles have
been removed citing copyright reasons (and it is not clear what those
reasons were) across all those years.

It was as a response to the growing desire of authors for OA
and the growing de facto practice of self-archiving (pre-green,
don't-ask/don't-tell) that publishers have given it their formal green
light. The rapid rate of growth of green in the past 3 years does not look
like evidence of *doubt,* but evidence of a natural adaptive response
to the growing demand for OA and the growing de-facto practice of
self-archiving by authors (now at about 20%). It's far easier and less
risky for a publisher to go green than gold, in response to this pressure
for OA.

Taking none of this into account, David simply advises authors to
focus more carefully on publishers' restrictions. I would counter-advise
that for the 92% that are green, go ahead and self-archive without giving
the details a further thought, and for the remaining 8% do the same,
but advising the journal that you are doing so. That's all.

This is not an assignment for a library permissions department, worrying
about what can be done with bought-in contents (and finding a colour-code for
every possible arbitrary stipulation). This is something entirely new:
It is about open-access provision, by authors, to their own give-away
work. Please don't try to fit it into old and irrelevant Procrustean
IP Beds, at the risk of wasting yet another 10 years on the road to the
optimal and inevitable (already long overdue).

Stevan Harnad

> -----Original Message-----
> From: on behalf of Stevan Harnad
> Sent: Tue 11/2/2004 3:20 PM
> To: AmSci Forum
> Subject: [OACI Working Group] Open Access and ISI-indexed journals andarticles
> "Open Access Journals in the ISI Citation Databases:
> Analysis of Impact Factors and Citation Patterns"
> ISI have reported (see excerpts below) that 395,052 (53%) of the 747,060
> articles indexed in the 2003 Journal Citation Reports were published
> in non-OA journals for which it is known that their publishers have
> given their authors the green light to make them OA by self-archiving them
> (56% if we add the 22,095 articles published in the OA journals). (ISI
> rightly ignores the superfluous yellow/blue subdistinctions, counting
> them as green.)
> This means, at the very least, that 56% of those ISI-indexed articles
> could be made immediately OA if their authors simply performed the few
> keystrokes needed to self-archive them.
> But the data are far stronger than that: ISI report that from the sample
> of ISI-indexed publishers for which their author self-archiving policy
> is known, 3056/3403 (90%) of the journals are green (which agrees quite
> well with the figures from ).
> If we assume (reasonably) that this sample can also be taken as an
> estimate of the percentage green among the remaining 2504 journals
> (whose publishers' self-archiving policy is not known), then a total of
> about 5316/5907 (90%) of the ISI-indexed journals are probably green.
> (Reckoned in terms of ISI-indexed articles, 417147/489824 (85%) of those
> articles come from the known green-journal sample, hence 635001/747060
> (85%) of them could already be self-archived by their authors with their
> publishers' blessing!)
> As neither (1) the publishers' green light nor (2) the growing
> evidence of the enhanced impact of OA vs. non-OA research
> yet seem sufficient to induce most of the authors of those articles
> to self-archive them (only about 20% are as yet doing so, according
> to our own estimates, gathered with the help of the ISI database),
> even though, for example, 34,000 authors signed an Open Letter
> demanding OA
> the time does appear to be ripe for a self-archiving mandate from
> researchers funders and employers in order to maximise the access
> and impact of their research output
> particularly as authors themselves, when surveyed, have declared that they
> will self-archive willingly if it is mandated (but not otherwise!) (Swan &
> Brown 2004).
> Stevan Harnad
> --------------------------------------------------------------------
> "Open Access Journals in the ISI Citation Databases:
> Analysis of Impact Factors and Citation Patterns"
> "The majority of publishers have only one or a few journals in the
> Thomson ISI citation databases, and were not listed on the Project
> Romeo site."
> "We found 133 publishers (and/or their subsidiaries) with information
> on author archiving policy. Of these publishers, 108 have a stated
> policy permitting some type of author archiving."
> "Along with the publishers of the covered OA journals, publishers
> supporting some form of author archiving produce nearly 52% of the
> journals in the 2003 JCR Science Edition.
> "It is possible that some of the publishers with no archiving policy
> yet listed on Project ROMEO would allow self-archiving by their
> authors, which would further increase the number of journals whose
> content is available for author archiving."
> "Because archiving is accomplished at the article level, we calculated
> the number of articles in journals that allow author archiving:
> 395,052 (53%) of 747,060 "citable items" in the 2003 Journal Citation
> Reports could be available, theoretically, for authors to post to
> individual or institutional archives. When the 22,095 articles and
> reviews in OA journals in 2003 are considered, the findings suggests
> that fully 56% of the article content indexed by Thomson could be
> deposited in one or more institutional archives."
> Visit the List Archives at:
Received on Wed Nov 03 2004 - 00:06:43 GMT

This archive was generated by hypermail 2.3.0 : Fri Dec 10 2010 - 19:47:39 GMT