Re: Study Identifies Factors That Could Lead to Cancelled Subscriptions

From: Stevan Harnad <>
Date: Mon, 11 Dec 2006 22:48:03 +0000

On Mon, 11 Dec 2006, David Goodman wrote:

> Are you seriously giving as a positive feature of self-archiving in
> institutional repositories that it is relatively inefficient and relatively
> confusing practice?

Nothing of the sort. You are conflating the question of efficiency
for the researcher-author and researcher-user, with the question of
efficiency for the publisher or the library. I am speaking only from
the researchers' viewpoint -- which is the *only* viewpoint relevant to
the question of Open Access self-archiving, by and for researchers,
and its efficiency in maximizing research access, usage and impact.

I said that unlike the OA content of a journal that goes gold -- 100%
-- the OA content of a green journal is subtotal, not easy to tally,
and grows anarchically and distributedly (depending on which of its
authors happen to self-archive, where, and when).

Self-archiving -- and especially mandated self-archiving -- is
highly efficient and not in the least confusing: for authors and
users. Anarchically distributed self-archived content, however, and
tallying it, is indeed confusing and inefficient: for publishers and

> But in another posting today* you praised them, as their
> findability was now fully equal to centralized repositories, and you
> called the most successful of all the centralized archives "obsolescent."

Correct. The user that is looking for a self-archived article will find it,
easily and reliably.

To repeat, what is uncertain at any point is what percentage of any
given journal's contents happens to be self-archived so far, and how fast
that percentage is growing. But those are not author or user concerns --
i.e., not researcher concerns. They are publisher and librarian concerns.

There are systematic, automatised ways to probe the percentage of
journal content that is self-archived, and how fast it is growing, for
example, by feeding a journal's metadata, issue by issue, into OAister,
Google Scholar and Google, checking the accuracy and the percentages,
and constructing and regularly updating a time-plot, issue by issue,
daily, weekly, monthly, or yearly. There is some possibility that a
journal publisher might take the trouble to do this, but it is unlikely
that they will want to publicise the outcome. Librarians are unlikely
to do this, and if they did, it's not clear they would know what to do
with the outcome.

Let me add that at this late date, after more than a decade of everybody's
swilling them around hypothetically without doing anything about them,
I have no further interest in these questions from either the publisher's
or the library's point of view. My own interest is now solely from the
researchers' point of view, and for researchers the problem is to get
all the research self-archived -- in IRs -- so they can all access it,
at long last. If it's self-archived, they'll find it, efficiently, and
without any confusion. (Please let's not start with the "problem" of
search engine efficiency again: There is no problem with today's search
engines. The problem is with their impoverished content: only about 15%
of the current research corpus. Crank that up to 100%, and not only will
the problems vanish, but the search engines will crank up their powers
creatively to levels unimaginable in the current content wilderness --
powers and levels not worth providing today, to be applied to next
to nothing.)

And, yes, CRs are indeed already obsolescent; but they will only obsolve
in earnest when the self-archiving mandates finally bite; that may be
after an initial false-start period when some of the mandates foolishly,
and short-sightedly, mandate self-archiving in a particular CR, instead
of in each researcher's own IR.

    Optimizing OA Self-Archiving Mandates: What? Where? When? Why? How?

   "Written evidence for UK Select Committee's Inquiry into Scientific
    Publications" (Dec 2003)

    "What Provosts Need to Mandate" (Dec 2003)

    "A Simple Way to Optimize the NIH Public Access Policy" (Oct 2004)

    "Please Don't Copy-Cat Clone NIH-12 Non-OA Policy!" (Jan 2005)

    "Maximising the Return on UK's Public Investment in Research"
    (Sep 2005)

    "Generic Rationale and Model for University Open Access
     Self-Archiving Mandate" (Mar 2006)

    "How to Counter All Opposition to the
    FRPAA Self-Archiving Mandate" (Sep 2006)

To repeat: IRs and CRs are OAI-compliant and equivalent, but each
researcher's institutional IR is the natural locus for each researcher's
own research output. Each researcher's institution is the primary research
provider. Universities consist of all disciplines. Universities can
mandate self-archiving and monitor compliance in their own IRs, locally,
with their own mandates, reinforced by the mandates of the funders of
their research. CRs, in contrast, are no sort of entity at all. They do
not cover all disciplines, and even if they did, it would be absurd to
have researchers at the same institution self-archiving left and right
in repositories other than their own, in an age when the OAI harvesting
protocol was created precisely to make that sort of thing unnecessary --
indeed obsolete.

