: Re: Critique of Research Fortnight article on RCUK policy proposal

From: Stevan Harnad <harnad_at_ecs.soton.ac.uk>
Date: Wed, 5 Oct 2005 16:29:00 +0100

On Wed, 5 Oct 2005, Sally Morris wrote:

> On 16 September Stevan Harnad said on this listserv:
> "[Research Fortnight] The Institute of Physics has already seen
> article downloads from its site diminish for journals whose content is
> substantially replicated in a repository, says ALPSP.
> [SH rejoinder] This statement is false, and is the exact opposite of
> what the Institute of Physics has said (Swan & Brown 2005)"

Sally is quite right. What I should have said was that the diminished
article downloads do not equal, nor do they imply, diminished
subscriptions, and that IOP had said exactly the opposite: That
despite replication in a repository (ArXiv) IOP had found no diminished
subscriptions, does not consider self-archiving a threat, cooperates
with Arxiv, and indeed will soon be hosting a mirror of Arxiv.

> on 5th September in a posting from Ken Lillywhite [IOP]:
> "Recent claims made by some of those who advocate the mandatory
> deposit of published articles in institutional and subject
> repositories prompt us to correct and clarify the relationship
> between physics journals and the physics e-print archive (arXiv).
> "It has been claimed that because physics publishers have
> co-existed with the arXiv over the past 14 years, publishers
> have nothing to fear about the future. We take the opportunity
> to point out that the past is not always a good predictor of
> future performance.

No, but as David Hume would remind us, the past is still the best predictor
of the future, not the opposite. But let me quickly agree that anything is
possible: That the 14-year peaceful co-existence between self-archiving and
subscription sales continues, or that it does not. If self-archiving should
reduce subscriptions, there will be ways for journals to adapt. Renouncing
the maximised access and impact -- and their benefits to research and
researchers -- by *not* self-archiving is not an option, regardless of which
course the future takes.

> "Ever since the launch of the physics e-print archive in 1991,
> authors publishing in IOP Publishing journals have had the choice
> to post their preprints to the service. However, we do note
> that article downloads from our site are significantly lower
> for those journals whose content is substantially replicated in
> the arXiv repository than for those which are not, after usage
> statistics have been normalized to take account of journal size.
> "Usage statistics (e.g., ProjectCOUNTER) are now increasingly
> used as a 'value for money' measure in the library community and
> elsewhere. Clearly, as usage statistics become more commonplace,
> it would be only natural for cash-strapped librarians to
> conclude that subscriptions to low-use - albeit high-quality,
> peer-reviewed - journals are no longer necessary. In this
> situation subscription-based journals published by a learned
> society such as ourselves would become economically unviable."

And if that so-far-counterfactual speculation should happen to prove true,
there are many speculations available as to how journals and learned societies
can and will adapt.


What is fact, and not speculation, is that self-archiving enhances
citation counts by 50%-250%, and that only 15% of researchers are
self-archiving spontaneously today. It follows that a self-archiving
mandate will remedy this needless impact loss, and should hence be
adopted as soon as possible.

> What's more, there is anecdotal evidence that some librarians, at least,
> do have both the power and the inclination to cancel subscriptions to
> journals which are not being used. See, for example, Bob Michaelson's
> (Northwestern University) posting to PAMNET on 15 September:
> "Surely you know that physicists use the free depository arXiv as their
> primary source for physics information - to such an extent that some
> long-established physics journals such as Nuclear Physics now have
> scarcely any readers (we know this because we have the online
> statistics)... " Bob subsequently confirmed to me that he does,
> indeed, use low usage statistics as one reason for cancellation.

The fact is unchanged that neither APS nor IOP have detected any cancellations
associated with self-archiving -- nor are they attempting (as Sally and the
ALPSP are) to try to persuade RCUK not to mandate self-archiving. Librarian
anecdotes there will always be, but it is the actual data on self-archiving
and subscriptions, across journals and institutions, that tell the true tale.

> I am assured that the IOP spokesperson to whom Alma Swan spoke included
> the all-important word 'yet' - as in 'we have not seen greater
> cancellations yet'. Other publishers tell me that they have also
> observed significantly lower usage for journals most of whose content
> can be found in arXiv.

"Yet" can quite safely and reasonably be appended to everything I have seen and
heard, and it makes not a whit of difference. (One could, after all, even
add "yet" to Newton's observation about apples not falling up but down: All
empirical data are merely data to-date, not proofs about the course of the

> Thus my statement was not false. If Stevan had been aware of the
> evidence upon which it was based, as I assume he was, then he should
> have known that my statement was not false and should, I feel, apologise
> publicly to ALPSP in consequence.

I apologise to ALPSP publicly: I should have said:

    "This statement [that IOP finds diminished downloads for self-archived
    articles] is perfectly true but in no way implies what ALPSP cites
    it to imply (i.e., that diminished downloads are evidence that
    self-archiving causes cancellations), for that is the exact opposite
    of what the Institute of Physics has said (Swan & Brown 2005)."

Now that Sally has been so focussed and scholarly on this one point,
which has no implications whatsoever for the substantive matters at
issue, it would be very gratifying to hear her replies to the many,
many other substantive points I and others have made about the absence
of evidence in support of her claim "that RCUK's proposed policy will
inevitably lead to the destruction of journals."


Stevan Harnad
Received on Wed Oct 05 2005 - 17:52:58 BST

This archive was generated by hypermail 2.3.0 : Fri Dec 10 2010 - 19:48:03 GMT