IRs are the natural and optimal locus for direct deposit; CRs (if any)
can and should harvest from the IRs. And the mandates should be uniformly
and consistently directed at the researchers' own IRs, not at depositing
willy-nilly in CRs.

Stevan Harnad

> ----- Original Message -----
> From: Stevan Harnad <harnad_at_ECS.SOTON.AC.UK>
> Date: Sunday, December 10, 2006 11:08 pm
> > On Sun, 10 Dec 2006, Sally Morris (Chief Executive) wrote:
> >
> > > As I hoped, a publisher has come up with some real figures about
> > > the effect of going OA after a short embargo. See below from
> > > PNAS (forwarded with Diane's permission).
> >
> > Dear Sally:
> >
> > Let's keep our eye on the ball: The question is and has always been:
> > Is there any evidence that self-archiving (green) causes
> > cancellations?
> > Answer is still: No.
> >
> > The PNAS report below is about making the journal freely accessible
> > (gold). That makes all of its contents, publisher's version, at the
> > publisher's website, free for all (gold) (within a month).
> >
> > I, for one, have never doubted that *that* could cause
> > cancellations. But
> > anarchic author self-archiving, of each author's postprints, in each
> > author's own IR, in uncertain proportions and at uncertain rates, are
> > another story.
> >
> > (But if/when mandated self-archiving should ever prove to cause
> > cancellations after all, publishing can and will adapt; research
> > should certainly not renounce
> > its impact in order to insure journals' current modus operandi
> > against all risk
> > from the new medium!)
> >
> >
> >
> > > I wonder whether there are other publishers on this list who have
> > > statistics they could share?
> >
> > Let's hope that if they do, their stats will be to the point
> > (green), rather than
> > off-topic (gold)!
> >
> > Chrs, Stevan
> >
> > > Sally Morris, Chief Executive
> > > Association of Learned and Professional Society Publishers
> > > Email:
> > > Website:
> > >
> > > ----- Original Message -----
> > > From: "Sullenberger, Diane" <>
> > > To: "Sally Morris (Chief Executive)" <>
> > > Sent: Monday, December 04, 2006 5:32 PM
> > > Subject: RE: Study Identifies Factors That Could Lead to Cancelled
> > > Subscriptions
> > >
> > > Hi Sally,
> > >
> > > In 2000, we were free after one month. We lost 11% of our paid
> > > subscribers in 2001, higher than the industry average, and we
> > > switched to 6 months in 2002. The move did not stem the loss in
> > > subscribers but it was reduced to 9% in 2002. We do not have hard
> > > data to show a causal effect of our one month policy, but the
> > > correlation certainly motivated a change.
> > >
> > > Best,
> > > Diane
> ==other posting by SH today:
> ----- Original Message -----
> From: Muriel Foulonneau <muriel.foulonneau_at_CCSD.CNRS.FR>
> Date: Monday, December 11, 2006 9:55 am
> Subject: Re: [AMERICAN-SCIENTIST-OPEN-ACCESS-FORUM] Central versus institutional self-archiving
> > Are there statistics supporting a current decrease in number of
> > deposits on ArXiv or CogPrints?
> >
> > Stevan Harnad a écrit :
> > >> May I ask you when CogPrints was first created? What was its
> > official launch date?
> > >>
> > >
> > > CogPrints was launched August 19, 1997
> > >
> > >
> > > I might add that CogPrints (as well as Arxiv) are obsolescent as
> > > primary loci for direct deposit: Since 1999 (the OAI
> > interoperability> protocol) the distributed network consisting of
> > authors' own OAI-compliant
> > > Institutional Repositories (IRs) has become the natural and
> > optimal locus
> > > for direct deposit. Central Repositories (CRs) (like Arxiv and
> > CogPrints) if they
> > > perdure at all, will become harvesters from the primary research
> > providers> (IRs), rather than the locus where papers are deposited
> > directly. The
> > > same applies to PubMed Central.
> > >
> > > Depositing directly in a CR is as silly today as depositing
> > directly in
> > > Google! Citeseer is a better model for an OA-age CR than Arxiv,
> > because> it already is (and always has been) a harvester rather
> > than a direct
> > > locus for central deposit. OAIster is another example, and there
> > are more.
> > >
> > > Stevan Harnad
> >
Received on Mon Dec 11 2006 - 23:22:04 GMT

This archive was generated by hypermail 2.3.0 : Fri Dec 10 2010 - 19:48:39 GMT